The spatial resolution of DEMs significantly influences the precision and accuracy of calculated morphometric parameters. Kervyn et al. (2008) compared SRTM to ASTER DEM, analysing the variability of resolution on morphometric parameters, using both modelled theoretical cones and global scoria cone fields. They found, for a regularly shaped cone on a flat surface, that error in height increases from 1 m using 10 m spatial resolution, to between a 4%-8% underestimate using 90 m spatial resolution. Width similarly increases in error from 10 m, to a 4.1% overestimate, and a 12.8% overestimate, using 10 m, 30m, and 90 m spatial resolutions respectively. These errors are further amplified with the introduction of a 7° underlying slope. Height increases in error from <5 m to an 8.2% underestimate and up to a 40% underestimate using 10 m, 30 m, and 90 m spatial resolutions respectively.

The study emphasized that over- and underestimation correlates directly with DEM resolution, slope steepness, and the sharpness of topographic breaks. This was tested on a sample of 40 Mauna Kea scoria cones for cone height, revealing that 90 m SRTM DEM products are inadequate for analysing smaller-scale features (<100 m in height), with identification challenges arising when DEM resolution exceeds one-third of the feature size (Table 1).

Subsequent investigations by Fornaciai et al. (2012) expanded these analyses by comparing 2 m LiDAR data with 10 m TINITALY DEM, 10 m USGS National Elevation Dataset (NED), and 30-m ASTER DEMs. Their results demonstrated that ASTER DEM exhibits a Root Mean Square Error (RMSE) for height of 21.1 m, translating to approximately 21% average error for 100 m high scoria cones. This analysis established that 30 m ASTER DEM should only be employed for cones with volumes exceeding 30 × 10⁶ m³, where relative error decreases below 20%.

Zarazúa-Carbajal and De la Cruz-Reyna (2021) further investigated DEM resolution impacts on morphological analysis, specifically targeting their Average Erosion Index (AEI) model for scoria cone degradation chronology. Their research confirmed that smaller-volume cones (<0.01 km³) cannot be accurately analysed using 12 m resolution DEMs, particularly for edifices situated on inclined terrain where vertical precision is most critically affected.

Zhang et al. (2022) assessed free global DEM accuracy (SRTM 30 m; ALOS AW3D30 30 m) across multiple volcanic fields, using a 12 m spatial resolution DEM as reference. Their analysis revealed average volume errors of 4.5%–7.4% for the SRTM DEM products, with cones smaller than 5 × 10⁶ m³ exhibiting volume errors ranging from 5.4% to 20.5%. The AW3D30 DEM demonstrated improved performance with average errors of 2.8%–4.5% for volume, 3.1%–8.3% for height, and 2.5%–6.2% for slope angle measurements. Additionally, pre-eruption surface flatness significantly affects accuracy, with SRTM-based pre-eruptive surface modelling yielding volume errors of 16.3% using average height methods or 30.6% using Triangulated Irregular Network (TIN) interpolation approaches. However, Zhang et al. (2022) did not consider the TanDEM-X 30 m global DEM, which can be used for morphometric studies. Van Wees et al. (2024) compared the TanDEM-X 30 m DEM to ALOS, ASTER, and SRTM 30 m DEMs for 16 stratovolcanoes, with relative standard deviations (RSD) between 0.13% and 2.03% for morphometric parameters. However, it is uncertain how the errors scale to smaller volume scoria cones.

Despite potential accuracy limitations for resolutions exceeding 30 m, diverse DEM products continue to be employed in cone morphology studies, Figure 3 (Supplementary Data 1).

The freely available, near-global AW3D30 DEM proves suitable for basic morphometric analysis of larger scoria cones (>100 m height) where high precision is not required, such as cone shape or size categorisations or analysing elongation for tectonic-based studies. However, detailed investigations on a cone-by-cone analysis, such as volcanic processes or age inference from morphometric parameters, necessitate DEMs with spatial resolutions better than 30 m to minimise error propagation and preserve critical topographic details such as crater rims, crater depths and abrupt slope transitions, such as the 12 m TanDEM-X DEM.

A critical limitation in existing accuracy assessments concerns the lack of quantification of methodological errors in width, height, and boundary delineation calculations. The comparative impact of these methodological differences on overall accuracy remains poorly constrained. Furthermore, studies have yet to adequately quantify external factors influencing DEM error, including cone morphological irregularity, pre-eruptive topography, and vegetation cover effects. Extreme differences between DEM products may partially result from dense vegetation coverage affecting surface detection capabilities. It is worth noting that the DEM vertical resolution, data source, and the possible artifacts should be considered during the selection process of DEMs.

