While existing research recognizes that policy conditions can influence the link between climate change and agricultural output, a critical question has long been overlooked—even if policy direction is correct, if policies are volatile and unpredictable, they can themselves become a source of risk. How this policy uncertainty alters the relationship between climate change and agricultural production lacks in-depth exploration in academia. This study aims to fill this critical gap by specifically revealing how Climate Policy Uncertainty (CPU) moderates the effects of temperature and precipitation on grain yields.
Methods
This study employs panel data from 286 prefecture-level cities in China spanning 2001–2020. Based on the C-D production function, it adopts an economic-climate interaction model, incorporates CPU as a moderating variable, and conducts empirical tests using regression method.
Results
There exists a significant inverse U-shaped relationship between climatic factors and grain yields. However, when CPU increases, this relationship curve becomes significantly steeper, meaning the negative impact of extreme climate on grain yields is amplified. In other words, policy instability may exacerbate the destructive force of climate change. Further regional analysis reveals that this “amplification effect” of CPU is more prominent in non-major grain-producing areas: in the north, it primarily intensifies the precipitation-yield relationship, while in the south, it amplifies the temperature-yield relationship. It is noteworthy that in non-climate adaptation pilot cities, CPU exhibits a more pronounced negative moderating effect.
Discussion
The core implication of this research is that in the face of climate change, the clarity, coherence, and predictability of policies are themselves a crucial form of adaptive capacity. Ensuring policy stability can effectively stabilize farmers’ production expectations and incentivize long-term investments to combat climate risks. Further, the results also underscore the necessity of driving systemic transformation to convert external risks into endogenous drivers, then CPUs’ responses can be internalized as assets against future extreme weather within a risk framework in the future. Ultimately, anchoring climate-policy regimes in stable, Nature-Based Solutions—especially by advancing them across the Global South—offers a scalable pathway to turn policy uncertainty into ecological and social resilience.
climate changeinverted-U shapeclimate policy uncertaintygrain yieldsmoderating effectThe author(s) declared that financial support was received for this work and/or its publication. This work was supported by the Fujian Province Innovation Strategy Research Plan Project (Grant No. 2024R0015).section-at-acceptanceConservation and Restoration EcologyIntroduction
Grain is the basis for human survival and development, which affects political, economic and social stability and has received widespread global attention (Bi et al., 2022; Çakir et al., 2025; Clapp et al., 2022; Pandey and Mishra, 2024). Grain production is the core support for food security (Li and Lin, 2022; Yan et al., 2023). However, in the 21st century, extreme weather events such as high temperatures, floods, droughts and cold waves have occurred frequent (Almazroui et al., 2021; Yan et al., 2022), severely impacting the global grain production environment (Lee and Zhao, 2023; Zheng et al., 2018). Both the Food and Agriculture Organization of the United Nations (FAO) and the United Nations Intergovernmental Panel on Climate Change (IPCC) have clearly pointed out that agriculture is one of the industries most significantly and vulnerably affected by the risks under climate change (Intergovernmental Panel On Climate Change (IPCC), 2022). Among these, developing countries, which are highly dependent on grain production for their livelihoods and face a lack of technological innovation and climate adaptation strategies, are far more affected than developed countries (Abbas et al., 2023; Warsame et al., 2021).
China is one of the countries with the most significant global climate change and frequentmeteorological disasters (Wang W. et al., 2024). Theseabnormal climate changes have severely impacted the stability and sustainability of China’s grain production, and their most direct and obvious manifestation on grain production is the fluctuation of grain yields (Li D. et al., 2023; Xiao et al., 2013). It is predicted that climate change could lead to a reduction in China’s wheat, rice and corn production by about 20%, which in turn has potential impacts on international grain trade (Xuan et al., 2021). As the world’s largest agricultural producer (Carter et al., 2012), it is of great significance to explore the impact of climate change on Chinese agriculture. In recent years, scholars have conducted a large number of studies in this area, and have found that climate change, especially rising temperatures and changing precipitation patterns, has a significant impact on China’s grain yields (Gao R. et al., 2025; Holst et al., 2013; Li et al., 2020; Saud et al., 2022).
Referring to the existing literature, agricultural climate balance and even stability of global grain yields were continuously threatening by aforementioned climate change risks (Luqman et al., 2023; Wheeler and Von Braun, 2013). Without an effective response in the coming decades, it will systematically jeopardize grain yields, destabilize food security systems and undermine the foundations of agricultural livelihoods (Hao W. et al., 2024; Ray et al., 2015). This highlight that coherent and stable climate policies are essential to provide grain producers with technical support, financial security and risk insurance to enhance their capacity to cope with climate risks (Rodríguez-Barillas et al., 2024). However, the formulation, deliberation and implementation of climate policies generally face significant uncertainties (Liu G. et al., 2023). The root causes of uncertainty are diverse, covering multiple dimensions such as public attitudes, policy environments, interest group games, and urgency of the problem (Kyselá et al., 2019; Sharp et al., 2011; Wood et al., 2014). Against this background, Gavriilidis (2021) proposes a new assessment methodology, called the Climate Policy Uncertainty (CPU) Index, to quantify the uncertainty that is closely associated with climate policy. Although the CPU index is a relatively new concept, it has attracted increasing attention from the academic community and has been applied in many studies (Ding et al., 2025; Kong et al., 2025; Ren et al., 2023).
However, there is still limited exploration related to the extension of CPU research to the closely related agricultural sector. In fact, the high dependence of the agricultural sector on climatic conditions and the institutional environment makes it more vulnerable to CPU (Lee et al., 2024). The climate policy environment not only profoundly affects the equilibrium of supply and demand and price fluctuations in agricultural markets (Sun et al., 2021; Xiao et al., 2019), but also directly shapes the risk-averse decision-making process of farmers and agribusinesses (Antón et al., 2013; Waldman et al., 2020). In a context of relatively stable climate policies, grain producers are able to plan their planting, select varieties and organize their farming schedules more effectively, based on climate patterns and policy guidance (Grigorieva et al., 2023; Osbahr et al., 2008). A stable policy environment also encourages producers to manage climate risks by introducing new technologies and diversifying cropping patterns (De Laporte et al., 2024; Vermeulen et al., 2012), and to have the confidence to make long-term investments, including improving soil and building water facilities, in order to increase the stability and resilience of grain production (Wei et al., 2025; Yang et al., 2007). However, the high level of CPU may be like a fog, seriously disrupting the expectations of future climate conditions and supporting policies. As it is difficult to predict the direction of policies (such as carbon emission limits, adjustment of agricultural subsidies), grain producers tend to be conservative and cautious in their decision-making (Ouattara et al., 2018). In the face of climate change (fluctuations in temperature and precipitation, etc.), they may avoid expanding grain production readily out of concern for potential policy adjustment risks (Yang et al., 2023). Simultaneously, long-term agricultural investment relies on a stable policy environment, and the CPU may create ambiguity in grain producers’ expectations of future returns, which inhibits their incentives to invest in production (Huang and Sun, 2024). Once climate variables deviate from the appropriate range for grain growth, grain yields are at risk of a sharp decline. Therefore, it is important to incorporate consideration of the role of CPU when examining the impact of climate change on grain yields. Especially in developing countries, policy uncertainty often accompanies rapid industrialization and frequent policy adjustments. The behavioral responses triggered by such uncertainty may further weaken the already limited climate resilience of agricultural systems, creating a cycle of “unclear policies → insufficient investment → weak adaptation”.
This study makes several contributions to the literature. First, this study employs a modified Cobb-Douglas (Cobb and Douglas, 1928) production function that incorporates climatic elements to construct an economic-climate model (C-D-C) and incorporate a quadratic term at the empirical level. From an economic perspective, it verifies that the impact of climate change impacts on grain production exhibits complex nonlinear characteristics involving multiple interacting factors. This establishes an empirical foundation for quantifying the economic costs of climate risks and optimizing food security response strategies. Second, while some studies suggest that policy conditions could moderate the relationship between climate change and grain yields (Gao et al., 2021; Hua et al., 2022; Zhu et al., 2024), the underlying mechanisms require further elaboration. This study employs empirical economic methods, incorporating CPU as a moderating variable within the “climate-agriculture” analytical framework, to reveal the role of policy uncertainty in mediating the relationship between climate change and grain production. The implications may extend beyond China to offer insights for various international contexts facing similar policy challenges, with a view to enhancing global climate change adaptation capacity and risk management capabilities. Finally, by analyzing data across multiple regional dimensions—including major grain-producing and non-major grain producing areas, northern regions and southern regions, and climate-adapted and non-climate-adapted cities—this study reveals the spatial heterogeneity of climate impacts and policy moderation effects. It provides decision-making references for countries worldwide in addressing food shortages and malnutrition, holding particular significance for nations currently facing food crises.
The subsequent structure of this study is as follows. Section 2 presents the literature review and theoretical analysis. Section 3 focuses on the model, describing variables and data. Section 4 introduces our empirical findings. Section 5 serves as the discussion section, providing an in-depth analysis of the empirical results. Section 6 concludes the study and offers corresponding policy recommendations.
Literature review and research hypothesesClimate change and grain yields
Grain production is highly dependent on the climatic environment (Chandio et al., 2023; Teng et al., 2023), and climatic factors are not only critical to the sustainability of agricultural production activities (Cabas et al., 2010; Chandio et al., 2024), but also key drivers of grain growth and yield in a given region (Baig et al., 2023; Herath and Thirumarpan, 2017; Panda et al., 2019). In this context, climate change has become a major force reshaping agroecosystems (Altieri et al., 2015; Semeraro et al., 2023). There is a global consensus that climate change has far-reaching implications for agricultural production (Schlenker and Roberts, 2009; Yuan et al., 2024), temperature and precipitation are the core influencing factors (Kukal and Irmak, 2018).
Evidences of significant impacts on grain yields have been found in many regions of the world (Lobell and Field, 2007; Wang et al., 2018). Some studies have shown that rising temperatures may have a positive impact on agricultural production in some developed countries such as the United State and Germany (Deschênes and Greenstone, 2007; Li et al., 2011; Lippert et al., 2009). Abbas (2022) investigated the dynamics between grain yields and precipitation in Pakistan from 2000 to 2019 and found that increased precipitation has a significant positive impact on grain yields. Conversely, a study by Tetteh et al. (2022) revealed that high temperatures have jeopardized overall grain production in Ghana. Similarly, Wu et al. (2021) found through their study that increased temperatures have harmed maize production in China from 1979 to 2016. Li (2023) suggested that increased precipitation has negatively impacted rice production in Japan. Overall, temperature and precipitation are the core influencing factors (Kukal and Irmak, 2018), and the relationship between them and grain yields may not be a simple linear pattern, but a complex and non-linear one (Burke et al., 2015). Regarding nonlinear effects, Li C. et al. (2024) used panel data from China’s major grain-producing regions between 1978 and 2018, finding a nonlinear relationship between precipitation, temperature, and rice yields, which is characterized by an inverted U-shaped pattern. Agnolucci et al. (2020) research indicates that grain yields do not follow a simple linear relationship with temperature but instead displays a nonlinear pattern centered around an “optimal temperature window.” Based on the literature, studies examining the nonlinear relationship between temperature and precipitation and grain yields are still relatively limited compared to those focusing on linear relationships.
According to the threshold theory in agronomy (AT), crop responses to environmental or biologicalfactors are often nonlinear. When factors exceed a critical value (threshold), yields decline rapidly or disaster risks rise significantly (Huang et al., 2022; Luo, 2011). Simultaneously, the law of diminishing marginal returns (DMR) in agricultural production functions indicates that, when other inputs (such as land and technology) remain constant, continuously increasing a single variable factor (like labor, fertilizer, or water) results in marginal productivity (the output increment per additional unit of that factor) first rising and then declining, ultimately following a diminishing trend (Just and Peterson, 2003; Layson, 2015). Currently, some studies have extended this theoretical logic to the analysis of how climate factors influence agricultural production reveals: when the climate variables are in the appropriate range, their resource attributes are dominant, and they contribute to grain yields by optimizing photosynthesis, ensuring cumulative temperatures and land moisture (Hao S. et al., 2024); once the ecological threshold is exceeded, they turn into stress factors, which inhibit physiology disturb the ecological balance, and induce disasters, leading to a decrease in grain yields (Mirón et al., 2023; Skarbø and VanderMolen, 2016). Overall, the relationship between temperature and precipitation and grain yields may not be a simple linear pattern, but could exist a potential complex nonlinear relationship (Burke et al., 2015).
Specifically, from a temperature perspective, when temperatures fall below the appropriate range,grain growth slows down or even stops. Low temperatures hinder cellular metabolism by inhibiting enzyme activity (Wang H. et al., 2023), and extreme low temperatures can cause frost damage that destroys cellular structure and interferes with nutrient transportation, leading to a reduction in grain yields (Lin et al., 2023). As temperatures gradually rise, grain yields increase. Increased temperatures accelerate physiological processes such as photosynthesis, respiration and nutrient conversion (Niu et al., 2023), creating conditions for seed development and activating soil microbial activity to release more nutrients (Wang M. et al., 2024). However, when temperatures exceed certain thresholds, grain yields decline. High temperatures and heat waves can exacerbate the volatilization of pesticides and fertilizers, increasing the risk of pests and diseases (Yin et al., 2023), while at the same time accelerating respiratory depletion but inhibiting photosynthesis at the physiological level, leading to a lack of net photosynthetically active product accumulation resulting in a reduction in grain yields (Zhang et al., 2022).
From a precipitation perspective, when the precipitation is insufficient, the lack of soil moisture will directly cause the physiological activities of the grain to be hindered (Mares and Mrva, 2014), which is manifested by the wilting of leaves, the decrease of photosynthetic efficiency and the restriction of root growth, which will ultimately lead to the reduction of grain yields (Garg et al., 2020). With the increase of precipitation, the improvement of soil moisture can guarantee the absorption of water and fertilizer to meet the needs of grain growth period, thus promoting the yields of grain (Qiu et al., 2022). However, when precipitation exceeds the capacity of grain crops, soil water saturation and deterioration of aeration lead to root hypoxia, which inhibits respiration and normal physiological functions, threatening grain growth and yields (Khalil et al., 2020). In addition, extreme flooding can even destroy farmland, which can be devastating to grain production (Shehzad, 2023).
Based on the above literature review and theoretical analysis, this study proposes.
Hypothesis 1a: The relationship between temperature and grain yields follows an inverted U-shaped curve.
Hypothesis 1b: There exists an inverted U-shaped relationship between precipitation and the grain yields.
Moderating effect of climate policy uncertainty
CPU refers to the uncertainty and lack of clarity inherent in government policies and regulationsrelated to climate change mitigation and adaptation strategies, which makes it difficult for policy participants (such as businesses, agricultural producers, investors, etc.) to accurately predict the direction of the policy and the impacts it will bring (Tian and Li, 2023). Currently, the measurement of CPU relies on mining news texts, which are processed using word frequency statistics and machine learning algorithms (Ma et al., 2023; Noailly et al., 2022). The existing discussions on CPU mainly focus on three levels. First of all, at the level of economic risk, CPU will increase the operating costs and risks of enterprises, make their investment decisions more prudent, and then inhibit the expansion of investment scale and cut international trade flows (Niu et al., 2023; Zhang Z. et al., 2025). Secondly, in terms of environmental risks, CPUs could undermine investment in green technologies, lead to a mismatch of resources and an inability to effectively channel financial flows to key environmental sectors, and slow down the green transition process (Gao D. et al., 2025; Wei et al., 2024). Finally, in terms of social risk, CPU may exacerbate social inequality and undermine public trust in climate policy (Emami Meybodi and Owjimehr, 2024; Kitt et al., 2021).
In the agricultural sector, frequent weather extremes, exacerbated by global warming, have severely disrupted grain growth cycles (Fatima et al., 2020; Onyeaka et al., 2024). Climate policies are designed to provide farmers with guidelines and safeguards against climate risks (Khan et al., 2023), and to buffer climate shocks and enhance system resilience through adaptive measures (Assan et al., 2018; Khan and Roberts, 2013). However, due to the complexity of the climate and the need to balance economic growth and carbon reduction, policy formulation and implementation are often characterized by discontinuity or ambiguity (Lee et al., 2024), making it difficult for agricultural actors to anticipate changes in the policy environment (Wang J. et al., 2024). For example, frequent adjustments in agricultural subsidies, irrigation investments or emission reduction targets further confuse expectations (Masud and Khan, 2024). While existing studies have separately highlighted the economic costs of CPU, its drag on the green transition, and the threats climate change poses to agriculture, empirical examinations of how CPU interacts with natural climate risks and analyses of regional vulnerability differences remain relatively limited.
The theory of real options (RO) represents the extension and application of financial option concepts and analytical tools within the real economy. This theory posits that investment opportunities held by enterprises can be viewed as options based on physical assets (Kogut and Kulatilaka, 2001; Trigeorgis and Reuer, 2017). The higher the level of uncertainty in the external environment, the greater the value of such options typically becomes. Consequently, from the perspective of real options, the CPU increases the “waiting value” of producers (Guo et al., 2023). This “waiting value” makes grain producers, especially risk averse farmers, more inclined to delay the implementation of countermeasures in response to climate risks, maintain the status quo, and avoid attempting to introduce new technologies or new varieties that may bring uncertainty, so as to accumulate more information and avoid potential risk (Lei et al., 2024). In addition, according to the Adaptive Capacity (AC) theory, which refers to the ability of socio-ecological systems, institutions, humans, and other organisms to adjust in order to reduce potential harm, seize opportunities, or cope with consequences (Chapagain et al., 2025; Seaborn et al., 2021). Therefore, under high CPU load, agricultural entities find themselves caught in a dilemma of dual uncertainties stemming from both climate and policy (Mittenzwei et al., 2017),and their adaptive behavior is degraded from “forward-looking investment” to “passive response” (Menapace et al., 2013), which significantly weakens the adaptive capacity of the system.
