We critically evaluate current techniques employed in measuring climate resilience, including machine learning, geospatial modeling, and econometric analyses. By assessing their strengths, limitations, and applicability in LMIC contexts, we refine our understanding of the methodological robustness required for accurate resilience assessment, and hence the standardization needed to enable comparability and scalability. We also identify cutting-edge applications of satellite data, critique its limitations, and highlight emerging tools for resolution and accessibility. Ultimately, we aim to help democratize the use of satellite data by recommending cost-effective, open-source solutions.

LMICs face unique challenges such as limited ground-truth data, socio-economic vulnerabilities, and diverse climate risks. We identify how analytical techniques and satellite data can be tailored to these specific contexts, underscoring the importance of integrating local knowledge, stakeholder engagement, and interdisciplinary approaches for more effective resilience measurement. Research gaps include the underrepresentation of socio-cultural dimensions of resilience and the lack of longitudinal studies tracking resilience over time. This provides a roadmap for innovations in data integration, policy alignment, and multi-scale analyses. In conclusion, we aim to provide a catalyst for climate resilience measurement in LMICs and facilitate informed decision-making and resource allocation, empowering policymakers, researchers, and practitioners to collaboratively address climate challenges in vulnerable regions.

Our systematic review employed internationally recognized literature search methodologies and includes a balanced, comprehensive, and critical view of the research area, following the PRISMA checklist (Page et al., 2021). The review was segmented into five categories: “literature search, inclusion criteria, methodological quality assessment, data extraction, and data analysis” (Page et al., 2021). The quality of each included study was evaluated, and findings critically appraised. The process comprised: identifying relevant records, selecting eligible studies, assessing the risk of bias, extracting data, and conducting a qualitative synthesis of the included studies. A meta-analysis was performed to integrate data from all pertinent studies to provide a more precise estimate than individual study-derived estimates. Results and conclusions were then summarized. Pertinent studies were drawn from peer-reviewed literature concerning interventions that utilize environmental informatics and satellite remote sensing applications (see search terms and strings; Table 1). Scopus, Web of Knowledge and IEEE Xplore were searched for EO interventions in LMICs; all studies underwent independent evaluation by the primary author, [PC]. Any issues regarding eligibility were addressed through discussions with a secondary reviewer [MG]. The search covered published studies in English (whilst English remains the predominant language for international scientific collaboration, many agencies do publish in their national languages with translation to facilitate multilingual access to satellite data and reports). Further sources were independently gathered by the primary author to identify studies that fulfilled the inclusion criteria in Table 1. Information from verified literature beyond traditional academic publishing was also scanned to gain region-specific information. Final searches were performed on January 24, 2025.

This systematic review aims to investigate good practices and successful examples of satellite remote sensing projects to deliver actionable climate and weather data, specific to Low and Middle-Income Countries (LMICs) where this information is most essential but currently is not widely accessible: “According to this year’s report on the Global Status of Multi-Hazard Early Warning Systems, the world is at its highest levels of early warning coverage since 2015. That said, progress remains uneven. Half of the countries in Africa and only 40 per cent of countries in the Americas and the Caribbean have reported the existence of Multi-Hazard Early Warning Systems” (United Nations Office for Disaster Risk Reduction and World Meteorological Organization, 2024). The LMIC classification system, utilized by international organizations such as the World Bank, categorizes nations based on their Gross National Income (GNI) per capita. Designated LMICs have a GNI per capita ranging from US$ 1,046 to US$ 4,095.

The criteria for inclusion and exclusion in the systematic review comprised of control conditions, study methodology, acceptability and effective outcomes. Predefined criteria were:

The review identified that while the literature contains numerous examples, the implementation of satellite observations is rarely conducted in a systematic manner due to the uncertainty of lasting continuity of EO missions. This may lead to underreporting within the existing literature. In addition, an additional constraint arises from the rapidly advancing field of satellite remote sensing. The technology and tools that support many of these projects are under continuous development, making project evaluation reflective only of past performance resulting in significant differences from estimated performance data. This underscores the necessity for timely strategic and impact analyses. The exclusion criteria for this review were designed to ensure the relevance and quality of the included studies (Figure 2). Studies lacking empirical data or those focused solely on conceptual models were excluded, as they do not provide actionable insights for measuring climate resilience. Research on high-income countries was excluded unless it involved comparative analyses with low-and middle-income countries, to maintain focus on the target context. Duplicate studies were removed to avoid redundancy, and non-English publications were excluded unless translation was available, as language barriers could limit the study’s accessibility. These criteria ensured a rigorous selection process. It is important to note that the scope of this paper does not permit an exhaustive account of all space agencies and their missions: each country has its own unique aims in the space sector, and many programmes, particularly from emerging spacefaring nations, may not gain widespread attention.