The potential risks posed by CPU is that it may systematically reconfigures the slope of themarginal effect of the inverted U-shaped relationship between climate variables and grain yields.The moderating effect could be mainly concluded in three dimensions, which could increase the“waiting value” and decrease the adaptive capacity. The first is the weakening of risk response and agricultural inputs, CPU makes the management behavior of grain producers tend to be conservative (Shah and Alharthi, 2022; Singh et al., 2020; Yuan et al., 2021), such as farmers worrying about the loss of policy changes and sticking to the traditional planting pattern (Brock and Haden, 2024), which hinders the improvement of grain climate resilience. At the same time, the willingness to invest in the long term will also contract (Hong et al., 2024; Liu et al., 2024), for example, delaying the upgrading of irrigation systems and the introduction of resilient varieties due to uncertainty about the direction of future climate policy support (Hallegatte, 2009; Wang S. et al., 2019). This could result in a lag in agricultural transformation and a weakening of grain system stability (Li et al., 2017). Second is the rigidity of production and business decisions, with CPUs interfering with market signals, exacerbating grain price volatility (Liu G. et al., 2023; Wang K. H. et al., 2023; Yin and Cao, 2024), and constraining grain producers from forming rational expectations (Zou et al., 2024). This leads to a tendency for farmers’ production decisions to be short-term risk averse, such as discarding high-risk, high-yield crops and conservatively adjusting acreage (Findlater et al., 2019; Ma et al., 2023; Tong et al., 2019). This significantly limits the flexibility of the grain system to respond to climate fluctuations and magnifies the negative impacts of temperature and precipitation changes on grain yields (Nam, 2021). The third is technology iteration and infrastructure blockage, as CPU increases the risk of sunk costs in agricultural technology development and inhibits the willingness of innovation agents to invest, especially when the technology may face policy rollbacks (Snitker et al., 2024; Sun et al., 2024), which will result in a delay in the diffusion of adaptation technologies. At the same time, because CPU increases the difficulty of risk assessment, financial institutions will tighten credit (Han et al., 2023; Hansen, 2022) and limit financial support for agricultural infrastructure and technological innovation (Liu Y. et al., 2023; Osabutey and Jackson, 2024), making grain production more vulnerable in the face of extreme weather.
Based on the above literature review and theoretical analysis, this study proposes.
Hypothesis 2a: CPU enhances the relationship between temperature and grain yields, that is, the higher the CPU, the steeper the inverted U-shaped relationship between temperature and grain yields.
Hypothesis 2b: The relationship between precipitation and grain yields is reinforced by CPU, with greater CPU levels translating into a steeper inverted U-shaped pattern.
Based on a progressive, sub-thematic literature review and analysis of relevant theories, we constructed the core theoretical framework and corresponding hypothesis mechanism diagram for our study. This provides a clarified presentation of our research approach and theoretical application process (Figure 1). As shown in Figure 1, the potential contribution of our research lies in two aspects. First, building upon the inverted U-shaped relationship between temperature and precipitation observed in existing literature, we employ an empirical model from an economic perspective to validate this agronomic principle, aligning with AT and DMR theories. Second, we further explore the possible moderating role of CPU in this inverted U effect—examining its potential amplifying or diminishing influence and the potential nonlinearity of its role—while attempting to construct a theoretical framework centered on AC and RO theories to analyze such potential phenomena.
Theoretical mechanism diagram.
Diagram illustrating the relationship between temperature, precipitation, and grain production. It features an inverted U-shape curve, indicating positive effects predominate initially, then negative. Key theories include AT and DMR, and AC and RO, emphasizing suitable and related higher temperatures and precipitation effects. The hypothesis section discusses the inverted U-shape's impact on grain production and the modulation by Climate Policy Uncertainty. Elements include photosynthesis, nutrient absorption, soil moisture, and risk response. The C-D-C production function denotes variables K, L, and T.Research designModel constructionC-D model
The Cobb-Douglas production function model (C-D model) (Cobb and Douglas, 1928) is widely recognized in economics as the classic model reflecting the relationship between production factors and output (Vasyl’Yeva, 2021). Compared to other models, the C-D production function model can more accurately depict the process of agricultural input-output and enable in-depth analysis in conjunction with other economic indicators (Chou et al., 2021). Its basic form is:
Q=f(K,L)=AKαLβ
In Equation 1, Q represents output; L and K denote the quantities of labor and capital inputs, respectively; A indicates the level of technology; α and β are the output elasticity coefficients for labor and capital, respectively.
“Economic-Climate” model
Since grain yields is constrained not only by factors of production such as land, labor, and fertilizer, but also directly influenced by climatic elements like temperature and precipitation (Wang P. et al., 2019). To enhance the applicability and explanatory power of the C-D production function model, agricultural economists have undertaken innovative efforts. They incorporate climate elements like temperature and precipitation as crucial production factors into the traditional C-D production function, forming an “economic-climate” model (C-D-C) that comprehensively considers both climate factors and agricultural input resource (Chou and Ye, 2006; Dong et al., 2007). The C-D-C model integrates both climatic and economic factors, treating climatic parameters under average climate conditions while accounting for long-term trends in climatic variables (Guo et al., 2024). Its specific form is:
Y=f(K,L,T,C)
In Equation 2, Y represents grain yields; L and K denote labor and capital inputs, respectively; T signifies technological progress; and C represents climate variables.
Quadratic form model
To further verify whether a nonlinear relationship exists between climate change and grain yields, this study extends the C-D-C model by introducing a quadratic term for climate factors. In agronomy, threshold theory (AT) emphasizes that crop growth exhibits a three-stage response to key environmental factors: “minimum threshold - optimal range - maximum threshold.” Regarding temperature: The minimum threshold is the base temperature required for seed germination and enzyme activation; The optimal range is the temperature interval where processes like photosynthesis achieve maximum efficiency; The maximum threshold is the temperature causing protein denaturation and a sharp increase in respiratory consumption. Regarding precipitation: The minimum threshold is the basic water requirement for sustaining cellular photosynthesis; The optimal range is the equilibrium point between water deficiency and waterlogging; while the maximum threshold represents the critical point for root rot due to waterlogging (Challinor et al., 2015). This indicates that the impact may rise initially (minimum threshold to optimal range), peak at the optimal range, and then decline (optimal range to maximum threshold), forming an inverted U-shaped influence curve. Therefore, integrating agronomic theory with economic modeling, we incorporate a quadratic term into the model to characterize the impact of climate on crop production.
Thus, the C-D-C model is transformed logarithmically, and its log-linear form is selected as the analytical framework for this study:
where y denotes total grain yields; climate denotes climate change factor; x denotes the control variable (This study selects control variables based on the three core factors of capital, labor, and technological progress within the C-D production function, while accounting for agricultural production characteristics: ①Capital Input Dimension: Crop planting area, fertilizer application volume, and government fiscal expenditure are selected, representing the land capital base, physical capital investment, and public capital investment sources for agricultural production, respectively. ②Labor Input Dimension: The number of persons employed in the primary sector is selected to reflect the scale of labor input in agricultural production. ③Technological progress dimension: The number of agricultural patents is selected as a quantitative indicator of agricultural technological invention and innovation output.); constant term is denoted by α, ϵit denotes random error term; γi denotes individual fixed effect; μt denotes time fixed effect, i denotes city; t denotes year. In agricultural production, the yield-increasing effect of marginal inputs (such as fertilizers, water, and labor) diminishes progressively as input quantities accumulate, rather than maintaining a constant rate of increase. The logarithmic form accurately captures this nonlinear relationship where “incremental inputs gradually decouple from incremental outputs,” aligning with the actual response characteristics of factor inputs in agricultural production (Li and Geng, 2013; Lobell et al., 2013).
CPU moderation model
Theoretical analyses have shown that CPU has a moderating effect of any input factor (including climatic factors) is typically not constant. Introducing a on the relationship between climate change factors and grain yields. Therefore, On the basis of model (3), the moderating effect model (4) of curvilinear regression is constructed by adding CPU and lnclimateit*CPUit, the interaction term between CPU and climate change factors, as well as (lnclimateit)2*CPUit, the interaction term between CPU and the squared term (ln(climateit)2) after taking the logarithm effectively captures this “initially increasing, then decreasing” nonlinear relationship, which aligns mathematically with threshold theory in agronomy.Furthermore, the logarithmic form effectively mitigates the heteroskedasticity issue of the model (Wooldridge, 2020). The climate change factors remain consistent with the previous specification, while the rest of the control variables are the same as those in model (3):
If the sign of the coefficient of the quadratic term is the same as that of the coefficient of the quadratic interaction term, there is an enhancement effect; if the sign is opposite, there is a weakening effect (Haans et al., 2016). If hypothesis H1 holds, the coefficient of the squared term in model (1) is significantly negative, then a significantly negative β5 may indicate that CPU strengthens the relationship between climate change factors and grain yields, while a significantly positive β5 indicates that CPU weakens the relationship between the two.
Variable selection
(1) Core explanatory variable: grain yields (Y). Grain yields is measured as the total amount of grain products, mainly including the annual production of rice, wheat, corn, soybean, tuber and other crops (at prefecture-level).
(2) Explained variable: Climate change. The impact of climate on agricultural production is primarily reflected in temperature and precipitation (Kukal and Irmak, 2018; Maitah et al., 2021). Considering that the annual mean temperature comprehensively reflects the heat accumulation status during the grain growing season and before sowing, it determines whether grain can complete their growth cycle normally and influences the selection of sowing dates (Arshad et al., 2021). Annual total precipitation reflects the overall level of water supply throughout the year, meeting the water requirements of the grain growing season while also affecting soil moisture conditions, thereby influencing the determination of sowing periods and seedling emergence rates (Drebenstedt et al., 2023). Both metrics also serve as macro indicators representing fundamental regional climate characteristics, thereby avoiding the cumulative climatic effects and prior influences that might be overlooked when selecting data from a single growing season period (Vogel et al., 2019; Yang et al., 2020). Therefore, this study employs annual mean temperature (TP) and annual total precipitation (RF) as climate indicators to assess their impact on grain production (at prefecture-level).
(3) Moderating variable: CPU. This study employs the China Prefecture-Level CPU Index calculated using the MacBERT model. This index, based on data from Ma et al. (2023) covering six major Chinese news outlets—including the People’s Daily, Guangming Daily, Economic Daily, Global Times, Science and Technology Daily, and China News Service. It employs the China Prefecture-Level CPU Index calculated using the MacBERT model, which encompasses uncertainties across all dimensions of climate policy: the entities responsible for policy formulation (who formulates climate policies), the timing and content of policy implementation (when and what types of policies are implemented), and the consequences of climate policy actions (the outcomes of climate policy actions). (See Appendix 1 for the data processing procedure).
(4) Control variables: Based on the three factors of capital (K), labor (L), and technological progress (T) in the C-D production function, and considering both literature conventions and data availability, this study sets the corresponding control variables as follows: Considering the characteristics of agricultural production, the area sown with crops determines the land size for grain production and is the agricultural capital base dimension; fertilizer application represents mobile physical capital and affects soil fertility and yields; and government fiscal expenditure is the main source of public capital inputs such as agricultural infrastructure and agricultural subsidies. Therefore, capital input is represented by area sown with crops (AC), fertilizer application (FER), and government fiscal expenditure (GE). The number of employees in primary industry directly reflects the scale of labor input into agricultural production. Therefore, labor input is measured by the number of employees in primary industry (LAB). The number of agricultural patents serves as a direct, quantifiable indicator of technological inventions and innovation output within the agricultural sector. Therefore, technological progress is reflected by the number of agricultural patents (TE). The specific definitions of the above variables are shown in Table 1.
Definition of variables.
Variable type
Variable definition
Variable symbol
Variable unit
Explained variable
Grain yields
Y
Ten thousand tons
Explanatory variable
Average annual temperature
TP
degrees centigrade
Total annual precipitation
RF
millimetre
Adjusting variable
Climate policy uncertainty
CPU
/
Control variable
Area sown with crops
AC
thousands of hectares
Fertilizer application
FER
Ten thousand tons
Number of employees in primary industry
LAB
ten thousand people
Number of agricultural patents
TE
item
Government fiscal expenditure
GE
billions
Data sources and data processing
In this study, we select the panel data of 286 prefecture-level cities in China from 2001 to 2020, and the data of the above explanatory variables and all control variables are from the National Bureau of Statistics of China, China’s official statistical yearbooks, and the work reports of local governments issued by the governments of each prefecture-level city at the end of the year. The missing data of a small number of prefecture-level cities are added and processed by linear interpolation in this study. (Linear interpolation can fill in missing values based on the linear relationship between adjacent known data points while maximally preserving the trend characteristics of the original data. It causes minimal disturbance to the original data and is suitable for statistical materials with relatively stable data quality over extended time spans (Wooldridge, 2020), such as panel data from prefecture-level cities.) The data source for the explanatory variables climate change factors (mean annual temperature and total annual precipitation) is the National Center for Environmental Information (NCEI) under the U.S. National Oceanic and Atmospheric Administration (NOAA) (See Appendix 1 for the data processing procedure). The CPU index is available for download from the ISETS Energy Finance Network website. In this study, all continuous variables are shrink-tailed at the 1% and 99% levels to minimize the effect of outliers. A statistical description of the sample intervals is presented in Table 2.
Descriptive statistics of variables.
Variable
Obs.
Mean
S.D.
Min.
Max
lnY
5720
4.823
1.046
1.065
6.957
lnTP
5720
2.594
0.459
0.841
3.175
lnRF
5720
6.875
0.552
5.262
7.842
lnAC
5720
5.878
0.905
2.832
7.554
lnFER
5720
2.486
0.944
-0.598
4.419
lnLAB
5720
-1.102
1.476
-4.605
2.707
lnTE
5440
4.245
1.834
0.693
8.32
lnGE
5720
4.818
1.2
2.153
7.58
CPU
5720
1.086
0.638
0.053
2.742
Empirical test and results analysisBaseline regression results
Based on Equation 3 in the model construction, we conducted regression analysis to investigate the relationship between temperature, precipitation, and grain yields. The relevant results are summarized in Table 3. Model (1) indicates that, without controlling for other variables, temperature exerts a positive effect on grain yields at the 5% significance level, while its squared term exerts a negative effect at the 5% significance level. Based on this, Model (3) incorporates economic input as a control variable. The results remain significant: the coefficient for lnTP is 0.544, the quadratic coefficient for lnTP is -0.150, revealing a nonlinear relationship between the two variables. By differentiating Equation 3 and incorporating the coefficients from the baseline regression results presented in Table 3, calculations reveal that within the sample range, when temperature reaches its minimum value, the slope of the curve is 0.292, indicating that a 1% increase in temperature leads to a 0.292% increase in grain yields. When temperature reaches its maximum value, the slope of the curve is -0.409, indicating that a 1% increase in temperature leads to a 0.409% decrease in grain yields. The inflection point of the curve is 1.8131, indicating that there is an inverted U-shape relationship between temperature and grain yields, which is “promoted first, then suppressed”. This indicates that when the temperature is lower than a turning point, increasing temperature can increase grain yields. However, when the temperature rises to the turning point, the continuous increase in temperature will lead to a decrease in grain yields, which aligns with the findings of Hogan and Schlenker (2024), validating the universality of the inverted U-shaped relationship between temperature and grain yields. The reason for this is that an appropriate increase in temperature accelerates physiological processes such as photosynthesis and creates favorable conditions for grain growth, thus increasing grain yields. However, when the temperature is too high, it will not only increase the volatilization of pesticides and fertilizers in the farmland, leading to an increase in pests and diseases, but will also lead to a decrease in the activity of photosynthesis enzymes, a decrease in the accumulation of organic matter in grain crops, and ultimately a decrease in yields.
Baseline regression results of the impact of climate change on grain yields.
Variable
Exclude control variables
Include control variables
(1)
(2)
(3)
(4)
Baseline-TP
Baseline-RF
Baseline-TP
Baseline-RF
lnTP
0.908**
0.544*
(2.303)
(1.670)
(lnTP)2
-0.243**
-0.150*
(-2.511)
(-1.848)
lnRF
2.455***
2.069***
(9.502)
(8.768)
(lnRF)2
-0.170***
-0.143***
(-9.087)
(-8.364)
Constant
4.153***
-3.972***
3.151***
-3.927***
(10.533)
(-4.444)
(8.957)
(-4.776)
Control
No
No
Yes
Yes
Entity FE
Yes
Yes
Yes
Yes
Time FE
Yes
Yes
Yes
Yes
N
5720
5720
5720
5720
R2
0.003
0.048
0.152
0.185
The values in parentheses are t statistics. ***, **, and * indicate significance at the confidence levels of<0.01,<0.05, and< 0.1. “Control Yes” indicates the inclusion of control variables in the regression.