We used an item checklist, and a four-phase workflow process as outlined in the PRISMA flow diagram, where Figure 2 illustrates the selection method whereby titles and abstracts were retrieved. The following specific objectives guided the review process:

Much of the literature that was screened in this review demonstrates that investment in space and services components form the foundation for effective implementation of the four essential pillars of the “Multi-Hazard Early Warning Systems (MHEWS); Pillar 1: Disaster risk knowledge. Pillar 2: Detection, observation, monitoring, analysis and forecasting of hazards. Pillar 3: Warning dissemination and communication. Pillar 4: Preparedness and response. It is acknowledged that this is a crucial decade for the future of geospatial technologies, observation, and geo-information services” (WMO, 2023). However, we found that evidence of sharing scientific models and data resources is lacking: North–South, South–South and triangular regional and international cooperation in infrastructure and utilization of these technologies for actionable data and environmental knowledge exchange is needed at a global level.

Those adversely impacted by climate-related disasters, especially in LMICS, and associated economic losses, are increasing. Experience gained from implementing people-centric, early warning, alert and response systems is essential to mitigate harm to individuals, assets, and livelihoods. By initiating early action plans that are well prepared and rigorously tested, such systems minimize disaster impacts and are among the most effective and economically viable methods for risk reduction and climate adaptation. As many climate hazards transcend national borders, disaster response measures must reflect this. Enhancing early warning systems to address the needs of both current and future generations necessitates international cooperation and just transition principles to protect those who have contributed least to the climate crisis. Despite their proven efficacy, as of 2024, only “half of all countries worldwide have adequate multi-hazard early warning systems” (WMO, 2025). Such protective coverage is even more limited in developing nations, with less than half of the Least Developed Countries and only one-third of Small Island Developing States equipped with such systems. The integration of satellite systems and EO data holds significant promise in advancing risk management strategies and capabilities, particularly in LMICs.

By integrating satellite data with AI and Deep Learning, early warning systems can deliver critical insights into environmental hazards, allowing proactive risk management (Jones and Brown, 2018). Additionally, the use of satellite imagery enables real-time disaster event monitoring, aiding swift response efforts (Wilson and Patel, 2020). The wide scope of examples identified in the peer reviewed and grey literature is indicative of the potential to apply a range of satellite applications to effectively deliver risk mitigation (Table 2). While existing reviews were informative, it was found that despite the implementation of projects within low resource settings, there is a lack of “comprehensive data for evidence-based research and decision-making is crucial but scarce” (Barteit et al., 2023; Table 2). It is necessary to provide evidence beyond technological integration to encompass stakeholder coordination and innovation fostering multi-sectoral collaboration to maximize the impact of early warning systems (Albris et al., 2020). Such collaborative approaches can promote knowledge-sharing and capacity-building among diverse stakeholders, including government agencies, NGOs, and local communities (Luna, 2017).

Remotely-sensed data for drought and food insecurity assessments includes the use of rainfall, temperature, and satellite images for anomaly assessment and early warning (Senay et al., 2015; Sudianto et al., 2023; Yildirim et al., 2022). Existing assessments have used various indices derived from remote sensing data in combination with precipitation and temperature data to explore environmental anomalies and indicate disasters. These indices include Precipitation Anomaly Percentage (PAP), Temperate Condition Index (TCI), Palmer Drought Severity Index (PDSI), Normalized Temperature Anomaly Index (NTAI), Soil Water Deficit Index (SWDI), Vegetation Condition Index (VCI), Normalized Vegetation Anomaly Index (NVAI) and Normalized Difference Vegetation Index (NDVI). For instance, remote sensing-derived indices such as NDVI were used as an indicator of vegetation/crop health, exploring their relationships with environmental variables including soil moisture, precipitation, and temperature anomalies (Ghazaryan et al., 2020). The integration of remote sensing data with weather data allows for the exploration of climatic stress and implications on crop productivity, forest fire, and rangeland availability that enable early warning (Antonelli et al., 2023; Krishnamurthy et al., 2020; van Ginkel and Biradar, 2021).