Model (2) (see Table 3) indicates that, without controlling for other variables, precipitation exerts a positive effect on grain yields at the 1% significance level, while its squared term exerts a negative effect on grain yields at the 1% significance level. Based on this, Model (4) incorporates economic input as a control variable. The results remain significant: the coefficient for lnRF is 2.096, the quadratic coefficient for lnRF is -0.143, indicating a nonlinear relationship between the two variables. Similarly, by differentiating Equation 3 and incorporating the coefficients from the baseline regression results presented in Table 3, calculations reveal that within the sample range, when precipitation reaches its minimum value, the slope of the curve is 0.564. This indicates that at this point, a 1% increase in precipitation leads to a 0.564% increase in grain yields. When precipitation reaches its maximum value, the slope of the curve is -0.174, indicating that a 1% increase in precipitation leads to a 0.174% decrease in grain yields. The inflection point of the curve is 7.391, indicating that there is an inverted U-shape relationship between precipitation and grain yields (when the precipitation is low, increasing precipitation can improve the yields; however, when precipitation increases to a turning point, continuing to increase precipitation will lead to a reduction in yields). This conclusion aligns with the findings of Saei et al. (2018), validating the universality of the inverted U-shaped relationship between precipitation and grain yields. The reason for this is that drought and low rainfall can inhibit the growth and development of grain. Moderate increase in precipitation is good for increasing grain set and hence grain yields. However, when precipitation is too high, it can lead to over-wetting of farmland, collapse of grain crops, and even bring about flooding, which washes away farmland and results in reduced grain yields.
As a result, this study reveals an inverted U-shaped relationship between temperature, precipitation, and grain yields. This finding aligns with Hoffman et al. (2018) research on the nonlinear association between climatic factors and grain yields, while also corroborating the threshold theory within the field of agronomy (Luo, 2011; Xiao et al., 2022). It should be noted that the inverted U-shaped relationship belongs to a special curvilinear relationship, in order to further ensure the reliability of the results of the above inverted U-shaped relationship, this study draws on the method of Lind and Mehlum (2010), and utilizes the “utest” command to test the regression (3) and (4). The test results show that the overall t-value is significant at the statistical level of 1%, and the slope of the set model has both positive and negative values, and the location of the extreme point falls within the sample interval, which indicates that the inverted U-shaped relationship is valid. Hypothesis H1 is verified.
Moderating effects
The above analysis shows that there is an inverted U-shaped relationship between climate change factors and grain yields. Based on the above, we expect that CPU enhancement may amplify the negative effect of climate change factors, i.e., strengthen the inverted U-shaped relationship between the two. To empirically verify the moderating effect of CPU, we introduced the interaction term between CPU and the squared term of climate change factors to construct the moderating model (Equation 4) The regression results are summarized in Table 4. The Model (3) tests the positive moderating effect of CPU on the relationship between temperature and grain yields. The results show that the coefficient of the quadratic cross-multiplier term between temperature and grain yields is significantly negative at the 5% level, indicating that CPU further strengthens the inverted U-shaped relationship between temperature and grain yields, that is the inverted U-shaped relationship between temperature and grain yields is steeper when CPU is higher. Similarly, model (4) (see Table 4) tests the positive moderating effect of CPU on the relationship between precipitation and grain yields. The results show that the coefficient of the quadratic cross-multiplier term between precipitation and grain yields is significantly negative at the 5% level, indicating that CPU likewise further strengthens the inverted U-shaped relationship between precipitation and grain yields, that is the inverted U-shape relationship between precipitation and grain yields is steeper when CPU is higher.
Regression results of moderating effect of CPU on the relationship between climate change and grain yields.
Variable
Exclude control variables
Include control variables
(1)
(2)
(3)
(4)
Moderation-TP/CPU
Moderation-RF/CPU
Moderation-TP/CPU
Moderation-RF/CPU
lnTP
1.102***
0.733**
(2.839)
(2.200)
lnTP*CPU
0.131
0.300***
(1.161)
(3.108)
(lnTP)2
-0.254***
-0.166**
(-2.690)
(-2.027)
(lnTP)2*CPU
-0.082***
-0.111***
(-3.401)
(-5.306)
CPU
-0.037***
-0.034***
-0.024***
-0.021***
(-5.339)
(-5.227)
(-3.730)
(-3.373)
lnRF
1.701***
1.419***
(6.689)
(6.161)
lnRF*CPU
0.240
0.251
(1.226)
(1.398)
lnRF2
-0.115***
-0.096***
(-6.229)
(-5.714)
(lnRF)2*CPU
-0.029**
-0.029**
(-2.020)
(-2.185)
Constant
3.765***
-1.341
2.788***
-1.653**
(9.545)
(-1.536)
(7.677)
(-2.069)
Control
No
No
Yes
Yes
Entity FE
Yes
Yes
Yes
Yes
Time FE
Yes
Yes
Yes
Yes
N
5720
5720
5720
5720
R2
0.099
0.109
0.223
0.233
The values in parentheses are t statistics. ***, **, and * indicate significance at the confidence levels of<0.01,<0.05, and< 0.1. “Control Yes” indicates the inclusion of control variables in the regression.
The findings suggest that the CPU indeed amplifies the adverse effects of climate change on grain yields, implying that the challenges originally posed by climate change may be further amplified against the backdrop of an uncertain and unstable policy environment, making agricultural production more difficult. This amplification effect may be more pronounced in agricultural production scenarios within developing countries, as their agricultural producers often have limited resources and lower resilience to policy fluctuations, making it difficult for them to cope with the combined impacts of climate risks and policy volatility. This aligns with the core argument proposed by Webber et al. (2020) that non-climatic stresses may exacerbate climate-induced yields losses. However, this argument extends the existing literature by treating policy uncertainty as an independent and understudied amplifying factor, rather than a generic contextual factor. Moreover, unlike previous studies such as Kan et al. (2023), which primarily examined the direct effects of policy adjustments on agricultural input allocation, our findings reveal a more complex and indirect pathway: policy volatility interacts with climate risks to exert compounding pressures on grain production systems.
The moderating effect of CPU is demonstrated more clearly in Figures 2, 3, where a steeper inverted U-shaped relationship between climate factors and grain yields can be seen for high CPU. Thus, hypothesis 2 is verified. According to the theoretical analysis above, for the main body of grain production, CPU amplifies their risk in climate change (it is difficult to accurately predict the direction of the climate and quickly respond to the related risks) and market risk (fluctuations in the price of agricultural products as well as changes in the market demand), which makes them more cautious and conservative in the process of production decision-making, which affects the efficiency of grain production, restricts the improvement of grain yields, and further amplifies the conflict between climate change and grain yields. This study confirms the core finding from existing research that the climate policy environment influences agricultural producers’ operational decisions (Fanchone et al., 2022; Singh et al., 2016). This also aligns with the conclusion that policy interventions can alter the relationship between climate change and grain production (Harkness et al., 2023). Notably, while previous studies have primarily focused on the regulatory effects of policies themselves (Gao D. et al., 2025), this research innovatively reveals the moderating role of the climate policy environment—it can influence the relationship between climate change and grain production by affecting agricultural producers’ operational decisions.
The inverted U-shaped curve diagram regarding lnTP and lnY.
A plot shows three parabolic curves representing grain yield as a function of temperature in logarithmic scales. The equations are \( y = a_1(\ln TP)^2 \) with no CPU, and \( y = (a_2 + b \text{CPU})(\ln TP)^2 \) with CPU. Parameters \( a_1, a_2, \) and \( b \) have specified ranges. The legend indicates that the curves correspond to no CPU, low CPU (\(\text{CPU} - \sigma\)), and high CPU (\(\text{CPU} + \sigma\)), shown in black and green.
The inverted U-shaped curve diagram regarding lnRF and lnY. (σ denotes the standard deviation).
Graph depicting the relationship between logarithm of grain yields and precipitation. The curve represents equations with and without climate potential units (CPU). No CPU is a black line, low CPU is a light green line, and high CPU is a bright green line. The equations and parameter ranges are displayed. The x-axis shows logarithm of precipitation, and the y-axis shows logarithm of grain yields.
Therefore, to mitigate the negative impacts of climate change on grain yields and ensure the sustainability of agricultural production, policymakers must not only establish stable, transparent, and consistent climate policy frameworks but also respond promptly to challenges posed by uncertainty. This approach prevents CPU from disrupting the adaptation rhythm of grain production systems to climate change, thereby avoiding the vicious cycle of “reduced grain yields → worsening poverty.” Further, this finding also underscores the necessity of driving systemic transformation to convert external risks into endogenous drivers (Vermeulen et al., 2018; Zagaria et al., 2021). By deeply integrating climate adaptation with green agricultural development and industrial upgrading—including promoting water-saving, energy-efficient, and pollution-reducing green production technologies (Loboguerrero et al., 2019; Wang et al., 2025)—and cultivating enhanced agricultural climate resilience as a new competitive advantage for regional agriculture (Kabato et al., 2025; Regmi and Paudel, 2024) (e.g., developing climate-smart agriculture that combines adaptation actions with resource efficiency improvements), CPUs’ responses can be internalized as assets against future extreme weather within a risk framework (Simpson et al., 2023).
Robustness tests
This study conducts robustness tests on the main regression results from three perspectives to ensure the reliability of the findings: (1) Lagged impact of policy; (2) Exclusion of municipalities; (3) Special year removal.
Lagged impact of policy
Based on the above analysis, the full impact of CPU on agricultural production may not be immediately apparent. By including a one-period lag term for CPU in the regression model, this study aims to capture the delayed response of farmers and agricultural systems. This allows for a more accurate assessment of how policy changes actually affect grain yields over time, rather than assuming instantaneous effects. The regression results are shown in Table 5 for models (1) and (2): the squared terms of temperature and precipitation and the interaction term of CPU both remain stable at a negative significant level of 1%, proving that CPU still reinforces the inverted U-shape relationship between climatic factors and grain yields, affirming the empirical results above and passing the robustness test.
Robustness tests results.
Variable
Lagged impact of policy
Exclusion of municipalities
Special year removal
(1)
(2)
(3)
(4)
(5)
(6)
Robust-TP
Robust-RF
Robust-TP
Robust-RF
Robust-TP
Robust-RF
lnTP
0.536*
0.671**
0.857**
(1.720)
(2.025)
(2.418)
(lnTP)2
-0.116
-0.156*
-0.205**
(-1.491)
(-1.915)
(-2.383)
lnTP*CPU
0.325***
0.330***
(4.005)
(3.915)
(lnTP)2*CPU
-0.112***
-0.115***
(-6.363)
(-6.309)
CPU
-0.009
-0.005
-0.012**
(-1.474)
(-0.766)
(-2.016)
lnTP*L.CPU
0.171
1.538***
(1.617)
(5.963)
(lnTP)2*L.CPU
-0.082***
-0.105***
(-3.680)
(-5.551)
L.CPU
-0.029***
-0.030***
-0.015***
0.461***
(-4.754)
(-5.004)
(-2.604)
(3.087)
lnRF
1.582***
1.411***
-0.044***
(7.390)
(6.122)
(-3.904)
(lnRF)2
-0.112***
-0.096***
(-7.281)
(-5.704)
lnRF*CPU
0.388***
(2.720)
(lnRF)2*CPU
-0.038***
(-3.530)
lnRF*L.CPU
0.244
(1.474)
(lnRF)2*L.CPU
-0.027**
(-2.234)
Constant
3.097***
-1.858**
2.803***
-1.647**
2.632***
-2.126**
(9.131)
(-2.450)
(7.701)
(-2.061)
(6.668)
(-2.392)
Control
Yes
Yes
Yes
Yes
Yes
Yes
Entity FE
Yes
Yes
Yes
Yes
Yes
Yes
Time FE
Yes
Yes
Yes
Yes
Yes
Yes
N
5434
5434
5640
5640
5148
5148
R2
0.218
0.215
0.234
0.237
0.243
0.246
The values in parentheses are t statistics. ***, **, and * indicate significance at the confidence levels of<0.01,<0.05, and< 0.1. “Control Yes” indicates the inclusion of control variables in the regression.
Exclusion of municipalities
Considering that municipalities directly under the central government are obviously different from other prefecture-level cities in terms of their agricultural economic methods and production and business activities, they have unique administrative structures, economic priorities, and resource allocations, which may distort the relationship between climatic factors and grain yields. In this study, the samples of four municipalities, namely Beijing, Tianjin, Shanghai and Chongqing, are excluded from the regression process and the regression is re-run using the remaining samples. The results of the basic regression are shown in Table 5 for models (3) and (4): the quadratic and linear coefficients of temperature and precipitation on grain yields are significant, exhibiting an inverted U-shaped relationship. Meanwhile, the significance and positivity of the interaction terms of squared terms of temperature and precipitation with CPU remain unchanged, consistent with previous analyses, suggesting that the main conclusions of this study are well robust.
Special year removal
The 2008 snowstorms in the south and the 2016 flooding events in the middle and lower reaches of the Yangtze River had severe and unusual impacts on grain production in China. To prevent the events in exceptional years from masking the potential relationship between regular climate change and grain yields. In this study, data from 2008 and 2016 are excluded and the regression is rerun using the remaining samples to mitigate the impact of extreme weather events. The regression results are displayed in Table 5 for models (5) and (6): The coefficient values and significance levels of the coefficients of the base regression results and the moderated effects regression results are basically the same as those of the regression results above, which indicates that the results of the study are robust and reliable.
Endogeneity test
In identifying endogenous variables, grain yields may also influence CPU, such as low yields potentially prompting governments to revise climate policies. Thus, a bidirectional causal endogenous relationship between CPU and grain yields is highly probable. Simultaneously, CPU may be associated with unobserved local economic or environmental shocks that also affect yields, implying potential omitted variable bias. To mitigate endogeneity issues arising from the above factors, this study employs three sequential approaches for endogeneity testing: two-stage least squares (2SLS) with instrumental variables (IV), system GMM, and the inclusion of additional control variables. For instrument selection, the study adheres to the principles of correlation and exogeneity, choosing two indicators as IVs for CPU: the lagged political cycle (one year) and the lagged CPU variable L.CPU. The political cycle is measured following Xu and Ma (2019), using changes in government officials as its proxy variable. Specifically, if a city mayor changes between January and June, it is recorded as a change in the current year; if between July and December, it is recorded as a change in the following year. When a mayor changes, cycle is coded as 1; otherwise, it is coded as 0.
The first four columns in Table 6 present regression results based on the two-stage least squares method. Notably, DWH yields a p-value of 0.00, indicating endogeneity issues with CPU in the model. Based on the results from columns (1) and (3) in the first stage, it is evident that both instrumental variables are significantly correlated with CPU. Moreover, the Cragg-Donald Wald F-statistics for identifying weak instrumental variables exceed the 10% critical value, confirming the absence of weak instruments and validating the instrumental variable correlation assumption. The P-values for the Sargan statistics in the over-identification test are all greater than 0.05, verifying the exogeneity of the instrumental variables. Furthermore, as shown in columns (2) and (4), the coefficients for the interaction terms between CPU and the linear and quadratic terms of precipitation and temperature exhibit no significant differences compared to those in Table 4. This further substantiates the negative moderating effect of CPU on the relationship between climate risk and grain yields.
Regression results of endogeneity tests.
Variable
2sls-TP
2sls-RF
GMM-TP
GMM-RF
Add control -TP
Add control -RF
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
CPU
lnY
CPU
lnY
lnY
lnY
lnY
lnY
CPU
-0.350***
-0.248***
-0.001***
0.004***
-0.020***
-0.017***
(-9.863)
(-6.551)
(-3.089)
(5.461)
(-3.338)
(-2.955)
lnTP
-0.063
-0.366***
-0.124***
0.348
(-0.710)
(-3.450)
(-25.396)
(1.157)
(lnTP)2
-0.001
-0.043*
0.008***
-0.045
(-0.040)
(-1.819)
(7.823)
(-0.599)
lnTP*CPU
-0.224
0.623***
0.229***
0.469***
(-1.110)
(3.747)
(30.204)
(5.286)
(lnTP)2*CPU
0.038
-0.174***
-0.043***
-0.143***
(0.830)
(-4.695)
(-26.873)
(-7.405)
lnRF
-0.644***
3.376***
-0.165***
1.258***
(-2.750)
(13.023)
(-7.693)
(5.927)
(lnRF)2
0.045**
-0.265***
0.008***
-0.086***
(2.580)
(-13.676)
(4.714)
(-5.576)
lnRF*CPU
-0.180
0.695*
0.499***
0.248
(-0.350)
(1.721)
(20.890)
(1.452)
(lnRF)2*CPU
0.013
-0.059*
-0.034***
-0.029**
(0.350)
(-1.956)
(-18.402)
(-2.305)
Cycle
-0.048***
-0.047***
(-3.660)
(-3.590)
L.CPU
0.414***
0.417***
(30.370)
(30.630)
L.lnY
0.744***
0.764***
(532.935)
(359.579)
L2.lnY
0.208***
0.211***
(192.369)
(103.144)
Constant
-0.001
3.166***
2.119***
-8.478***
0.325***
0.763***
1.776***
-2.175***
(0.010)
(25.144)
(2.710)
(-9.985)
(34.500)
(10.822)
(5.119)
(-2.931)
Control
YES
YES
YES
YES
YES
YES
YES
YES
Pvalue of DWH
0.000
0.000
P value of Sargan
0.812
0.469
Cragg-D F(10%)
528.380(19.93)
538.265(19.93)
AR(1)
0.000
0.000
AR(2)
0.636
0.672
Hansen
1.000
0.997
N
5434
5434
5148
5148
5720
5720
R2
0.783
0.758
0.972
0.973
***p<0.01”, “**p<0.05”, “*p<0.10. Z-values are indicated in parentheses in columns (2) and (4)-(6), while t-values are indicated in parentheses in the remaining columns.”Control Yes” indicates the inclusion of control variables in the regression.