The availability of open-access data has significantly enhanced the utilization of remote sensing technologies for Early Warning Systems. Initiatives such as the European Space Agency’s EO AFRICA initiative promotes the sustainable integration of Earth Observation (EO) and associated space technologies across Africa through its collaborative African-European research and development (R&D) partnership. It focuses on capacity building and offers customized, domain-specific training programs. Scientific collaboration and research capacity is co-designed and executed by African and European research teams. The “Famine Early Warning Systems Network” (FEWSNET) originally created by the United States Agency for International Development (USAID) during the severe famines of the 1980s in East and West Africa, serves as a key tool for early drought warning, providing timely data and analysis through remote sensing. It has been instrumental in identifying areas vulnerable to humanitarian crises, such as droughts, floods, and natural disasters in LMICs, a challenge often difficult to address through traditional ground surveys. Cross-disciplinary consortia and citizen scientists also enhance data-driven decision-making processes, enabling alert responses to global food security challenges. The NASA Harvest consortium connects NASA Earth observations to support a wide range of stakeholders, including leading researchers, humanitarian organizations, economists, policymakers, and other decision-makers. By providing actionable information, NASA Harvest facilitates preparedness and response to events such as food price volatility, weather-induced food shortages and other agricultural disruptions (NASA, 2024).

Our review presents a novel synthesis of satellite-based climate resilience measures in low-and middle-income countries, integrating advanced analytical techniques and artificial intelligence (AI) to enhance decision-making and policy implementation. Research and development partnerships, with facilitation of sustainable adoption of EO and related space technologies, could expedite the integration of small satellites utilizing standardized technologies, allowing populations with limited research budgets or technological capacity to access and benefit from space technology advances. The review identified key factors that could enable implementation and scale-up by deploying small, cost-effective satellites. Thus, facilitating greater participation of LMICs in EO missions as they can operate efficiently and affordably, enabling effective utilization of data and facilitating the development of application-driven missions and space system management. It is predicted that over the next decade, advancements in resolution and reductions in the cost per image will enable (small) satellite-based remote sensing to expand to previously unobtainable scales.

While many high-income countries worldwide continue to advance their infrastructure and capabilities for EO through satellite remote sensing, this progress creates both challenges and opportunities for collaboration in the adoption and application of these technologies tailored to low-and middle-income countries (LMICs). Over 40% of the world’s natural disaster events occur in Asia-Pacific, where international cooperation in EO is key to mitigating these disasters and ensuring “preparedness for future calamities” (Kaku, 2023). The China National Space Administration (CNSA) advances EO technologies for disaster risk reduction in China and beyond. CNSA’s satellite missions, like Gaofen, provide critical data for earthquake monitoring, landslide detection, and urban planning (Chen et al., 2022). The Indian Space Research Organization (ISRO) plays a vital role in EO for disaster management in India and neighboring South-East Asia regions. ISRO’s satellites provide critical data for cyclone tracking, drought monitoring, and urban planning to mitigate disaster risk (Khatri et al., 2022). The Japan Aerospace Exploration Agency (JAXA) utilizes satellite technology for disaster response and recovery efforts in Japan and the pacific. JAXA’s ALOS satellites contribute to land-use mapping, earthquake monitoring, and forest fire detection, supporting disaster resilience initiatives (Harvey, 2023). The Korea Aerospace Research Institute (KARI) develops EO capabilities for disaster risk reduction in South Korea. KARI’s satellites monitor typhoons, landslides, and forest fires, enhancing national resilience to natural hazards (Park et al., 2024).

The United Arab Emirates Space Agency (UAESA) and Mohammed Bin Rashid Space Center (MBRSC) advance EO technologies for disaster management in the Middle East region (Al Rashedi et al., 2020). UAE’s satellites provide critical data for desertification monitoring, coastal zone mapping, and disaster response (Al Hosani et al., 2021). The Egyptian Space Agency (EgSA) is one of the leading space players in Africa and is working towards greater independence in satellite manufacturing and space technology in the region. Its EgyptSat-3 (Planned for 2025–2026) will enhance Earth observation, agriculture monitoring and environmental tracking and aims to reduce reliance on foreign-built satellites.

South American agencies include the Comisión Nacional de Actividades Espaciales (CONAE), Argentina contributing to disaster risk reduction, leveraging satellite missions and EO applications like SAOCOM for flood monitoring and agricultural assessment. CONAE collaborates with regional partners to enhance disaster resilience in South America (Sahade, 1992). The Bolivarian Agency for Space Activities (ABAE), supports EO initiatives for disaster risk reduction in Venezuela, focusing on satellite applications for flood mapping and agriculture monitoring. ABAE contributes to regional collaborations to strengthen disaster preparedness. Instituto Nacional de Pesquisas Espaciais (INPE), Brazil INPE plays a critical role in EO for disaster monitoring in Brazil’s Amazon rainforest and coastal regions. INPE’s satellite data supports deforestation monitoring, fire detection, and biodiversity conservation efforts (Dantas, 2020).

Similarly, the European Union Space Programme components, specifically Galileo, EGNOS, Copernicus and GOVSATCOM, cover more than a third of the global data processing market and provides a diversity of specialist products and services offered by the geo-information sector. A notable milestone has been the “constellation of satellites (Sentinel family) and contributing missions (existing commercial and public satellites) providing datasets for Copernicus services” (ESA, 2023). These satellites are key to operational environmental information services, providing unprecedented volumes of data with assured continuity (Table 3). The launch of the “EarthCARE (Earth Clouds, Aerosols and Radiation Explorer) satellite a joint European/Japanese satellite will specifically address climate change and rising temperatures” (Eisinger et al., 2024). The European Organization for the Exploitation of Meteorological Satellites (EUMETSAT) maintains services for meteorological satellites in disaster risk management, e.g., severe weather monitoring, tropical cyclone tracking and climate studies. The German Aerospace Center (DLR) contributes to EO research and applications for disaster management, focusing on innovative satellite technologies. DLR’s TerraSAR-X and TanDEM-X missions support disaster response efforts through high-resolution radar imaging. In addition, France’s CNES plays pivotal roles in EO for disaster risk reduction, focusing on high-resolution imagery and radar satellite missions. CNES contributes to international collaborations, such as the International Charter on Space and Major Disasters, facilitating rapid access to satellite data during emergencies (Deniel and Tabary, 2023). SRON Netherlands Institute for Space Research (SRON) specializes in developing advanced technology in astrophysical research, Earth observation and design of satellite instruments for space research missions conducted by the European Space Agency (ESA), NASA and other space agencies. The UK Space Agency (UKSA) supports EO applications for disaster risk management through collaborations with European and international partners. UKSA-funded missions, such as Sentinel satellites, contribute to disaster response and environmental monitoring (Lappa et al., 2024). Private sector companies such as UK based Disaster Monitoring Constellation International Imaging Ltd. (DMCii) operate a constellation of EO satellites for disaster monitoring and response at the International Charter for Space and Major Disasters. Its rapid imaging capabilities enable timely assessments of floods, wildfires, and humanitarian crises worldwide. It also sells satellite imaging services and manages the operations of spacecraft such as UK-DMC1 and UK-DMC2.

The State Space Corporation (ROSCOSMOS) of Russia also contributes to global EO efforts through its satellite missions, providing data for disaster response and environmental monitoring. ROSCOSMOS collaborates with international partners to enhance disaster resilience (Kartamyshev et al., 2021) and methods for “forming multi-dimensional data in the information financial and economic system at the enterprise of state space corporation” (Gumenyuk, 2023).

In North America, the US National Oceanic and Atmospheric Administration (NOAA) is instrumental in disaster risk assessment, offering a wealth of environmental datasets through its satellite systems. NOAA’s geostationary and polar-orbiting satellites monitor weather patterns, hurricanes, and ocean conditions, aid in disaster forecasting and response. “Radiances and Environmental Data Records (EDR) from the Joint Polar Satellite System (JPSS) are critical for weather prediction. A variety of EDRs such as fire detection, retrievals of aerosols and water vapor are used by forecasters to provide warnings and alerts during disasters. Several JPSS products also provide continuity of essential climate variables such as upper atmospheric temperature and ozone concentration for climate change monitoring″ (Kalluri et al., 2021). The United States Geological Survey (USGS) is a key provider of EO data for disaster risk assessment, particularly through its Landsat programme. USGS satellites monitor land cover changes, volcanic activity, and coastal erosion, supporting disaster preparedness in the United States and globally (Teodoro and Duarte, 2022). In addition, the Canadian Space Agency (CSA) supports EO applications for disaster management through its RADARSAT program. RADARSAT satellites provide all-weather, day-and-night imaging capabilities for monitoring floods, landslides, and ice hazards in Canada and globally (Garneau, 2005).