Columns (5) and (6) in Table 6 present the estimation results of the dynamic GMM model. Given the persistence characteristics of grain yields over time, this study further incorporates lagged terms l.lncl and l2.lncf to isolate potential unobservable factors. First, AR tests indicate first-order serial correlation but no second-order serial correlation, confirming no significant serial correlation in the original model’s error term. Second, Hansen tests yield P-values exceeding 0.05, confirming the validity of the over-identification constraint. Estimation results show that the direction and significance of interaction coefficients align with the main model, indicating that CPU’s negative regulatory effect remains reliable after mitigating endogeneity issues.
The final two columns in Table 6 present the estimation results after incorporating additional control variables. This study selected four indicators—livestock manure utilization rate, effective irrigated area, total agricultural machinery power, and GDP growth rate—from four perspectives that may influence grain production: soil quality, irrigation coverage, agricultural technology, and economic shocks. These were included in the control variable set. Based on the interaction term coefficients, the reliability of the moderation effect results was further validated.
Heterogeneity test
The above studies show that two major factors of climate change significantly affect grain yields, and that the CPU can exacerbate the negative impacts of climate change on grain yields. China is a vast country with different scales across the country. On this basis, the scope of the study is further subdivided to investigate the different impacts of climate change as well as the moderating effect of CPU on grain yields in different regions, which can help to more accurately predict and evaluate the potential risks in agricultural production. Based on previous studies, the heterogeneity analysis will be carried out in terms of both major grain-producing and non-major grain-producing regions, as well as southern and northern regions, while this study also incorporates the heterogeneity analysis of climate-adaptive policy pilot cities and non-climate-adaptive policy pilot cities.
Heterogeneity between major grain-producing and non-major grain-producing regions
Based on the agricultural resource endowments, crop structure characteristics, and primary grain production functions of China’s various provinces, the Ministry of Finance issued the “Opinions on Reforming and Improving Comprehensive Agricultural Development Policies” in 2003, categorizing China’s 31 provinces into two main types: primary grain-producing regions and non-primary grain-producing regions (Lu et al., 2020). Given the differences in agricultural production conditions and policy allocations between the two regions, we explore the heterogeneity of climate and CPU fluctuations in these regions through subsample regression analysis.
From Table 7, it can be seen that there is a significant inverted U-shaped relationship between climatic factors (temperature and precipitation) and grain yields in both major grain-producing and non-major grain-producing areas in China, with the relationship of “promoting first and suppressing later”. However, the coefficients of the quadratic cross-multipliers of CPU with temperature and precipitation in grain-producing areas are not significant, but the coefficients of the quadratic cross-multipliers of CPU with precipitation in non-major grain-producing areas are significantly negative at 1%, indicating that CPU in non-major grain-producing areas further strengthens the inverted U-shape relationship between climate change and grain yields, and magnifies the systemic risk of climate change.
Heterogeneity test results for major grain-producing and non-major grain producing regions.
Variable
Major grain producing regions
Non-major rain producing regions
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
Baseline -TP
Baseline -RF
Moderate -TP/CPU
Moderate -RF/CPU
Baseline -TP
Baseline -RF
Moderate -TP/CPU
Moderate-RF/CPU
lnTP
0.473
0.595*
2.238***
1.913***
(1.481)
(1.812)
(2.968)
(2.814)
(lnTP)2
-0.185**
-0.199**
-0.513***
-0.471***
(-2.200)
(-2.319)
(-3.014)
(-3.048)
lnTP*CPU
-0.179*
0.689***
(-1.859)
(3.711)
(lnTP)2*CPU
0.022
-0.186***
(0.985)
(-5.045)
CPU
-0.002
-0.010
0.027**
0.010
(-0.266)
(-1.409)
(2.335)
(1.023)
lnRF
2.096***
1.635***
1.764***
1.133***
(6.673)
(5.187)
(5.630)
(3.776)
(lnRF)2
-0.146***
-0.112***
-0.121***
-0.075***
(-6.345)
(-4.828)
(-5.452)
(-3.506)
lnRF*CPU
0.225
-0.047
(1.168)
(-0.226)
(lnRF)2*CPU
-0.023
-0.005
(-1.570)
(-0.349)
Constant
4.149***
-3.385***
3.868***
-1.913*
0.848
-3.333***
1.404*
-1.162
(12.472)
(-3.166)
(11.312)
(-1.784)
(0.944)
(-2.922)
(1.737)
(-1.082)
Control
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Entity FE
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Time FE
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
N
3180
3180
3180
3180
2540
2540
2540
2540
R2
0.262
0.298
0.282
0.315
0.090
0.112
0.199
0.174
The values in parentheses are t statistics. ***, **, and * indicate significance at the confidence levels of<0.01,<0.05, and< 0.1. “Control Yes” indicates the inclusion of control variables in the regression.
The possible reason of such heterogeneity between major grain-producing and non-major grain-producing regions may be that grain-producing regions have well-developed infrastructure, rich experience in coping with climate change, strong policy support, and sound market regulation mechanisms, which can effectively buffer the impacts of CPU and safeguard the stability of grain yields. However, in non-major grain-producing areas where agricultural infrastructure is weak and most of the local small-scale farmers have limited funds, an increase in the CPU will lead to a lack of guaranteed sustained investment in agricultural infrastructure construction in the region, affecting grain output. At the same time, local agricultural products are more dependent on external markets, and CPUs will interfere with market expectations and prices, making it difficult for farmers to adjust production in a timely manner, weakening their ability to resist risks and affecting the stable supply of grain. Under the dual risks of extreme weather and CPU constraints, the comparative advantage of major grain-producing regions over non- major grain-producing regions lies in their long-term, stable national policies that have invested in building a multidimensional resilient ecosystem (Shangguan et al., 2025). This result of CPU’s heterogeneous effect underscores that future policy priorities in grain-producing regions should focus on upgrading from “yield resilience” to “green industrial chain resilience,” while leveraging their radiating influence to drive development in non-grain-producing areas (Ren et al., 2025).
Heterogeneity between the northern and southern regions
China’s climate exhibits a pronounced north-south differentiation across natural zones, with the division between northern and southern regions primarily based on the Qinling-Huaihe Line as a natural geographical boundary. The climatic differences between the north and the south are mainly reflected in rainfall, temperature and seasons (Lei et al., 2023). The south has abundant and relatively stable precipitation, warm temperatures, and is prone to high temperatures in summer; the north has less total precipitation and uneven seasonal distribution, and moderate temperatures. This climatic difference leads to differences in the climate dependence of agriculture between the north and the south.
As can be seen from Table 8, the primary terms of temperature and precipitation in both the southern and northern regions of China significantly and positively affect grain yields, and their squared terms significantly and negatively affect grain yields, suggesting that the climatic factors (temperature and precipitation) in both regions have an inverted U-shaped relationship with grain yields. The coefficients of the quadratic cross-multiplication terms of CPU and temperature were significant at the 1% level in the southern region, while the quadratic cross-multiplication terms of precipitation were not significant. The probable reason is that precipitation is abundant and stable in the south, so the grain water demand is guaranteed and less affected by the fluctuation of precipitation, and at the same time, the water conservancy facilities in the south are perfect, and the ability to cope with the changes of precipitation is strong. However, with the intensification of climate warming, summer high temperature events in the South have become more frequent and more intense, resulting in significant fluctuations in grain yields. Measures to cope with temperature changes in the South are more influenced by policies, such as the construction of sunshades and the adoption of heat-resistant crop varieties that require policy and financial support, making them more vulnerable to CPU. At the same time, in the face of CPU, farmers will choose a more conservative production strategy and may reduce agricultural inputs in areas at risk of high temperatures, thus affecting grain production. The coefficient of the quadratic cross-multiplier term between CPU and precipitation is significant at the 1% level in the northern region, while the quadratic cross-multiplier term for temperature is not significant. The probable reason for this is the higher latitude of the northern region, where grain yields is relatively less affected by the increase in temperature due to warming. With relatively low total precipitation and uneven seasonal distribution in the northern region, grain growth is susceptible to the risk of water shortage. Simultaneously, the lack of investment in water infrastructure to meet the needs of agricultural production, and the CPU’s further constraints on the maintenance and upgrading of facilities, the promotion of new irrigation technologies, and the ability to prevent flooding and drainage in the event of a sudden downpour make grain production capacity even more vulnerable in the face of extreme precipitation.
Heterogeneity test results for the south region and the north region.
Variable
North region
South region
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
Baseline -TP
Baseline -RF
Moderate -TP/CPU
Moderate -RF/CPU
Baseline -TP
Baseline -RF
Moderate -TP/CPU
Moderate-RF/CPU
lnTP
0.637*
0.769**
4.276*
7.130***
(1.842)
(2.142)
(1.916)
(2.945)
(lnTP)2
-0.222**
-0.243**
-0.754*
-1.278***
(-2.362)
(-2.518)
(-1.872)
(-2.949)
lnTP*CPU
-0.345***
5.441***
(-2.809)
(9.345)
(lnTP)2*CPU
0.062**
-1.023***
(2.074)
(-9.873)
CPU
-0.008
-0.092***
-0.157***
0.028
(-0.653)
(-5.498)
(-5.771)
(1.590)
lnRF
1.117***
0.611
1.493**
0.792
(2.972)
(1.575)
(2.351)
(1.194)
(lnRF)2
-0.075***
-0.035
-0.101**
-0.053
(-2.593)
(-1.157)
(-2.312)
(-1.167)
lnRF*CPU
1.312***
-0.081
(4.746)
(-0.119)
(lnRF)2*CPU
-0.112***
-0.002
(-4.963)
(-0.047)
Constant
3.843***
0.079
3.612***
1.649
-3.330
-2.801
-7.015**
-0.155
(10.321)
(0.064)
(9.444)
(1.300)
(-1.053)
(-1.207)
(-2.026)
(-0.064)
Control
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Entity FE
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Time FE
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
N
2600
2600
2600
2600
3120
3120
3120
3120
R2
0.143
0.167
0.165
0.188
0.175
0.175
0.262
0.193
The values in parentheses are t statistics. ***, **, and * indicate significance at the confidence levels of<0.01,<0.05, and< 0.1.”Control Yes” indicates the inclusion of control variables in the regression.
Heterogeneity between climate-adapted and non-climate-adapted cities
To address the challenges posed by climate change, China launched the first batch of pilot projects for building climate- adapted cities in 2017, aiming to comprehensively enhance the capacity of cities to adapt to climate change. Twenty-eight cities were selected for the first batch of pilot projects, covering three regions: eastern, central and western (He et al., 2025). To explore whether climate resilience policies have a positive impact on climate resilience of grain production, we conducted a subsample regression analysis. It is worth noting that the development of climate-resilient cities in China remains at an early stage, with a relatively limited overall number of such cities. This directly resulted in a sample size of only 100 climate-resilient cities (N = 100) being included in the sub-sample regression analysis of this study. The limitations of the sample size may to some extent reduce the statistical power and robustness of the regression results.
Table 9 shows that the coefficients of the quadratic cross-multiplier terms of CPU with both temperature and precipitation in climate-adapted cities are not significant, but the coefficients of the quadratic cross-multiplier terms of CPU with temperature and precipitation in non-climate-adapted cities are significant and negative at the 1% level, indicating that CPU in non-climate-adapted cities further strengthens the inverted U-shaped relationship between climate change and grain yields, amplifying the systematic risk of climate change, whereas CPU does not significantly affect the climate resilience of grain yields in climate-adapted cities. The likely reason for this is that non-adapted cities could merely implement climate-resilient policies and may have unstable policy support systems (including agricultural insurance, disaster subsidies, and technology extension), which brought risks posed by CPU. However, long-term implementation of climate adaptation policies may enable farmers to accumulate a wealth of coping experience and develop a strong adaptive capacity, which enables them to adjust their production measures in response to changes in temperature and precipitation in a timely manner, and to minimize the negative impacts of CPUs on grain yields. This finding provides crucial evidence for new insights into the theory of agricultural adaptation in grain production. Against the backdrop of increasingly frequent extreme weather events, shifting from passive responses—such as attempting to reduce CPU while adapting to extreme weather—to proactive development transforms climate resilience into an endogenous driver for agricultural green upgrading (Bailey-Serres et al., 2019; Zheng et al., 2024). Merely defending against risks is static and passive. The long-term goal should leverage stability to transition agricultural systems toward more productive and sustainable forms (Pretty et al., 2018).
Heterogeneity test results for climate-adapted and non-climate-adapted cities.
Variable
Adapted cities
Non–adapted cities
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
Baseline -TP
Baseline -RF
Moderate -TP/CPU
Moderate -RF/CPU
Baseline -TP
Baseline -RF
Moderate -TP/CPU
Moderate-RF/CPU
lnTP
-5.487**
-5.348**
0.531
0.659**
(-2.382)
(-2.203)
(1.621)
(1.986)
(lnTP)2
1.027*
0.994
-0.148*
-0.154*
(1.738)
(1.618)
(-1.804)
(-1.877)
lnTP*CPU
-0.643
0.325***
(-0.701)
(3.990)
(lnTP)2*CPU
0.137
-0.112***
(0.709)
(-6.352)
CPU
0.033
-0.001
-0.007
-0.013**
(1.148)
(-0.058)
(-1.167)
(-2.390)
lnRF
-0.492
-0.815
2.070***
1.405***
(-0.465)
(-0.663)
(8.714)
(6.094)
(lnRF)2
0.040
0.065
-0.143***
-0.096***
(0.529)
(0.737)
(-8.328)
(-5.687)
lnRF*CPU
0.850
0.404***
(0.408)
(2.828)
(lnRF)2*CPU
-0.062
-0.039***
(-0.421)
(-3.634)
Constant
12.660***
6.864
12.334***
7.998*
3.155***
-3.933***
2.791***
-1.652**
(4.865)
(1.652)
(4.368)
(1.683)
(8.956)
(-4.751)
(7.724)
(-2.067)
Control
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Entity FE
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Time FE
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
N
100
100
100
100
5620
5620
5620
5620
R2
0.191
0.134
0.207
0.142
0.153
0.186
0.239
0.241
The values in parentheses are t statistics. ***, **, and * indicate significance at the confidence levels of<0.01,<0.05, and< 0.1.”Control Yes” indicates the inclusion of control variables in the regression.
Discussion
Against the backdrop of escalating climate threats to global food security, this study employs an economic-climate model (C-D-C) to systematically validate a significant inverted U-shaped relationship between climate factors (temperature, precipitation) and total grain yields. The findings corroborate previous literature on the nonlinear relationship between climate and yields (Chen et al., 2016; Guntukula, 2020; Schlenker et al., 2006), aligning with principles of agronomy and ecology (Schauberger et al., 2017; Song et al., 2022; Tschurr et al., 2023). Additionally, it provides empirical insights into how policy uncertainty shapes extreme temperatures’ impact on grain productio—while scholars have noted institutional conditions’ direct effects on grain production (Gao et al., 2021; Hua et al., 2022; Zhu et al., 2024), research on institutional role in moderating the climate-grain production relationship remains limited, with existing studies indicating that agricultural infrastructure and insurance subsidies can mitigate climate adverse impacts (Wang et al., 2022; Wu et al., 2019) and agricultural R&D investment enhances crop adaptability (Bollington et al., 2021; Carter et al., 2021).
What sets this study apart is its integration of CPU as a moderating variable within the climate-agriculture framework, emphasizing policy quality and stability over mere existence, and ambiguous or discontinuous climate policies may act as “institutional barriers” to agricultural adaptation (Oberlack, 2017). Specifically, high CPU exposes farmers to dual climate and policy uncertainties (Mittenzwei et al., 2017), which increases the “waiting value” of investments (Yanore et al., 2023) and delays adaptive adoption. Consistent with RO theory, CPU enhances the option value of waiting for future information, leading farmers to delay irreversible long-term investments (e.g., drip irrigation) amid policy ambiguities and leaving systems vulnerable to climate risks (Dai, 2025; Ginbo et al., 2021). From the perspective of AC theory, CPU erodes key elements—resource availability and action freedom—thereby reducing farmers’ability to manage risks via consistent resource access and planning (Lennert et al., 2024). In summary, CPU amplifies climate impacts through dual mechanisms: raising the “waiting premium” on long-term investments and weakening immediate “action capacity.” Notably, CPU’s moderating effect also presents spatially heterogeneous. In detail, non-major grain-producing areas are more sensitive to CPU shocks due to weaker infrastructure, insufficient institutional support, and higher market dependency, which supports the “institutional buffer” hypothesis (Nilsson et al., 2012). In contrast, climate adaptation pilot cities show reduced grain yields sensitivity to CPU, underscoring the role of policy credibility in boosting agricultural resilience (Awokuse et al., 2024), while regionally, southern areas are more sensitive to CPU-temperature interactions and northern regions are more susceptible to CPU-precipitation interactions, reflecting differences in agricultural climate exposure and adaptive capacity (Li et al., 2016). Furthermore, three theoretical sub-pathways can explain how CPU amplifies extreme weather’s impact on grain production: firstly, farmers adopt precautionary strategies to prepare for policy changes, curtailing adaptive inputs for climate risks; secondly, CPU reduces long-term signal reliability, leading to short-term experience reliance and slower, narrower system adjustments; thirdly, CPU prompts caution in public goods supply and long-term financing and reduces long-term adaptive technology R&D, though adaptive cities and staple grain regions can mitigate these issues via stable policies, forming resilient assets unaffected by CPU. Through aforementioned results, we further contend that policies enhancing the climate resilience of agricultural systems should not only focus on infrastructure and subsidies, but also incorporate Nature-based Solutions into their core strategy. By harnessing natural processes, Nature-based Solutions can deliver critical regulatory services—such as temperature regulation, water retention, and carbon sequestration—in a more cost-effective and synergistic manner. Crucially, once established, many Nature-based Solutions measures exhibit robust self-sustaining ecosystem service functions with reduced subsequent fiscal dependency. This capacity can partially buffer the disruption caused by CPU to long-term agricultural investments.