Private sector Earth Observation (EO) suppliers are also pivotal in advancing global geospatial intelligence and remote sensing. Companies such as Maxar Technologies, Planet Labs, and Airbus Defence and Space are leading providers of satellite imagery and related data services. These leverage technological advancements in high-resolution imaging, and AI-driven data analysis to offer an array of products and services across various sectors, including agriculture, urban planning, environmental monitoring, and disaster management. The rise of private sector EO suppliers may present challenges, particularly in terms of data ownership, privacy, and regulatory oversight. Unlike publicly funded EO programs, private companies operate within a competitive, profit-driven environment, raising concerns about equitable access to data, especially in regions with limited economic investment or political constraints. Furthermore, use of EO data for commercial gain brings additional complexity to data security and ethical considerations, particularly in sensitive fields such as surveillance and national security. However, international collaboration can address shared global challenges while ensuring equitable access to EO technologies. This collaboration not only empowers LMICs to build resilience and improve decision-making but also fosters innovation and strengthens global partnerships in sustainable development. Ultimately, inclusive approaches to advancing EO capabilities have the potential to benefit all stakeholders, contributing to a more connected and informed world.

Satellite observations provide a wealth of data on a range of Essential Climate Variables (ECVs), which are critical for climate research, policymaking, and climate adaptation planning, particularly in LMICs that are most vulnerable to climate change (GCOS, 2016). ECV variables include land and sea surface temperature, precipitation, and CO₂ levels. Despite these challenges, there are immense opportunities for collaboration between high-income countries and LMICs to bridge these gaps. Knowledge-sharing initiatives, technology transfers, and partnerships focused on developing affordable, context-specific solutions can enable LMICs to harness the potential of EO technologies. International frameworks and programs, such as those spearheaded by the UN, play crucial roles in such collaborations.

Established standards, guidance, and good practices are vital for generating, sustaining and curating reliable global datasets to meet the needs of LMICs (Table 4). Monitoring standards for ECVs are often outdated, inadequate or absent. Addressing these issues will improve climate observations, predictions, and warnings, and ultimately climate adaptation planning in LMICs. The Global Climate Observing System (GCOS, 2016) has outlined standards and best practices for many ECVs to ensure data consistency, interoperability and reliability across different regions and time periods. Inconsistent data can lead to poor climate predictions and inadequate early warning systems (UNEP, 2023) and exacerbate LMICs’ vulnerability to climate-related hazards. For example, accurate precipitation data is essential for managing water resources and agricultural practices under rain-fed agriculture. Inconsistencies can result in unreliable hydrological models, poor water management decisions and increased vulnerability to droughts and floods. Similarly, inadequate or incorrect monitoring of temperature extremes can affect public health responses to heatwaves.

Capacity development for harnessing earth observation data from satellite remote sensing is essential, training of local scientists and technicians, establishment of necessary technical infrastructure and the creation of sustainable funding mechanisms are key to deliver impact. This involves both technology transfer in the use of EO technologies and developing capabilities for data interpretation and application (e.g., training on software tools, data processing techniques, and the integration of EO data with other data sources). Additionally, capacity development initiatives need to be tailored to the specific country need and contexts to ensure they are relevant and effective.

Improving the global consistency of EO observations is a critical goal. A concerted effort to harmonize data collection, data formats and reporting practices globally can only be achieved through international cooperation and establishment of centralized data repositories that maintain standardized protocols and curate data effectively. “Reliable data and effective disaster risk reduction policies are crucial to saving lives. Disaggregated data – showing the frequency, triggers and sequence of displacement – can help response and development planners mitigate impacts on displaced people and host communities” (WMO, 2024). Such repositories (ideally open access) would greatly facilitate data sharing, interoperability and integration, enhancing the overall quality and usability of global climate datasets. To address the risks associated with data discontinuity for LMICs, several strategies could be implemented at the national, regional, and international levels. Active collaboration is necessary between space agencies, research institutions and international organizations to advocate for the continuation of critical satellite missions and the development of follow-on missions (GEO, 2021). By engaging in partnerships and knowledge exchange, LMICs can better leverage existing expertise and resources to mitigate f data discontinuity impacts on climate monitoring. Investment in capacity building and technology transfer programs would enhance in-country ability to develop and operate satellite missions independently (UNESCO, 2019) and avoid vendor “lock in”. By building local technical expertise and infrastructure, LMICs can reduce their dependence on external sources of satellite data and ensure the continuity of climate observations over the long term.

For example, the European Commission is proposing an African-European partnership for digital transformation of the African continent. In this context, Earth Observation by satellite remote sensing is identified as essential digital technology providing information for evidence-based policy and decision making (ESA, 2024). The European Copernicus program and its Sentinel satellites offer open and free data of up to 1.2 petabytes per year across the African continent. The Sentinel-1 and -2 missions can now provide geospatial satellite remote sensing data that can strengthen climate resilience and disaster preparedness across African countries and subregions (ESA, 2024). Additionally, LMICs can combine EU data with alternative and complementary streams of climate data, such as ground-based monitoring networks, citizen science initiatives, and unmanned aerial vehicles (UAVs), to complement satellite observations and fill data gaps (Conrad and Hilchey, 2011).