The findings of this paper provide significant implications for China and nations worldwide. FromChina’s standpoint, as a leading global grain producer and consumer (He et al., 2022), its energy and agricultural transformation is steered by the “dual carbon” objectives (achieving carbon peak by 2030 and carbon neutrality by 2060) (Chen et al., 2022; Dang et al., 2025). While this policy ensures long-term sustainability and fosters green transition, short-term adjustments may entail some challenges (Hao et al., 2022). Based on previous research and our findings, three key approaches can be summarized to mitigate the impact of policy uncertainty on grain producers who may be affected in the short term. Firstly, long-term expectations for this policy should be stabilized to reduce systemic risks in agricultural production (Li J. et al., 2024). Uncertainty in climate policies hinders producers’ investment in green agricultural innovations and sustainable agricultural machinery (Lai et al., 2026; Sun et al., 2024). Therefore, the design of the “dual carbon” policy should establish a clear long-term pathway, reduce frequent adjustments, and create stable fiscal and financial support tailored for agriculture (Zhai and Hu, 2024). This will stabilize expectations for small and micro agricultural producer, preventing policy volatility from undermining their adaptive capacity (Zhang X. et al., 2025). Second, promote the strengthening of adaptation capacity (based on the AC theory) among agricultural entities to mitigate the inverted U-shaped risks posed by extreme weather (Chepkoech et al., 2020; Kangogo et al., 2020). Building on the established inverted U-shaped relationship with climate change, supporting policies for “dual carbon” should invest in climate-resilient infrastructure and promote adaptive agricultural technologies (Elahi et al., 2022; Mthembu et al., 2025; Webb et al., 2017), which will enhance the agricultural system’s capacity to withstand climate fluctuations, flatten the inverted U-shaped curve (reducing the slope in Figures 2, 3). Third, implement differentiated policies by region to avoid a one-size-fits-all approach(Li L. et al., 2023). Tailored emission reductiontargets, technology promotion lists, and compensation mechanisms should be established based on each region’s position on the inverted U-curve, variations in green production efficiency, and agricultural carbon emission characteristics (Xia et al., 2024; Zhang X. et al., 2023). Moreover, in line with the rural revitalization policy requirements (the “Comprehensive Plan for Promoting Rural Revitalization (2024-2027)”, it is imperative to establish market-driven incentive mechanisms, such as agricultural carbon sink markets and horizontal ecological compensation, harmonizing the “dual carbon” goals with rural revitalization aspirations (Song et al., 2024; Tang et al., 2023). Additionally, leveraging international experience, a 10–15 year cross-cycle plan should be formulated, integrating fiscal support with climate resilience enhancement and emission reduction and carbon sequestration targets, thereby converting inverted U-shaped risks into stable income streams (Bréchet et al., 2013; Fodha and Yamagami, 2025).
From an international standpoint, extending the relevant analysis to other developing nations, such as rice-producing countries in Southeast Asia and smallholder economies in Africa, uncovers inherent instability within their policy frameworks (Aryal et al., 2021; Liu et al., 2020). Variables like fluctuations in international climate financing and domestic government transitions intensify CPU (Ma et al., 2024; Monasterolo et al., 2019), prompting resource-constrained agricultural producers to opt for conservative survival strategies rather than climate-adaptive innovative investments. This creates a vicious cycle of “policy uncertainty → inadequate investment → diminished adaptability → declining grain production → worsening poverty”. The primary finding of this study is that CPU significantly undermines the climate resilience of grain production systems. This challenge is particularly pronounced in developing countries experiencing rapid industrialization and policy shift (Borsetta et al., 2025; Fahad and Hossain, 2025), offering valuable insights for regions with comparable global climate conditions and developmental stages to tackle climate change and safeguard food security. Moreover, it is imperative to further investigate methods for balancing short-term, medium-term, and long-term policy interests within a framework that ensures positive social externalities, while steering clear of policy discontinuity (Medina Hidalgo et al., 2022; Vogt-Schilb and Hallegatte, 2017).
In a word, in addition to extend the application of Gavriilidis (2021) CPU concept to the agricultural sector, the study also provides a critical theoretical lens for understanding the complex decision-making environment faced by developing countries under the triple pressures of “development-emission reduction-adaptation.” Our findings underscore the critical importance of establishing stable, transparent, and continuous climate policy frameworks for developing countries—particularly for advancing Nature-based Solutions to address the impacts of climate change within the Global South, where policy stability could unlock the potential of ecological approaches to enhance agricultural resilience. For China, this implies that while advancing macro-level climate strategies, it must provide clear, predictable implementation guidelines and long-term incentives in the agricultural sector to stabilize producer confidence (Ju et al., 2020; Wu, 2023). For the international community, it means providing financial and technical assistance to help vulnerable developing countries build more resilient policy systems and implementation capacities (Rauniyar et al., 2025; Savvidou et al., 2021).
Conclusions and policy implications
This study examines the nonlinear effects of climate change factors (temperature and precipitation) on grain yields and innovatively explores the moderating role of CPU in this relationship, using panel data for 286 prefecture-level cities in China from 2001 to 2020. The study draws the following conclusions: (1) Climate change factors (temperature and precipitation) have a significant inverted U-shaped relationship with grain yields. Moderate increases in temperature and precipitation initially boost grain yields, but exceeding ecological thresholds may have adverse effects, leading to yield declines. (2) CPU exacerbates the negative impacts of climate extremes by enabling farmers to invest less in adaptation and make conservative production decisions. the higher the level of CPU, the steeper the inverted U-shape curve, indicating that the adverse impacts of climate change on grain yields are more pronounced in high CPU environments. (3) The impact of the CPU exhibits regional heterogeneity. Specifically, non-major grain-producing regions are more susceptible to CPU-amplified climate risks due to weaker agricultural infrastructure; precipitation fluctuates widely in the north, and precipitation extremes are more likely to be amplified by CPU effects; hot weather is frequent in the south, and temperature extremes are more likely to be amplified by CPU effects; and non-climate-adapted cities without adaptive climates aren’t able to mitigate the CPU effects by stabilizing policies and infrastructure.
Based on the research findings, this study proposes the following policy recommendations using China as an example, aiming to provide insights and references for national grain systems—particularly those in developing countries with high gain system vulnerability—to address climate change and extreme weather events.
First, strengthen the capacity building of agriculture to adapt to climate change: on the one hand, the government should increase the construction of advanced climate monitoring and early-warning system facilities to provide accurate and timely climate data and early-warning information, and to help farmers to make informed decisions and optimize the use of their resources under the changing climate conditions, reduce potential losses and enhance the resilience of the grain system. For instance, enhance the integrated digital support platform for “monitoring-early warning-decision-making,” leveraging digital technologies to develop disaster warning modules for crops such as rice, wheat, and corn (e.g., high-temperature heat damage, drought, and waterlogging warnings) (Rojas, 2021; Zhang W. et al., 2023). Utilize channels like mobile apps to achieve simultaneous, precise delivery of warning information alongside adaptive cultivation techniques (e.g., irrigation and fertilizer adjustments). On the other hand, the promotion of drought- and flood-tolerant crop varieties, the improvement of water-saving irrigation techniques, as well as the optimization of the layout of agricultural production and the allocation of resources, will increase the adaptive capacity of agriculture to climate change. Disaster-avoidance cropping systems (such as adjusting planting dates and developing conservation tillage) could be enhanced by establishing county-level demonstration farms and implementing subsidy programs.
Second, enhancing climate resilience through policy stabilization: Governments should strive to reduce the uncertainty of climate policy and provide a stable policy environment for agricultural production, while actively advancing Nature-based Solutions to address the impacts of climate change within the Global South. The agricultural departments could lead the development of a medium-to-long-term action plan for agricultural adaptation to climate change. By 2030 and 2060, a phased approach may be established to achieve emission reduction and carbon sequestration targets, build adaptive capacity indicators, promote key technologies, and establish a central government funding framework. Moreover, annual progress reports shall be published in accordance with the plan and stable “ecological subsidies” will directly benefit farmers implementing adaptation measures. Through these measures, the Government can help farmers anticipate and adapt to climate and policy changes, minimize the adverse impacts of the CPU, and encourage long-term investment in climate change mitigation and technological innovation (Khatri-Chhetri et al., 2021).
Third, develop targeted regional strategies:
(1) for non-major grain-producing regions that are subject to greater fluctuations in CPU, the Government should take measures to buffer the impact of CPU and increase the resilience of the grain system by strengthening infrastructure, improving water resource management and providing specialized agricultural extension services (Osumba et al., 2021; Priya et al., 2025). By encouraging leading agricultural enterprises in major grain-producing regions to form partnerships with leading enterprises in non-major grain-producing regions, and through methods such as contract farming and technical management services, non-grain-producing production areas can benefit from the resilient production models and market channels of major grain-producing regions. In non-major grain-producing regions, funds promoting production resilience can be established. By linking funding to improvements in local climate resilience indicators, these funds can incentivize local agricultural producers—including small-scale farmers and agricultural cooperatives—to implement measures addressing climate risks.
(2) Accurate policy implementation should be tailored to regional variations. Southern China could establish medium-to-long-term protection mechanisms against high-temperature disasters, including a special long-term fiscal program for “heat-adaptive planting technologies.” This would ensure the continuity and predictability of support for heat-tolerant varieties, shade structures, and related resources (Hao and He, 2025; Ma et al., 2022). while innovating financial instruments to diversify high-temperature production risks (e.g., heatwave insurance) that automatically trigger payouts when consecutive high temperatures reach specific thresholds. these instruments stabilize farmers’ income expectations while countering their tendency to reduce investments due to concerns over policy uncertainty risks. In northern China, combining conservation tillage and dryland farming with the restoration of native vegetation to comprehensively enhance soil health and water retention capacity (Nature-based Solutions). Simultaneously, promoting long-term maintenance agreements for water conservancy facilities can partially insulate against CPU disruptions by contractually locking in the attention and responsibilities of government and social capital. Moreover, in financial instruments, developing “dual drought and flood insurance” enhances farmers’ confidence in weathering precipitation fluctuations in the context of CPU risks.
(3) The scope of climate adaptation pilot projects could be further expanded. Building upon the ongoing advancement of pilot climate-adaptive cities, and drawing on the advanced experience of adaptive cities (such as establishing high-density meteorological observation networks and utilizing smart platforms for precise disaster early warning—a typical example is Shenzhen City’s “electronic water gauge” flood warning system), develop and train a “Climate Adaptive Grain Production Decision Support System” tailored for non-adaptive cities. In the long run, this system could contribute to integrating localized climate forecasts, resilient technical solutions, and policy tool information, while enhancing adaptation tools such as urban climate change health risk monitoring and assessment, and comprehensive indicators for heat island, rain island, drought island, and smog island effects. Thus, those successful practices, technologies and experiences that have already yielded significant results in climate-resilient cities will be systematically replicated in other regions.
This study still has limitations that require further refinement in subsequent research. Although this study selected annual mean temperature and annual total precipitation as climate factors, which may have obscured the effects of water and heat stress on grain crops during their critical growth stages. Future research should refine climate indicators by focusing on crop critical growth stages and employing more precise metrics such as accumulated temperature, days of extreme temperatures, and precipitation concentration to achieve accurate analysis of the climate-yield relationship. The research unit could be extended to the county level, utilizing county-level grain production data and policy text data to improve variable matching. In this study, we have employed two-stage least squares and system GMM methods to address endogeneity issues. Moving forward, we will utilize more cutting-edge approaches such as machine learning to mitigate endogeneity concerns. Additionally, this study introduces the moderating effect of CPU into the field of grain production, and future research could introduce CPU into other areas of agricultural sector research, such as agro-ecosystem services, rural economic development, and agricultural supply chain management, in order to comprehensively assess the impacts of CPU on the sustainable development of agriculture in various aspects.
Data availability statement
The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/Supplementary Material.
The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Generative AI statement
The author(s) declared that generative AI was not used in the creation of this manuscript.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
Supplementary material
The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fevo.2026.1753076/full#supplementary-material
ReferencesAbbasS. (2022).
Climate change and major crop production: Evidence from Pakistan. Environ. Sci. pollut. Res.29, 5406–5414. doi: 10.1007/s11356-021-16041-4, PMID: 34417972AbbasM.RibeiroP. F.SantosJ. L. (2023).
Farming system change under different climate scenarios and its impact on food security: An analytical framework to inform adaptation policy in developing countries. Mitig. Adapt. Strateg. Global Change28, 43. doi: 10.1007/s11027-023-10082-5AgnolucciP.RaptiC.AlexanderP.De LipsisV.HollandR. A.EigenbrodF.. (2020).
Impacts of rising temperatures and farm management practices on global yields of 18 crops. Nat. Food1, 562–571. doi: 10.1038/s43016-020-00148-x, PMID: 37128016AlmazrouiM.SaeedF.SaeedS.IsmailM.EhsanM. A.IslamM. N.. (2021).
Projected changes in climate extremes using CMIP6 simulations over SREX regions. Earth Syst. Environ.5, 481–497. doi: 10.1007/s41748-021-00250-5AltieriM. A.NichollsC. I.HenaoA.LanaM. A. (2015).
Agroecology and the design of climate change-resilient farming systems. Agron. Sustain. Dev.35, 869–890. doi: 10.1007/s13593-015-0285-2AntónJ.CattaneoA.KimuraS.LankoskiJ. (2013).
Agricultural risk management policies under climate uncertainty. Global Environ. Change23, 1726–1736. doi: 10.1016/j.gloenvcha.2013.08.007ArshadA.RazaM. A.ZhangY.ZhangL.WangX.AhmedM.. (2021).
Impact of climate warming on cotton growth and yields in China and Pakistan: A regional perspective. Agriculture11, 97–119. doi: 10.3390/agriculture11020097AryalJ. P.SapkotaT. B.RahutD. B.MarenyaP.StirlingC. M. (2021).
Climate risks and adaptation strategies of farmers in east Africa and south Asia. Sci. Rep.11, 10489. doi: 10.1038/s41598-021-89391-1, PMID: 34006938AssanE.SuvediM.Schmitt OlabisiL.AllenA. (2018).
Coping with and adapting to climate change: A gender perspective from smallholder farming in Ghana. Environments5, 86. doi: 10.3390/environments5080086AwokuseT.LimS.SanteramoF.SteinbachS. (2024).
Robust policy frameworks for strengthening the resilience and sustainability of agri-food global value chains. Food Policy127, 102714. doi: 10.1016/j.foodpol.2024.102714BaigI. A.MohammadS.AkramV.ChandioA. A.GuptaY. (2023).
Examining the impacts of climatological factors and technological advancement on wheat production: A road framework for sustainable grain production in India. Environ. Dev. Sustain.26, 12193–12217. doi: 10.1007/s10668-023-03746-4Bailey-SerresJ.ParkerJ. E.AinsworthE. A.OldroydG. E. D.SchroederJ. I. (2019).
Genetic strategies for improving crop yields. Nature575, 109–118. doi: 10.1038/s41586-019-1679-0, PMID: 31695205BiW.SongY.LiuY.LiZ.ZhangY. (2022).
Food consumption structure and food security—Through mediating effect analysis of agricultural R&D and agricultural investment. Int. J. Environ. Res. Public Health19, 12524. doi: 10.3390/ijerph191912524, PMID: 36231821BollingtonA.DeLongeM.MungraD.HayekM.SaifuddinM.McDermidS. S. (2021).
Closing research investment gaps for a global food transformation. Front. Sustain. Food Syst.5. doi: 10.3389/fsufs.2021.794594BorsettaG.ZoviA.VittoriS. (2025).
Long-term frameworks for food security and sustainability through climate-smart interconnected agrifood systems. Sci7, 15–29. doi: 10.3390/sci7010015BréchetT.HritonenkoN.YatsenkoY. (2013).
Adaptation and mitigation in long-term climate policy. Environ. Resour. Econ55, 217–243. doi: 10.1007/s10640-012-9623-xBrockC.HadenV. R. (2024).
Amish and non-amish farmer perspectives on climate change causes, effects, and adaptation strategies. Sustainability16, 897. doi: 10.3390/su16020897BurkeM.HsiangS. M.MiguelE. (2015).
Global non-linear effect of temperature on economic production. Nature527, 235–239. doi: 10.1038/nature15725, PMID: 26503051CabasJ.WeersinkA.OlaleE. (2010).
Crop yield response to economic, site and climatic variables. Clim. Change101, 599–616. doi: 10.1007/s10584-009-9754-4ÇakirM. A.MercanY.İÇyerN. C.BozkurtF. (2025).
Post-disaster food safety and food security: An example of the Türkiye earthquake. Food Secur.17, 671–685. doi: 10.1007/s12571-025-01539-3CarterR.ChoulartonR.FerdinandT.DingH.GinoyaN.PreethanP. (2021). Food systems at risk: Transformative adaptation for long-term food security (Washington, DC, United States:
WRI World Resources Institute). doi: 10.46830/wrirpt.19.00042CarterC. A.ZhongF.ZhuJ. (2012).
Advances in chinese agriculture and its global implications. Appl. Econ. Perspect. Policy34, 1–36. doi: 10.1093/aepp/ppr047ChallinorA. J.ParkesB.Ramirez-VillegasJ. (2015).
Crop yield response to climate change varies with cropping intensity. Global Change Biol.21, 1679–1688. doi: 10.1111/gcb.12808, PMID: 25581316ChandioA. A.GokmenogluK. K.DashD. P.KhanI.AhmadF.JiangY. (2024).
Exploring the energy-climate-agriculture (ECA) nexus: A roadmap toward agricultural sustainability in asian countries. Environ. Dev. Sustain27, 12769–12795. doi: 10.1007/s10668-023-04418-zChandioA. A.OzdemirD.JiangY. (2023).
Modelling the impact of climate change and advanced agricultural technologies on grain output: Recent evidence from China. Ecol. Model.485, 110501. doi: 10.1016/j.ecolmodel.2023.110501ChapagainP. S.BanskotaT. R.ShresthaS.KhanalN. R.YiliZ.YanJ.. (2025).
Studies on adaptive capacity to climate change: A synthesis of changing concepts, dimensions, and indicators. Humanit. Soc. Sci. Commun.12, 331. doi: 10.1057/s41599-025-04453-3ChenS.ChenX.XuJ. (2016).
Impacts of climate change on agriculture: Evidence from China. J. Environ. Econ Manage.76, 105–124. doi: 10.1016/j.jeem.2015.01.005ChenM.CuiY.JiangS.ForsellN. (2022).
Toward carbon neutrality before 2060: Trajectory and technical mitigation potential of non-CO2 greenhouse gas emissions from chinese agriculture. J. Clean. Prod.368, 133186. doi: 10.1016/j.jclepro.2022.133186ChepkoechW.MungaiN. W.StöberS.Lotze-CampenH. (2020).
Understanding adaptive capacity of smallholder african indigenous vegetable farmers to climate change in Kenya. Climate Risk Manage.27, 100204. doi: 10.1016/j.crm.2019.100204ChouJ.XuY.DongW.ZhaoW.LiJ.LiY. (2021).
An economy-climate model for quantitatively projecting the impact of future climate change and its application. Front. Phys.9. doi: 10.3389/fphy.2021.723306ChouJ. M.YeD. (2006).
Assessing the effect of climate changes on grains yields with a new economy-climate model. Clim. Environ. Res.11, 347–353.
ClappJ.MoseleyW. G.BurlingameB.TermineP. (2022).
Viewpoint: The case for a six-dimensional food security framework. Food Policy106, 102164. doi: 10.1016/j.foodpol.2021.102164CobbC. W.DouglasP. H. (1928).
A theory of production. Am. Econ. Rev.18, 139–165.
DaiJ. (2025).
Impact of climate policy uncertainty on agriculture development: multidimensional analysis from land use, food structure, and carbon emissions. Land. Degrad. Dev.36, 4545–4561. doi: 10.1002/ldr.5652DangH.DengY.HaiY.ChenH.WangW.ZhangM.. (2025).
Integrating geodetector and GTWR to unveil spatiotemporal heterogeneity in China’s agricultural carbon emissions under the dual carbon goals. Agriculture15, 1302. doi: 10.3390/agriculture15121302De LaporteA.SchuurmanD.SkolrudT.SladeP.WeersinkA. (2024).
Business risk management programs and the adoption of beneficial management practices in canadian crop agriculture. Can. J. Agric. Econ/Rev. Can. D’agroecon.72, 309–324. doi: 10.1111/cjag.12349DeschênesO.GreenstoneM. (2007).
The economic impacts of climate change: Evidence from agricultural output and random fluctuations in weather. Am. Econ. Rev.97, 354–385. doi: 10.1257/aer.97.1.354DingQ.ZhuX.ZhangS. (2025).
Asymmetric effects of climate policy uncertainty on the renewable energy consumption. Environ. Dev. Sustain1–34. doi: 10.1007/s10668-025-06144-0DongW.ChouJ.FengG. (2007).
A new economic assessment index for the impact of climate change on grain yield. Adv. Atmos. Sci.24, 336–342. doi: 10.1007/s00376-007-0336-yDrebenstedtI.MarhanS.PollC.KandelerE.HögyP. (2023).
Annual cumulative ambient precipitation determines the effects of climate change on biomass and yield of three important field crops. Field Crops Res.290, 108766. doi: 10.1016/j.fcr.2022.108766ElahiE.KhalidZ.TauniM. Z.ZhangH.LirongX. (2022).
Extreme weather events risk to crop-production and the adaptation of innovative management strategies to mitigate the risk: A retrospective survey of rural punjab, Pakistan. Technovation117, 102255. doi: 10.1016/j.technovation.2021.102255Emami MeybodiM.OwjimehrS. (2024).
The nexus of climate policy uncertainty, renewable energy and energy inequality under income disparities. Energy Strateg. Rev.53, 101381. doi: 10.1016/j.esr.2024.101381FahadS.HossainM. (2025).
Advancing climate resilience in agri-food systems: The role of green innovation and strategic policy interventions. Food Energy Secur.14, e70122. doi: 10.1002/fes3.70122FanchoneA.NelsonL.DodetN.MartinL.AndrieuN. (2022).
How agro-environmental and climate measures are affecting farming system performances in Guadeloupe?: Lessons for the design of effective climate change policies. Int. J. Agric. Sustain.20, 1348–1359. doi: 10.1080/14735903.2022.2136836FatimaZ.AhmedM.HussainM.AbbasG.Ul-AllahS.AhmadS.. (2020).
The fingerprints of climate warming on cereal crops phenology and adaptation options. Sci. Rep.10, 18013. doi: 10.1038/s41598-020-74740-3, PMID: 33093541FindlaterK. M.SatterfieldT.KandlikarM. (2019).
Farmers’ risk-based decision making under pervasive uncertainty: Cognitive thresholds and hazy hedging. Risk Anal.39, 1755–1770. doi: 10.1111/risa.13290, PMID: 30830976FodhaM.YamagamiH. (2025).
Mitigation or adaptation to climate change? The role of fiscal policy. Macroecon. Dynam.29, e105. doi: 10.1017/S1365100525000252GaoR.CaiH.XuX. (2025).
Climate variability and agricultural inputs: Effects on grain production and yield stability in China, (1991–2020). Food Energy Secur.14, e70096. doi: 10.1002/fes3.70096GaoY.ShuY.CaoH.ZhouS.ShiS. (2021).
Fiscal policy dilemma in resolving agricultural risks: Evidence from China’s agricultural insurance subsidy pilot. Int. J. Environ. Res. Public Health18, 7577. doi: 10.3390/ijerph18147577, PMID: 34300028GaoD.ZhangT.LiuX. (2025).
The urban renewable energy transition: impact assessment and transmission mechanisms of climate policy uncertainty. Energies18, 2089. doi: 10.3390/en18082089GargA.BordoloiS.GanesanS. P.SekharanS.SahooL. (2020).
A relook into plant wilting: Observational evidence based on unsaturated soil–plant-photosynthesis interaction. Sci. Rep.10, 22064. doi: 10.1038/s41598-020-78893-z, PMID: 33328549GavriilidisK. (2021).
Measuring climate policy uncertainty. SSRN Electron. J. doi: 10.2139/ssrn.3847388GinboT.Di CoratoL.HoffmannR. (2021).
Investing in climate change adaptation and mitigation: A methodological review of real-options studies. Ambio50, 229–241. doi: 10.1007/s13280-020-01342-8, PMID: 32451968GrigorievaE.LivenetsA.StelmakhE. (2023).
Adaptation of agriculture to climate change: A scoping review. Climate11, 202. doi: 10.3390/cli11100202GuntukulaR. (2020).
Assessing the impact of climate change on Indian agriculture: Evidence from major crop yields. J. Public Aff.20, e2040. doi: 10.1002/pa.2040GuoQ.HeZ.WangZ. (2024).
Monthly climate prediction using deep convolutional neural network and long short-term memory. Sci. Rep.14, 17748. doi: 10.1038/s41598-024-68906-6, PMID: 39085577GuoK.LiY.ZhangY.JiQ.ZhaoW. (2023).
How are climate risk shocks connected to agricultural markets? J. Commod. Mark.32, 100367. doi: 10.1016/j.jcomm.2023.100367HaansR. F. J.PietersC.HeZ.-L. (2016).
Thinking about U: Theorizing and testing U- and inverted U-shaped relationships in strategy research. Strat. Manage. J.37, 1177–1195. doi: 10.1002/smj.2399HallegatteS. (2009).
Strategies to adapt to an uncertain climate change. Global Environ. Change19, 240–247. doi: 10.1016/j.gloenvcha.2008.12.003HanX.HsuS.LiJ.AnR. (2023).
Economic policy uncertainty, non-financial enterprises’ shadow banking activities and stock price crash risk. Emerg. Mark. Rev.54, 101003. doi: 10.1016/j.ememar.2023.101003HansenL. P. (2022).
Central banking challenges posed by uncertain climate change and natural disasters. J. Monet. Econ125, 1–15. doi: 10.1016/j.jmoneco.2021.09.010HaoJ.ChenL.ZhangN. (2022).
A statistical review of considerations on the implementation path of China’s “double carbon” goal. Sustainability14, 11274. doi: 10.3390/su141811274HaoC.HeS. (2025).
Research on atmospheric temperature change and its adaptive strategies in winter wheat and summer maize dual-cropping region of North China Plain during 1961–2021—A case study of Jiaozuo City in Henan Province. PloS One20, e0325781. doi: 10.1371/journal.pone.0325781, PMID: 40577319HaoS.WangG.YangY.ZhaoS.HuangS.LiuL.. (2024).
Promoting grain production through high-standard farmland construction: Evidence in China. J. Integr. Agric.23, 324–335. doi: 10.1016/j.jia.2023.11.021HaoW.YangJ.HuX.ZhangZ.ShiZ.ZhouH. (2024).
Can fiscal expenditure for agriculture mitigate the impact of climate change on agricultural production? Front. Sustain. Food Syst.8. doi: 10.3389/fsufs.2024.1349840HarknessC.ArealF. J.SemenovM. A.SenapatiN.ShieldI. F.BishopJ. (2023).
Towards stability of food production and farm income in a variable climate. Ecol. Econ204, 107676. doi: 10.1016/j.ecolecon.2022.107676HeW.GuoX.ZhangC. (2025).
Assessing the impact of climate-resilient city development on urban sustainability: evidence from China. Sustainability17, 4381. doi: 10.3390/su17104381HeJ.LiZ.ZhangX.WangH.DongW.DuE.. (2022).
Towards carbon neutrality: A study on China’s long-term low-carbon transition pathways and strategies. Environ. Sci. Ecotechnol.9, 100134. doi: 10.1016/j.ese.2021.100134, PMID: 36157858HerathH. M. L.ThirumarpanK. (2017).
Climate change induced adaptation strategies by paddy farmers: Special emphasis on socio economic insights. J. Agric. Sci. – Sri. Lanka.12, 124–137. doi: 10.4038/jas.v12i2.8230HoffmanA. L.KemanianA. R.ForestC. E. (2018).
Analysis of climate signals in the crop yield record of sub-saharan africa. Global Change Biol.24, 143–157. doi: 10.1111/gcb.13901, PMID: 28892592HoganD.SchlenkerW. (2024).
Non-linear relationships between daily temperature extremes and US agricultural yields uncovered by global gridded meteorological datasets. Nat. Commun.15, 4638. doi: 10.1038/s41467-024-48388-w, PMID: 38821938HolstR.YuX.GrünC. (2013).
Climate change, risk and grain yields in China. J. Integr. Agric.12, 1279–1291. doi: 10.1016/S2095-3119(13)60435-9HongN. T. H.KienP. T.LinhH. G.ThanhN. V. H.TuanN. L.AnhP. D. (2024).
Do climate policy uncertainty and economic policy uncertainty promote firms’ green activities? Evidence from an emerging market. Cogent. Econ Financ.12, 2307460. doi: 10.1080/23322039.2024.2307460HuaW.ChenZ.LuoL. (2022).
The effect of the major-grain-producing-areas oriented policy on crop production: evidence from China. Land11, 1375. doi: 10.3390/land11091375HuangN.SongY.WangJ.ZhangZ.MaS.JiangK.. (2022).
Climatic threshold of crop production and climate change adaptation: A case of winter wheat production in China. Front. Ecol. Evol.10. doi: 10.3389/fevo.2022.1019436HuangT.SunZ. (2024).
Climate policy uncertainty and firm investment. Int. J. Financ. Econ29, 4358–4371. doi: 10.1002/ijfe.2881Intergovernmental Panel On Climate Change (IPCC) (2022). Climate change 2022 – impacts, adaptation and vulnerability: Working group II contribution to the sixth assessment report of the intergovernmental panel on climate change. 1st ed (Cambridge, UK:
Cambridge University Press). doi: 10.1017/9781009325844JuH.LiuQ.LiY.LongX.LiuZ.LinE. (2020).
Multi-stakeholder efforts to adapt to climate change in China’s agricultural sector. Sustainability12, 8076. doi: 10.3390/su12198076JustD. R.PetersonH. H. (2003).
Diminishing marginal utility of wealth and calibration of risk in agriculture. Am. J. Agric. Econ85, 1234–1241. doi: 10.1111/j.0092-5853.2003.00536.xKabatoW.GetnetG. T.SinoreT.NemethA.MolnárZ. (2025).
Towards climate-smart agriculture: Strategies for sustainable agricultural production, food security, and greenhouse gas reduction. Agronomy15, 565–589. doi: 10.3390/agronomy15030565KanI.ReznikA.KaminskiJ.KimhiA. (2023).
The impacts of climate change on cropland allocation, crop production, output prices and social welfare in Israel: A structural econometric framework. Food Policy115, 102311. doi: 10.1016/j.foodpol.2022.102311KangogoD.DentoniD.BijmanJ. (2020).
Determinants of farm resilience to climate change: the role of farmer entrepreneurship and value chain collaborations. Sustainability12, 868–883. doi: 10.3390/su12030868KhalilA. M.MurchieE. H.MooneyS. J. (2020).
Quantifying the influence of water deficit on root and shoot growth in wheat using X-ray Computed Tomography. AoB. Plants12, plaa036. doi: 10.1093/aobpla/plaa036, PMID: 32905427KhanN. A.ChowdhuryA.ShahA. A.KhanP.AlotaibiB. A. (2023).
The institutional support index: A pragmatic approach to assessing the effectiveness of institutions’ climate risk management support-a case study of farming communities in Pakistan. Climate Risk Manage.42, 100560. doi: 10.1016/j.crm.2023.100560KhanM. R.RobertsJ. T. (2013).
Adaptation and international climate policy. WIREs. Climate Change4, 171–189. doi: 10.1002/wcc.212Khatri-ChhetriA.SapkotaT. B.SanderB. O.ArangoJ.NelsonK. M.WilkesA. (2021).
Financing climate change mitigation in agriculture: Assessment of investment cases. Environ. Res. Lett.16, 124044. doi: 10.1088/1748-9326/ac3605KittS.AxsenJ.LongZ.RhodesE. (2021).
The role of trust in citizen acceptance of climate policy: Comparing perceptions of government competence, integrity and value similarity. Ecol. Econ183, 106958. doi: 10.1016/j.ecolecon.2021.106958KogutB.KulatilakaN. (2001).
Capabilities as real options. Organ. Sci.12, 744–758. doi: 10.1287/orsc.12.6.744.10082KongQ.WangZ.WuP.PengD. (2025).
Climate policy uncertainty and synergistic industrial agglomeration: What role does financial development play? J. Int. Money. Financ.152, 103282. doi: 10.1016/j.jimonfin.2025.103282KukalM. S.IrmakS. (2018).
Climate-driven crop yield and yield variability and climate change impacts on the U.S. great plains agricultural production. Sci. Rep.8, 3450. doi: 10.1038/s41598-018-21848-2, PMID: 29472598KyseláE.ŠčasnýM.ZvěřinováI. (2019).
Attitudes toward climate change mitigation policies: A review of measures and a construct of policy attitudes. Climate Policy19, 878–892. doi: 10.1080/14693062.2019.1611534LaiY.ZhouJ.YaoL. (2026).
Signal of sustainable agriculture! Nonlinear impact of climate policy uncertainty: New evidence from China’s agricultural sector. Sustain. Dev.34, 127–147. doi: 10.1002/sd.70167LaysonS. K. (2015).
The increasing returns to scale CES production function and the law of diminishing marginal returns. South. Econ. J.82, 408–415. doi: 10.4284/0038-4038-2013.202LeeC.-C.ZengM.LuoK. (2024).
How does climate change affect food security? Evidence from China. Environ. Impact. Assess. Rev.104, 107324. doi: 10.1016/j.eiar.2023.107324LeeC.-C.ZhaoY.-N. (2023).