Governments, NGOs, and companies can facilitate seamless data sharing by implementing mechanisms that are user-friendly, anonymous, and add value. In particular, the major cloud computing providers are well positioned to play a more significant role in reducing redundancy in imagery, streamlining processing costs, and reduce overall costs per data point (Lee et al., 2023). It can be anticipated that the cost of satellite imagery will continue to decrease to a level comparable with alternative imaging solutions.

Satellite-based measurements offer a cost-effective means of monitoring large and remote areas, providing critical data that may be difficult or very expensive to obtain through ground-based observations alone (Zhao et al., 2020). Integrating satellite data with ground-based measurements can also improve the reliability and comprehensiveness of datasets. By strengthening local capabilities, LMICs can contribute more effectively to global climate monitoring efforts and benefit from improved climate predictions and early warning systems to inform their Nationally Determine Contributions (NDCs) and National Adaptation Plans (UNEP, 2023).

The rapid advancement of analytical methods and AI technologies is revolutionizing the interpretation of satellite data, holding significant potential where access to timely and accurate data is crucial to address natural resource management and disaster risk reduction. International collaborations and partnerships can play a significant role in capacity development in EO and analytical techniques for LMICs. The combined resources of these agencies and technologies globally have unlimited capacity to deliver large-scale environmental monitoring services to terrestrial populations particularly in resource-constrained settings. These existing missions are all fully operational and highlight that the international satellite remote sensing community has the capacity to provide financial assistance and technical support to LMICs to strengthen their climate monitoring capabilities and resilience (UNFCCC, 2015). The Green Climate Fund, the Global Environment Facility and international funding mechanisms can allocate resources to LMICs for the procurement of satellite data, the development of climate information systems and implementation of capacity building initiatives (The World Bank Group, 2020). Likewise, international agreements and conventions can promote collaboration and knowledge sharing among countries to address common challenges related to climate monitoring. The Group on Earth Observations (GEO) and the African Space Policy and Strategy provide frameworks for cooperation, enabling knowledge transfer and resource sharing (GEO, 2021). It is critical to ensure the reliability and longevity of the satellites, developing appropriate data processing and analysis tools, and integrating satellite data with other sources of climate information (Choubey et al., 2020). International collaborations and partnerships can support the development and deployment of satellite cubes, leveraging shared expertise and resources.

Partnerships to develop CubeSats, or small satellites, presents a promising avenue for bolstering climate mitigation efforts in LMICs. These compact and cost-effective satellites offer the capacity to collect high-resolution data essential for monitoring various climate variables, thus supporting actions to mitigate and adapt to climate change. CubeSats hold particular significance for LMICs, providing a more accessible and affordable means of acquiring Earth Observation (EO) data and deliver crucial insights for climate mitigation strategies, monitoring deforestation patterns, tracking urban heat islands and coastal erosion dynamics, “advances in artificial intelligence and deep learning underpin all these technologies by enabling insights to be drawn from increasingly large data volumes” (Cavanaugh et al., 2024). The European Space Agency’s (ESA) new Φsat-2 mission, launched in August 2024, will push the boundaries of AI for Earth observation – demonstrating the transformative potential of AI for space technology. The Φsat-2 is a dedicated AI mission to explore the capabilities of utilizing extended onboard processing and further demonstrate the benefits of using AI for innovative EO. “With six AI applications running onboard, the satellite is designed to turn images into maps, detect clouds in the images, classify them and provide insight into cloud distribution, detect and classify vessels, compress images on board and reconstruct them in the ground reducing the download time, spot anomalies in marine ecosystems and detect wildfires” (ESA, 2024).

Building upon the foundation established by its predecessor, the African Regional Data Cube (ARDC), the Digital Earth Africa initiative secures support for a comprehensive, whole-of-government and multi-stakeholder approach to the utilization of Earth observation data. This collaboration is essential to promote the effective integration of these data into decision-making processes across various sectors and has helped countries such as Ghana, Kenya, Sierra Leone, Senegal, and Tanzania to address issues from agriculture and food security to deforestation and urbanization (Mubea et al., 2020). By facilitating access to EO data and fostering collaboration, initiatives like Digital Earth Africa can empower LMICs to monitor socioeconomic and biophysical conditions. Geographical coverage limitations, notably with geostationary satellites, do still exist, and stem from disparities in data availability across satellite agencies: collaboration is needed to address these challenges and ensure the reprocessing of data and their inclusion in climate data records.

In addition, assessing the accuracy and completeness of reliable model accuracy demands substantial investment in human resources, validation and ground truthing. This involves collecting field data to validate and calibrate satellite measurements, improving the accuracy and reliability of EO data. Manual data collection on the ground is indispensable for training analytics models and validating predictive imagery (Miller and Clark, 2020). Comprehensive ground-truthing data must encompass variables and observed events all geotagged to the observation location and compared against imagery from the same period (Robinson and White, 2023). However, the widespread acquisition of such data incurs significant costs.

Ground-truthing and validation are essential components of EO data implementation, ensuring that satellite observations accurately reflect on-the-ground conditions. In LMIC settings, ground-truthing is challenging due to financial, logistical and resource constraints. Ground-truthing efforts can be significantly impeded by remote and inaccessible areas, limited infrastructure, and scarcity of trained personnel (Smith et al., 2021; Johnson and Lee, 2023), challenges re particularly pronounced in LMICs, where logistical and resource constraints often limit the efficacy and accuracy of ground-based data validation (Garcia et al., 2022). Overcoming these challenges requires innovative approaches, including the use of citizen science and community-based monitoring. These reference-quality observations play a crucial role in monitoring climate system changes, bolstering confidence in assessments of future climate change and variability to facilitate the monitoring and quantification of the effectiveness of internationally agreed climate adaptation measures.

Enhancing accessibility and inclusivity in satellite remote sensing for the EO sector requires fostering equitable partnerships with local scientists and institutions in LMICs. This is essential to address data gaps and advance the sustainable development and utilization of satellite cubes by these countries (Table 5). These technical inputs could also be augmented through community-based approaches to make use of the resulting data, for example along the lines of the IDeAMapS approach proposed by Kuffer et al. (2020) with respect to LMIC urban areas. By establishing mutually beneficial partnerships to leverage local expertise, LMICs can build technical capacity and strengthen their ability to harness EO data for climate adaptation (Parkinson et al., 2016). Localisation is critical to ensure that satellite remote sensing contributes to local needs, and for long-term sustainability and impact in LMICs. Notably, few studies of community-based initiatives have been carried out for EO, in contrast to community-based and citizen-science approaches in biodiversity monitoring or hydrology (for example).

Advancements in machine learning, deep learning (DL) and big data analytics are transforming the landscape of satellite data interpretation, enabling more efficient, accurate, and scalable analysis. Machine learning algorithms, such as convolutional neural networks (CNNs) and recurrent neural networks (RNNs), can be used to extract complex patterns and features from satellite imagery (Ma et al., 2020). These algorithms can automate tasks such as land cover classification, vegetation monitoring, and disaster detection, significantly reducing the time and effort required for data analysis. Integration of satellite data with other sources of geospatial information, such as geographic information systems (GIS), crowd-sourced data, and social media feeds, has enabled the development of comprehensive analytical frameworks for understanding dynamic environmental processes and human-environment interactions (Qi and Hankey, 2021). These integrated approaches leverage the complementary strengths of different data sources to generate actionable insights holding immense potential to address key challenges faced by LMICs.

Disaster monitoring and response is a key arena where “timely and accurate information is critical for effective emergency management” (Sun, 2023). In this context, machine learning (ML) algorithms can synthesize satellite imagery to detect changes in land cover, identify affected areas, and assess the extent of damage following natural disasters such as floods, earthquakes, and wildfires (Feng et al., 2022). By providing real-time information to disaster response teams, satellite remote sensing technologies can improve the efficiency and effectiveness of relief efforts, saving lives and minimizing economic losses. Satellite remote sensing applications in agricultural monitoring and food security assessment are delivering on its potential, particularly in regions vulnerable to climate variability and extreme weather events. ML models trained on satellite data can analyze crop health, predict yields, and identify areas at risk of drought or pest infestation (Lobell et al., 2021). By providing EWS and decision support tools to farmers and policymakers, these technologies can enhance resilience to climate shocks and improve food security outcomes in LMICs.

Case studies evidencing the impact of this data on the prediction of future flood events are driven by the generation of extreme precipitation patterns. These patterns are produced by AI-driven weather generators, which are trained on historical satellite weather data and conditioned on climate change projections derived from global climate models. Techniques such as traditional Markov chain sampling (Steinschneider et al., 2015) or more advanced deep learning models, including variational autoencoders (Oliveira et al., 2022), have been shown to successfully generate synthetic weather data. This enables both the retrospective assessment of flood events and the forward-looking prediction of flood risks and damages, providing valuable insights into flood impact and future hazard preparedness (Annex A).