Heterogeneity analysis of factors influencing CO2 emissions: The role of human capital, urbanization, and FDI. Renewable Sustain. Energy Rev.185, 113644. doi: 10.1016/j.rser.2023.113644LeiH.LiF.LiuC.LiuY. (2024).
Cognitive load and economic decision making of smallholder farmers in China: An experimental study. Curr. Psychol.43, 465–480. doi: 10.1007/s12144-023-04288-8LeiY.LiaoR.SuY.ZhangX.LiuD.ZhangL. (2023).
Variation characteristics of temperature and rainfall and their relationship with geographical factors in the qinling mountains. Atmosphere14, 696–716. doi: 10.3390/atmos14040696LennertJ.KovácsK.KoósB.SwainN.BálintC.HamzaE.. (2024).
Climate change, pressures, and adaptation capacities of farmers: Empirical evidence from Hungary. Horticulturae10, 56. doi: 10.3390/horticulturae10010056LiC. (2023).
Climate change impacts on rice production in Japan: A cobb-douglas and panel data analysis. Ecol. Indic.147, 110008. doi: 10.1016/j.ecolind.2023.110008LiR.GengS. (2013).
Impacts of climate change on agriculture and adaptive strategies in China. J. Integr. Agric.12, 1402–1408. doi: 10.1016/S2095-3119(13)60552-3LiL.HanJ.ZhuY. (2023).
Does environmental regulation in the form of resource agglomeration decrease agricultural carbon emissions? Quasi-natural experimental on high-standard farmland construction policy. J. Clean. Prod.420, 138342. doi: 10.1016/j.jclepro.2023.138342LiS.Juhász-HorváthL.HarrisonP. A.PintérL.RounsevellM. D. A. (2017).
Relating farmer’s perceptions of climate change risk to adaptation behaviour in Hungary. J. Environ. Manage.185, 21–30. doi: 10.1016/j.jenvman.2016.10.051, PMID: 28029477LiJ.KongT.GuL. (2024).
The impact of climate policy uncertainty on green innovation in chinese agricultural enterprises. Financ. Res. Lett.62, 105145. doi: 10.1016/j.frl.2024.105145LiD.LiZ.LiuZ.YangY.KhosoA. G.WangL.. (2023).
Climate change simulations revealed potentially drastic shifts in insect community structure and crop yields in China’s farmland. J. Pest Sci.96, 55–69. doi: 10.1007/s10340-022-01479-3LiY.LiF.YangF.XieX.YinL. (2020).
Spatiotemporal impacts of climate change on food production: Case study of shaanxi province, China. Environ. Sci. pollut. Res.27, 19826–19835. doi: 10.1007/s11356-020-08447-3, PMID: 32222925LiJ.LinQ. (2022).
Can the adjustment of China’s grain purchase and storage policy improve its green productivity? Int. J. Environ. Res. Public Health19, 6310. doi: 10.3390/ijerph19106310, PMID: 35627849LiC.MaoX.ZhengM.HanM. (2024).
Adapting seasonal rice cultivation strategies for food security in response to climate change impacts. Sustainability16, 6748. doi: 10.3390/su16166748LiX.PhilpJ.CremadesR.RobertsA.HeL.LiL.. (2016).
Agricultural vulnerability over the chinese loess plateau in response to climate change: Exposure, sensitivity, and adaptive capacity. Ambio45, 350–360. doi: 10.1007/s13280-015-0727-8, PMID: 26563383LiX.TakahashiT.SuzukiN.KaiserH. M. (2011).
The impact of climate change on maize yields in the United States and China. Agric. Syst.104, 348–353. doi: 10.1016/j.agsy.2010.12.006LinF.LiC.XuB.ChenJ.ChenA.HassanM. A.. (2023).
Late spring cold reduces grain number at various spike positions by regulating spike growth and assimilate distribution in winter wheat. Crop J.11, 1272–1278. doi: 10.1016/j.cj.2023.03.014LindJ. T.MehlumH. (2010).
With or without U? The appropriate test for a U-shaped relationship*: practitioners’ corner. Oxf. Bull. Econ Stat72, 109–118. doi: 10.1111/j.1468-0084.2009.00569.xLippertC.KrimlyT.AurbacherJ. (2009).
A ricardian analysis of the impact of climate change on agriculture in Germany. Clim. Change97, 593–610. doi: 10.1007/s10584-009-9652-9LiuY.CuiJ.JiangH.YanH. (2023).
Do county financial marketization reforms promote food total factor productivity growth?: A mechanistic analysis of the factors quality of land, labor, and capital. Front. Sustain. Food Syst.7. doi: 10.3389/fsufs.2023.1263328LiuG.LuoK.XuP.ZhangS. (2023).
Climate policy uncertainty and its impact on major grain futures. Financ. Res. Lett.58, 104412. doi: 10.1016/j.frl.2023.104412LiuF.SuC. W.TaoR.LobonţO.-R. (2024).
Does economic and climate policy uncertainty matter the oil market? Resour. Policy95, 105188. doi: 10.1016/j.resourpol.2024.105188LiuJ.WangM.YangL.RahmanS.SriboonchittaS. (2020).
Agricultural productivity growth and its determinants in south and southeast asian countries. Sustainability12, 4981. doi: 10.3390/su12124981LobellD. B.FieldC. B. (2007).
Global scale climate–crop yield relationships and the impacts of recent warming. Environ. Res. Lett.2, 14002. doi: 10.1088/1748-9326/2/1/014002LobellD. B.HammerG. L.McLeanG.MessinaC.RobertsM. J.SchlenkerW. (2013).
The critical role of extreme heat for maize production in the United States. Nat. Climate Change3, 497–501. doi: 10.1038/nclimate1832LoboguerreroA.CampbellB.CooperP.HansenJ.RosenstockT.WollenbergE. (2019).
Food and Earth systems: Priorities for climate change adaptation and mitigation for agriculture and food systems. Sustainability11, 1372. doi: 10.3390/su11051372LuC.LiuA.XiaoY.LiuX.XieG.ChengS. (2020).
Changes in China’s grain production pattern and the effects of urbanization and dietary structure. J. Resour. Ecol.11, 358. doi: 10.5814/j.issn.1674-764x.2020.04.004LuoQ. (2011).
Temperature thresholds and crop production: A review. Clim. Change109, 583–598. doi: 10.1007/s10584-011-0028-6LuqmanM.MahmoodF.Al-AnsariT. (2023).
Supporting sustainable global food security through a novel decentralised offshore floating greenhouse. Energy Convers. Manage.277, 116577. doi: 10.1016/j.enconman.2022.116577MaY.GugaS.XuJ.SuY.LiuX.TongZ.. (2022).
Agricultural vulnerability assessment of high-temperature disaster in shaanxi province of China. Agriculture12, 980. doi: 10.3390/agriculture12070980MaY.-R.LiuZ.MaD.ZhaiP.GuoK.ZhangD.. (2023).
A news-based climate policy uncertainty index for China. Sci. Data10, 881. doi: 10.1038/s41597-023-02817-5, PMID: 38065994MaD.ZhangD.GuoK.JiQ. (2024).
Coupling between global climate policy uncertainty and economic policy uncertainty. Financ. Res. Lett.69, 106180. doi: 10.1016/j.frl.2024.106180MaitahM.MalecK.MaitahK. (2021).
Influence of precipitation and temperature on maize production in the Czech Republic from 2002 to 2019. Sci. Rep.11, 10467. doi: 10.1038/s41598-021-89962-2, PMID: 34001991MaresD. J.MrvaK. (2014).
Wheat grain preharvest sprouting and late maturity alpha-amylase. Planta240, 1167–1178. doi: 10.1007/s00425-014-2172-5, PMID: 25257145MasudS.KhanA. (2024).
Policy implementation barriers in climate change adaptation: The case of Pakistan. Environ. Policy Govern.34, 42–52. doi: 10.1002/eet.2054Medina HidalgoD.NunnP. D.BeazleyH.BurkhartS.RantesJ. (2022).
Adaptation, sustainable food systems and healthy diets: An analysis of climate policy integration in Fiji and Vanuatu. Climate Policy22, 1130–1145. doi: 10.1080/14693062.2022.2095969MenapaceL.ColsonG.RaffaelliR. (2013).
Risk aversion, subjective beliefs, and farmer risk management strategies. Am. J. Agric. Econ95, 384–389. doi: 10.1093/ajae/aas107MirónI. J.LinaresC.DíazJ. (2023).
The influence of climate change on food production and food safety. Environ. Res.216, 114674. doi: 10.1016/j.envres.2022.114674, PMID: 36341795MittenzweiK.PerssonT.HöglindM.KværnøS. (2017).
Combined effects of climate change and policy uncertainty on the agricultural sector in Norway. Agric. Syst.153, 118–126. doi: 10.1016/j.agsy.2017.01.016MonasteroloI.RoventiniA.FoxonT. J. (2019).
Uncertainty of climate policies and implications for economics and finance: An evolutionary economics approach. Ecol. Econ163, 177–182. doi: 10.1016/j.ecolecon.2019.05.012MthembuB. E.CeleT.MkhizeX. (2025).
Climate change impacts on agricultural infrastructure and resources: Insights from communal land farming systems. Land14, 1150. doi: 10.3390/land14061150NamK. (2021).
Investigating the effect of climate uncertainty on global commodity markets. Energy Econ96, 105123. doi: 10.1016/j.eneco.2021.105123NilssonM.ZamparuttiT.PetersenJ. E.NykvistB.RudbergP.McGuinnJ. (2012).
Understanding policy coherence: Analytical framework and examples of sector–environment policy interactions in the EU. Environ. Policy Govern.22, 395–423. doi: 10.1002/eet.1589NiuX.SongZ.XuC.WuH.LuanQ.JiangJ.. (2023).
Prediction of needle physiological traits using UAV imagery for breeding selection of slash pine. Plant Phenomics.5, 28. doi: 10.34133/plantphenomics.0028, PMID: 36939412NoaillyJ.NowzohourL.Van Den HeuvelM. (2022). Does environmental policy uncertainty hinder investments towards a low-carbon economy? (Cambridge, MA:
National Bureau of Economic Research). doi: 10.3386/w30361OberlackC. (2017).
Diagnosing institutional barriers and opportunities for adaptation to climate change. Mitig. Adapt. Strateg. Global Change22, 805–838. doi: 10.1007/s11027-015-9699-zOnyeakaH.NwauzomaU. M.AkinsemoluA. A.TamasigaP.DuanK.Al-SharifyZ. T.. (2024).
The ripple effects of climate change on agricultural sustainability and food security in Africa. Food Energy Secur.13, e567. doi: 10.1002/fes3.567OsabuteyE. L. C.JacksonT. (2024).
Mobile money and financial inclusion in Africa: Emerging themes, challenges and policy implications. Technol. Forecast. Soc. Change202, 123339. doi: 10.1016/j.techfore.2024.123339OsbahrH.TwymanC.Neil AdgerW.ThomasD. S. G. (2008).
Effective livelihood adaptation to climate change disturbance: Scale dimensions of practice in Mozambique. Geoforum39, 1951–1964. doi: 10.1016/j.geoforum.2008.07.010OsumbaJ. J. L.RechaJ. W.OromaG. W. (2021).
Transforming agricultural extension service delivery through innovative bottom–up climate-resilient agribusiness farmer field schools. Sustainability13, 3938. doi: 10.3390/su13073938OuattaraP. D.KouassiE.EgbendeweA. Y. G.AkinkugbeO. (2018).
Climate uncertainty and agricultural soil conservation investment decisions. Climate Change Econ9, 1850004. doi: 10.1142/S2010007818500045PandaA.SahuN.BeheraS.SayamaT.SahuL.AvtarR.. (2019).
Impact of climate variability on crop yield in kalahandi, bolangir, and koraput districts of odisha, India. Climate7, 126. doi: 10.3390/cli7110126PandeyD. K.MishraR. (2024).
Towards sustainable agriculture: Harnessing AI for global food security. Artif. Intell. Agric.12, 72–84. doi: 10.1016/j.aiia.2024.04.003PrettyJ.BentonT. G.BharuchaZ. P.DicksL. V.FloraC. B.GodfrayH. C. J.. (2018).
Global assessment of agricultural system redesign for sustainable intensification. Nat. Sustain.1, 441–446. doi: 10.1038/s41893-018-0114-0PriyaN. K.KhatriA.KarkiP.SamotaS. D.VishwakarmaS. K.SukdeveE. K.. (2025).
The important role of extension services in strengthening the capacity of farmers’ resilience to climate change in India. J. Exp. Agric. Int.47, 204–223. doi: 10.9734/jeai/2025/v47i33329QiuW.MaX.CaoH.HuangT.SheX.HuangM.. (2022).
Improving wheat yield by optimizing seeding and fertilizer rates based on precipitation in the summer fallow season in drylands of the loess plateau. Agric. Water Manage.264, 107489. doi: 10.1016/j.agwat.2022.107489RauniyarS. K.ShiibaN.YamasakiL.TomoiH.TomoiH.WestM. K.. (2025).
Climate-related bilateral official development assistance (ODA) and vulnerability: A comparative study of allocation and effectiveness. PloS Climate4, e0000382. doi: 10.1371/journal.pclm.0000382RayD. K.GerberJ. S.MacDonaldG. K.WestP. C. (2015).
Climate variation explains a third of global crop yield variability. Nat. Commun.6, 5989. doi: 10.1038/ncomms6989, PMID: 25609225RegmiS.PaudelB. (2024).
Climate-smart agriculture: A review of sustainability, resilience, and food security. Arch. Agric. Environ. Sci.9, 832–839. doi: 10.26832/24566632.2024.0904028RenX.LiJ.HeF.LuceyB. (2023).
Impact of climate policy uncertainty on traditional energy and green markets: Evidence from time-varying granger tests. Renewable Sustain. Energy Rev.173, 113058. doi: 10.1016/j.rser.2022.113058RenC.LiuW.XiaoT.CastelliniA.MaesanoG. (2025).
The impact of technological progress and industrial structure upgrading on agricultural economic resilience: An empirical study in China. Sustainability17, 8062. doi: 10.3390/su17178062Rodríguez-BarillasM.KlerkxL.PoortvlietP. M. (2024).
Transformative policy mix or policy pandemonium? Insights from the Climate Smart Agriculture policy mix in Costa Rica. Environ. Innovation Societ. Trans.50, 100791. doi: 10.1016/j.eist.2023.100791RojasO. (2021).
Next generation agricultural stress index system (ASIS) for agricultural drought monitoring. Remote Sens.13, 959. doi: 10.3390/rs13050959SaeiM.MohammadiH.ZiaeeS.BarkhordariS. (2018).
Economic effect of climate alteration on grain production in Iran. Appl. Ecol. Environ. Res.16, 6691–6707. doi: 10.15666/aeer/1605_66916707SaudS.WangD.FahadS.AlharbyH. F.BamagoosA. A.MjrashiA.. (2022).
Comprehensive impacts of climate change on rice production and adaptive strategies in China. Front. Microbiol.13. doi: 10.3389/fmicb.2022.926059, PMID: 35875578SavvidouG.AtteridgeA.Omari-MotsumiK.TrisosC. H. (2021).
Quantifying international public finance for climate change adaptation in Africa. Climate Policy21, 1020–1036. doi: 10.1080/14693062.2021.1978053SchaubergerB.ArchontoulisS.ArnethA.BalkovicJ.CiaisP.DeryngD.. (2017).
Consistent negative response of US crops to high temperatures in observations and crop models. Nat. Commun.8, 13931. doi: 10.1038/ncomms13931, PMID: 28102202SchlenkerW.HanemannW. M.FisherA. C. (2006).
The impact of global warming on U.S. agriculture: An econometric analysis of optimal growing conditions. Rev. Econ Stat88, 113–125. doi: 10.1162/rest.2006.88.1.113SchlenkerW.RobertsM. J. (2009).
Nonlinear temperature effects indicate severe damages to U.S. crop yields under climate change. Proc. Natl. Acad. Sci.106, 15594–15598. doi: 10.1073/pnas.0906865106, PMID: 19717432SeabornT.GriffithD.KliskeyA.CaudillC. C. (2021).
Building a bridge between adaptive capacity and adaptive potential to understand responses to environmental change. Global Change Biol.27, 2656–2668. doi: 10.1111/gcb.15579, PMID: 33666302SemeraroT.ScaranoA.LeggieriA.CalisiA.De CaroliM. (2023).
Impact of climate change on agroecosystems and potential adaptation strategies. Land12, 1117. doi: 10.3390/land12061117ShahJ.AlharthiM. (2022).
The association between farmers’ Psychological factors and their choice to adopt risk management strategies: the case of Pakistan. Agriculture12, 412. doi: 10.3390/agriculture12030412ShangguanC.MaR.LuY.ZhangH.TianM.TengY.. (2025).
Assessment of grain production levels and analysis of obstacle factors in China: A sustainable, stable and green perspective. Front. Sustain. Food Syst.9. doi: 10.3389/fsufs.2025.1631845SharpE. B.DaleyD. M.LynchM. S. (2011).
Understanding local adoption and implementation of climate change mitigation policy. Urban. Aff. Rev.47, 433–457. doi: 10.1177/1078087410392348ShehzadK. (2023).