EO also supports natural resource management and biodiversity conservation. ML algorithms can use EO to detect deforestation, monitor changes in land use, and identify habitat loss hotspots, helping target interventions to mitigate environmental degradation. The widespread adoption of advanced analytical methods for satellite data interpretation may be constrained as “limited access to high-quality training data, inadequate technical infrastructure, and a shortage of skilled personnel are significant barriers to implementation” (Anderson et al., 2023). Data privacy, security, and equity challenges must be addressed to ensure that the benefits of these technologies are shared equitably across different socioeconomic groups and geographic regions. International collaborations, knowledge sharing initiatives, and capacity building programs can facilitate technology transfer and skill development, enabling LMICs to build local expertise and infrastructure for satellite data analysis (GEO, 2021). In theory, “public-private partnerships and innovative financing mechanisms can also support the development and deployment of tailored solutions that address the specific needs and priorities of LMICs” (UNEP, 2023).

Much of the literature recommends harnessing the potential of satellite remote sensing to deliver innovative climate adaptation solutions and a range of significant benefits for LMICs. As evidenced in the review, the potential to apply satellite remote sensing applications to remotely gather EO data and information can enable multiple applications across different sectors. Advanced ML techniques could identify the most “effective methods for distinguishing EQ intensities, enhancing risk evaluation, and minimizing incorrect hazard assessments. Additionally, integration with IoT systems ensures reliable and swift alerts to decision-makers, crucial for effective EEWSs and evacuation plans” (Abdalzaher et al., 2024). Technological progression swiftly enhances the capacity of the satellite sector to monitor, measure, and disseminate information pertaining to various adverse climate events. The review found that 11.67% of the studies demonstrated improvements in disaster response using satellite data, whilst 37.22% highlighted gaps in data accessibility in LMICs. It also plays pivotal roles in forecasting, validating, and alleviating natural disaster impacts, “leveraging their distinctive capacity for remote monitoring of environmental conditions pre and post occurrences like droughts, wildfires, and floods. This capability holds significant utility for insurers and has the potential to facilitate parametric insurance schemes for the agriculture sector” (Born et al., 2019). Parametric insurance schemes, being detached from the underlying asset, can be linked to specific triggering events such as extreme weather conditions, averting the necessity for physical assessments.

The consideration of space science trends, market dynamics, policy landscapes, strategic frameworks, and regulatory frameworks, along with associated social, economic, and political challenges present detailed insights from emerging space-capable countries within these regions is crucial for establishing and enhancing efficient and mutually beneficial cooperation mechanisms within the realm of space technology to adapt to climate change shocks and disasters. Recent innovations bridging AI with EO infrastructure offer a range of platforms to access and process data at scale. Despite these trends, the current landscape of EO initiatives remains disjointed and is characterized by varying degrees of user engagement and impacts (Di Leo et al., 2023). Advances in AI applied to EO data platforms are facilitating a new generation of AI-based climate forecasting solutions to improve precision. However, access to computing power, large EO datasets, technology transfer resources and expertise all pose challenges for LMICs. Despite this, many LMICs are adopting impactful climate resilience strategies based on satellite data (Annex C). Examples include Pakistan, which has expanded EWS, reduced flood-related fatalities by over 90%, and built flood-resilient infrastructure, helping mitigate $30 billion of losses. Kenya promotes climate-smart agriculture, increasing crop yields by 30% while supporting 1 million farmers with drought-resistant practices and financial support. Bangladesh has pioneered floating schools and homes, providing education to thousands of children and strengthening coastal defenses to protect from rising sea levels and floods.

We have identified various gaps in research on this topic. While many of the studies presented satellite remote sensing projects and provided an immediate “snapshots ‘of project implementation, there was a lack of evaluation of the long-term impact of satellite remote sensing application in EO. Effective implementation of satellite remote sensing initiatives hinges on delivery of impact, which is achieved when driven by local or national needs, while adhering to monitoring standards, guidance, and best practices. International organizations play a pivotal role in establishing standardized protocols for data acquisition, processing and dissemination. By promoting data interoperability and harmonization, these initiatives facilitate cross-border collaboration and information sharing, thereby enhancing the scalability and replicability of satellite-based monitoring systems. Moreover, capacity-building efforts aimed at training local stakeholders in EO data analysis and interpretation will foster sustainable institutional capacities, ensuring the long-term viability of satellite remote sensing initiatives in LMICs.