Extreme flood in Pakistan: Is Pakistan paying the cost of climate change? A short communication. Science of the Total Environment. 880, 162973. doi: 10.1016/j.scitotenv.2023.162973, PMID: 36958563SimpsonN. P.WilliamsP. A.MachK. J.Berrang-FordL.BiesbroekR.HaasnootM.. (2023).
Adaptation to compound climate risks: A systematic global stocktake. iScience26, 105926. doi: 10.1016/j.isci.2023.105926, PMID: 36866045SinghC.DorwardP.OsbahrH. (2016).
Developing a holistic approach to the analysis of farmer decision-making: Implications for adaptation policy and practice in developing countries. Land. Use Policy59, 329–343. doi: 10.1016/j.landusepol.2016.06.041SinghA. S.EanesF.ProkopyL. S. (2020).
Climate change uncertainty among american farmers: An examination of multi-dimensional uncertainty and attitudes towards agricultural adaptation to climate change. Clim. Change162, 1047–1064. doi: 10.1007/s10584-020-02860-wSkarbøK.VanderMolenK. (2016).
Maize migration: Key crop expands to higher altitudes under climate change in the andes. Climate Dev.8, 245–255. doi: 10.1080/17565529.2015.1034234SnitkerA. J.YungL.MetcalfE. C.BocinskyR. K.HassaneinN.JenscoK.. (2024).
How agricultural producers use local knowledge, climate information, and on-farm “experiments” to address drought risk. Agric. Hum. Vals.41, 1857–1875. doi: 10.1007/s10460-024-10582-3SongC.HuangX.LesO.MaH.LiuR. (2022).
The economic impact of climate change on wheat and maize yields in the north China plain. Int. J. Environ. Res. Public Health19, 5707. doi: 10.3390/ijerph19095707, PMID: 35565106SongS.SuM.KongL.KongM.MaY. (2024).
Assessing the economic value of carbon sinks in farmland using a multi-scenario system dynamics model. Agriculture15, 69–90. doi: 10.3390/agriculture15010069SunG.FangJ.LiT.AiY. (2024).
Effects of climate policy uncertainty on green innovation in chinese enterprises. Int. Rev. Financ. Anal.91, 102960. doi: 10.1016/j.irfa.2023.102960SunT.-T.SuC.-W.MirzaN.UmarM. (2021).
How does trade policy uncertainty affect agriculture commodity prices? Pac-Basin. Financ. J.66, 101514. doi: 10.1016/j.pacfin.2021.101514TangX.HaoG.WangY.LiD.LuoY. (2023).
Research on an equilibrium development model between urban and rural areas of henan including carbon sink assets under the dual carbon goal. Front. Environ. Sci.10. doi: 10.3389/fenvs.2022.1037286TengZ.ChenY.MengS.DuanM.ZhangJ.YeN. (2023).
Environmental stimuli: A major challenge during grain filling in cereals. Int. J. Mol. Sci.24, 2255. doi: 10.3390/ijms24032255, PMID: 36768575TettehB.BaidooS. T.TakyiP. O. (2022).
The effects of climate change on food production in Ghana: Evidence from maki, (2012) cointegration and frequency domain causality models. Cogent. Food Agric.8, 2111061. doi: 10.1080/23311932.2022.2111061TianL.LiX. (2023).
Does climate policy uncertainty affect carbon emissions in China? A novel dynamic ARDL simulation perspective. Humanit. Soc. Sci. Commun.10, 1–10. doi: 10.1057/s41599-023-02102-1TongQ.SwallowB.ZhangL.ZhangJ. (2019).
The roles of risk aversion and climate-smart agriculture in climate risk management: Evidence from rice production in the Jianghan Plain, China. Climate Risk Manage.26, 100199. doi: 10.1016/j.crm.2019.100199TrigeorgisL.ReuerJ. J. (2017).
Real options theory in strategic management. Strat. Manage. J.38, 42–63. doi: 10.1002/smj.2593TschurrF.KirchgessnerN.HundA.KronenbergL.AndereggJ.WalterA.. (2023).
Frost damage index: The antipode of growing degree days. Plant Phenomics.5, 104. doi: 10.34133/plantphenomics.0104, PMID: 37799632Vasyl’YevaO. (2021).
Assessment of factors of sustainable development of the agricultural sector using the cobb-douglas production function. Baltic. J. Econ. Stud.7, 37–49. doi: 10.30525/2256-0742/2021-7-2-37-49VermeulenS. J.AggarwalP. K.AinslieA.AngeloneC.CampbellB. M.ChallinorA. J.. (2012).
Options for support to agriculture and food security under climate change. Environ. Sci. Policy15, 136–144. doi: 10.1016/j.envsci.2011.09.003VermeulenS. J.DineshD.HowdenS. M.CramerL.ThorntonP. K. (2018).
Transformation in practice: A review of empirical cases of transformational adaptation in agriculture under climate change. Front. Sustain. Food Syst.2. doi: 10.3389/fsufs.2018.00065VogelE.DonatM. G.AlexanderL. V.MeinshausenM.RayD. K.KarolyD.. (2019).
The effects of climate extremes on global agricultural yields. Environ. Res. Lett.14, 054010. doi: 10.1088/1748-9326/ab154bVogt-SchilbA.HallegatteS. (2017).
Climate policies and nationally determined contributions: Reconciling the needed ambition with the political economy. WIREs. Energy Environ.6, e256. doi: 10.1002/wene.256WaldmanK. B.ToddP. M.OmarS.BlekkingJ. P.GirouxS. A.AttariS. Z.. (2020).
Agricultural decision making and climate uncertainty in developing countries. Environ. Res. Lett.15, 113004. doi: 10.1088/1748-9326/abb909WangH.ChengX.ShiQ.XuJ.ChenD.LuoC.. (2023).
Cold tolerance identification of nine rosa L. materials and expression patterns of genes related to cold tolerance in rosa hybrida. Front. Plant Sci.14. doi: 10.3389/fpls.2023.1209134, PMID: 37441175WangP.DengX.JiangS. (2019).
Global warming, grain production and its efficiency: Case study of major grain production region. Ecol. Indic.105, 563–570. doi: 10.1016/j.ecolind.2018.05.022WangK.-H.KanJ.-M.QiuL.XuS. (2023).
Climate policy uncertainty, oil price and agricultural commodity: From quantile and time perspective. Econ. Anal. Policy78, 256–272. doi: 10.1016/j.eap.2023.03.013WangM.LiD.LiuX.ChenC.FreyB.SuiX.. (2024).
Global hierarchical meta-analysis to identify the factors for controlling effects of antibiotics on soil microbiota. Environ. Int.192, 109038. doi: 10.1016/j.envint.2024.109038, PMID: 39357259WangH.LiuH.WangD. (2022).
Agricultural insurance, climate change, and food security: Evidence from chinese farmers. Sustainability14, 9493. doi: 10.3390/su14159493WangJ.MishraS.SharifA.ChenH. (2024).
Dynamic spillover connectedness among green finance and policy uncertainty: Evidence from QVAR network approach. Energy Econ131, 107330. doi: 10.1016/j.eneco.2024.107330WangM. H.RykalinaO.BorzunovaT.KolobovE.TakhumovaO. (2025).
Eco-integrated agriculture: Merging environmental sustainability and energy efficiency in the modern agro-production complex. E3S. Web Conf.633, 2002. doi: 10.1051/e3sconf/202563302002WangS.TianY.LiuX.FoleyM. (2019).
How farmers make investment decisions: Evidence from a farmer survey in China. Sustainability12, 247. doi: 10.3390/su12010247WangJ.VangaS.SaxenaR.OrsatV.RaghavanV. (2018).
Effect of climate change on the yield of cereal crops: A review. Climate6, 41. doi: 10.3390/cli6020041WangW.ZhuX.LuP.ZhaoY.ChenY.ZhangS. (2024).
Spatio-temporal evolution of public opinion on urban flooding: Case study of the 7.20 Henan extreme flood event. Int. J. Dis. Risk Reduct.100, 104175. doi: 10.1016/j.ijdrr.2023.104175WarsameA. A.Sheik-AliI. A.AliA. O.SarkodieS. A. (2021).
Climate change and crop production nexus in Somalia: An empirical evidence from ARDL technique. Environ. Sci. pollut. Res.28, 19838–19850. doi: 10.1007/s11356-020-11739-3, PMID: 33410024WebbN. P.MarshallN. A.StringerL. C.ReedM. S.ChappellA.HerrickJ. E. (2017).
Land degradation and climate change: Building climate resilience in agriculture. Front. Ecol. Environ.15, 450–459. doi: 10.1002/fee.1530WebberH.LischeidG.SommerM.FingerR.NendelC.GaiserT.. (2020).
No perfect storm for crop yield failure in Germany. Environ. Res. Lett.15, 104012. doi: 10.1088/1748-9326/aba2a4WeiZ.PengJ.ZhaoY.LiX.WangC. (2025).
Reform of agricultural land property rights system and grain production resilience: Empirical evidence based on China’s “three rights separation” reform. PloS One20, e0319387. doi: 10.1371/journal.pone.0319387, PMID: 40117226WeiH.XianjunD.KhanK. (2024).
The race to zero emission: Can climate policy uncertainty threaten decarbonization? Environ. Dev. Sustain1–19. doi: 10.1007/s10668-024-05118-yWheelerT.Von BraunJ. (2013).
Climate change impacts on global food security. Science341, 508–513. doi: 10.1126/science.1239402, PMID: 23908229WoodR. S.HultquistA.RomsdahlR. J. (2014).
An examination of local climate change policies in the great plains. Rev. Policy Res.31, 529–554. doi: 10.1111/ropr.12103WooldridgeJ. M. (2020). Introductory econometrics: A modern approach (7th ed.) (Boston, MA:
Thomson South Western).
WuS. (2023).
A systematic review of climate policies in China: Evolution, effectiveness, and challenges. Environ. Impact. Assess. Rev.99, 107030. doi: 10.1016/j.eiar.2022.107030WuQ.GuanX.ZhangJ.XuY. (2019).
The role of rural infrastructure in reducing production costs and promoting resource-conserving agriculture. Int. J. Environ. Res. Public Health16, 3493. doi: 10.3390/ijerph16183493, PMID: 31546849WuJ.ZhangJ.GeZ.XingL.HanS.ShenC.. (2021).
Impact of climate change on maize yield in China from 1979 to 2016. J. Integr. Agric.20, 289–299. doi: 10.1016/S2095-3119(20)63244-0XiaY.GuoH.XuS.PanC. (2024).
Environmental regulations and agricultural carbon emissions efficiency: Evidence from rural China. Heliyon10, e25677. doi: 10.1016/j.heliyon.2024.e25677, PMID: 38370207XiaoS.HeZ.ZhangW.QinX. (2022).
The agricultural green production following the technological progress: evidence from China. Int. J. Environ. Res. Public Health19, 9876. doi: 10.3390/ijerph19169876, PMID: 36011508XiaoX.TianQ.HouS.LiC. (2019).
Economic policy uncertainty and grain futures price volatility: Evidence from China. China Agric. Econ. Rev.11, 642–654. doi: 10.1108/CAER-11-2018-0224XiaoG.ZhangF.QiuZ.YaoY.WangR.HuangJ. (2013).
Response to climate change for potato water use efficiency in semi-arid areas of China. Agric. Water Manage.127, 119–123. doi: 10.1016/j.agwat.2013.06.004XuY.MaG. (2019).
Local officials’ turnover and enterprises’ overcapacity. Econ. Res. J.54, 129–145.
XuanX.LiuB.ZhangF. (2021).
Climate change and adaptive management: Case study in agriculture, forestry and pastoral areas. Land10, 832. doi: 10.3390/land10080832YanD.LiuX.HaoX.LiJ. (2023).
Tracing environmental impacts of grain losses along the supply chain in the north China plain: An integrated framework. Resour. Conserv. Recycl.189, 106771. doi: 10.1016/j.resconrec.2022.106771YanY.XiongX.LiS.LuL. (2022).
Will temperature change reduce stock returns? Evidence from China. Int. Rev. Financ. Anal.81, 102112. doi: 10.1016/j.irfa.2022.102112YangT.ChandioA. A.ZhangA.LiuY. (2023).
Do farm subsidies effectively increase grain production? Evidence from major grain-producing regions of China. Foods12, 1435. doi: 10.3390/foods12071435, PMID: 37048255YangX.LinE.MaS.JuH.GuoL.XiongW.. (2007).
Adaptation of agriculture to warming in northeast China. Clim. Change84, 45–58. doi: 10.1007/s10584-007-9265-0YangQ.ZhengJ.ZhuH. (2020).
Influence of spatiotemporal change of temperature and rainfall on major grain yields in southern Jiangsu province, China. Global Ecol. Conserv.21, e00818. doi: 10.1016/j.gecco.2019.e00818YanoreL.SokJ.Oude LansinkA. (2023).
Anticipate, wait or don’t invest? The strategic net present value approach to study expansion decisions under policy uncertainty. Agribusiness39, 535–548. doi: 10.1002/agr.21780YinL.CaoH. (2024).
Understanding climate policy uncertainty: Evidence from temporal and spatial domains. Int. Rev. Financ. Anal.95, 103482. doi: 10.1016/j.irfa.2024.103482YinX.FengL.GongY. (2023).
Mitigating ecotoxicity risks of pesticides on ornamental plants based on life cycle assessment. Toxics11, 360. doi: 10.3390/toxics11040360, PMID: 37112587YuanX.LiS.ChenJ.YuH.YangT.WangC.. (2024).
Impacts of global climate change on agricultural production: A comprehensive review. Agronomy14, 1360. doi: 10.3390/agronomy14071360YuanS.LiX.DuE. (2021).
Effects of farmers’ behavioral characteristics on crop choices and responses to water management policies. Agric. Water Manage.247, 106693. doi: 10.1016/j.agwat.2020.106693ZagariaC.SchulpC. J. E.ZavalloniM.ViaggiD.VerburgP. H. (2021).
Modelling transformational adaptation to climate change among crop farming systems in Romagna, Italy. Agric. Syst.188, 103024. doi: 10.1016/j.agsy.2020.103024ZhaiJ.HuF. (2024).
Selection and application of China environmental sustainability policy instrumental: A quantitative analysis based on “dual carbon” policy text. Front. Environ. Sci.12. doi: 10.3389/fenvs.2024.1418253ZhangZ.ChengShiqinWangCongSong. (2025).
Climate policy uncertainty and corporate investment efficiency: Evidence from China. J. Environ. Plann. Manage.68, 957–977. doi: 10.1080/09640568.2023.2276062ZhangX.DuanX.ZhouX.PangZ.DuanH.TengY.. (2025).
Study on the behavioral decision of multiple subjects of agricultural green production under the double carbon target in China. Front. Public Health13. doi: 10.3389/fpubh.2025.1575121, PMID: 40809755ZhangW.PengL.GeX.YangL.ChenL.LiW. (2023).
Spatio-temporal knowledge graph-based research on agro-meteorological disaster monitoring. Remote Sens.15, 4403. doi: 10.3390/rs15184403ZhangH.ZhaoDanTangZiyanZhang. (2022).
Exogenous brassinosteroids promotes root growth, enhances stress tolerance, and increases yield in maize. Plant Signaling Behav.17, 2095139. doi: 10.1080/15592324.2022.2095139, PMID: 35775499ZhangX.ZhouX.LiaoK. (2023).
Regional differences and dynamic evolution of China’s agricultural carbon emission efficiency. Int. J. Environ. Sci. Technol.20, 4307–4324. doi: 10.1007/s13762-022-04196-7ZhengH.MaW.HeQ. (2024).
Climate-smart agricultural practices for enhanced farm productivity, income, resilience, and greenhouse gas mitigation: A comprehensive review. Mitig. Adapt. Strateg. Global Change29, 28–66. doi: 10.1007/s11027-024-10124-6ZhengJ.YuY.ZhangX.HaoZ. (2018).
Variation of extreme drought and flood in north China revealed by document-based seasonal precipitation reconstruction for the past 300 years. Climate Past.14, 1135–1145. doi: 10.5194/cp-14-1135-2018ZhuM.ZhengK.LiuB.JinF. (2024).
Can agricultural support and protection subsidy policies promote high-quality development of grain industry? A case study of China. Agriculture14, 1664. doi: 10.3390/agriculture14101664ZouZ.OuyangC.LiX. (2024).
Uncertainty impacts on China’s agricultural commodity futures: A quantile perspective. Appl. Econ56, 5090–5106. doi: 10.1080/00036846.2024.2364105
Edited by: Oyetola Oyebanji, Botanical Research Institute of Texas, United States
Reviewed by: Alban Kuriqi, University of Lisbon, Portugal
Raza Ali Tunio, Sichuan Agricultural University, China
Zhou Shiwei, Anhui Provincial Institute of Water Resources Science, China
It is worth noting that the turning point calculated in our model represents a “macro-statistical optimum for economic output,” rather than the “micro-physiological optimum for individual crops” often cited in the literature. Macro-aggregated data incorporates complex factors such as crop structure, cropping systems, regional variations, and non-growing season periods. We based our analysis on prefecture-level city-level data for annual, multi-crop (rice, wheat, corn, soybean, tuber and other crops.) mixed yields, using annual average temperature and precipitation. Thus, the turning point number calculated reflects the “climate-economic equilibrium point” that maximizes total annual grain output across all regions under China's current multi-crop planting structure, geographic distribution, and agricultural scheduling—not the optimal physiological temperature for any single crop at any specific stage.