The manual approach represents the most widely employed method for identifying scoria cone boundaries, involving analysis of slope breaks and topographic discontinuities that distinguish volcanic edifices from surrounding terrain (e.g. Scott and Trask, 1971; Favalli et al., 2009; Haag et al., 2019). This user-dependent method analyses slope breaks using:

• Linear topographic profiles extracted from DEMs or topographic maps

Early morphometric studies frequently included debris aprons within cone boundaries (Hasenaka and Carmichael, 1985; Sucipta et al., 2006), however Grosse et al. (2012) suggested that delimitation should not include debris aprons due to a lack of clear morphometric signature. To reduce subjectivity, researchers have implemented:

• Field validation (Sucipta et al., 2006; Inbar et al., 2011; Kereszturi et al., 2013a)

Van Wees et al. (2024) conducted a comprehensive assessment of delimitation uncertainty by having seven volcano geomorphology experts manually outline 16 composite volcanic edifices using 30-m SRTM DEMs. Initial delineations were performed using only topographic data, followed by a second round 6 months later incorporating slope thresholds ranging from 1° to 6°. Results demonstrated that a 3° slope threshold achieved >50% inter-analyst consensus, significantly improving boundary consistency and reducing subjective variation. It is key that such studies are expanded to include smaller-volume scoria cones across various volcanic settings, where the impact of DEM resolution may pose as an additional challenge to identifying cone boundaries.

The manual topographic method for base delimitation remains highly subjective and is largely dependent on the resolution of the DEM, satellite imagery accuracy, surrounding vegetation, surrounding topography and the irregularity of the scoria cones morphology from flank collapse or lava flow burial. It can become unclear where the edifice of a scoria cone starts and ends, and deviations in boundaries can cause outlier values of morphometric parameters. The extent to which the user-defined identification of scoria cone bases influences derived morphometric parameters remains unresolved in the literature, particularly for cones whose bases are difficult to delineate within complex surrounding topography. However, it is safe to assume that care has been taken during base delimitation, and future studies should continue to apply multiple methods (e.g. field confirmation, satellite imagery, orthophotos, slope thresholds) to identify the base of a scoria cone, particularly when using courser DEM resolutions. Furthermore, following Van Wees et al. (2024) using a slope threshold of 3° can support delimitation and progress to an agreed consensus on methodology.

Recognition of the subjectivity inherent in manual boundary identification has driven development of computational approaches designed to improve objectivity and consistency while reducing time requirements for large-scale morphometric studies.

Early studies calculated the scoria cone width as the maximum diameter of the cone based on a simple, circular based cone (Scott and Trask, 1971; Wood, 1980a; b). Settle (1979) defined cone width (W_co) as the average of the maximum (Wmax) and minimum (Wmin) base diameters, Equation 1 and Figure 4. Clear descriptions regarding how the maximum and minimum diameters are measured is limited, particularly for an irregular shape. Some studies generate a best-fit ellipse of a cone outline, measuring the diameters as the perpendicular maximum and minimum diameters (e.g. Kereszturi et al., 2013a; Becerra-Ramírez et al., 2022). It is likely that many studies applied this method but did not provide adequate descriptions. It is worth noting that measurements of crater diameters are calculated in a similar way to the diameter of the cone, however, are more sensitive to the effects of DEM resolution and crater outlines due to the smaller size. W c o = W m a x + W m i n 2 (1)

This methodology, representing the arithmetic mean of the major and minor axes of the cone base, continues to be employed in contemporary studies, demonstrating its enduring utility for morphometric characterisations (e.g. Pedrazzi et al., 2020; Vörös et al., 2022; Sieron et al., 2023; Azizah, 2025).

Bemis et al. (2011) and Bemis and Farencz (2017) adapted the traditional diameter approach to accommodate irregularly shaped cones by incorporating multiple elevation profiles across the cone structure, Equation 2: W c o = W c o 1 + W c o 2 + W c o 3 + W c o 4 … . n (2) where Wco1, 2, 3, 4… Represent the widths measured along individual profiles passing through the centre of the cone and n represents the total number of profiles analysed. This approach provides enhanced characterisation of cone dimensions by sampling multiple orientations, thereby accounting for morphological irregularities that may not be captured by simple diameter measurements.

Zarazúa-Carbajal and De la Cruz-Reyna (2021) implemented a similar multi-profile methodology, calculating width across four cone and four crater profiles using the distance formula, Equation 3: W c o = x 2 − x 1 2 + y 2 − y 1 2 (3) where x₁, y₁ and x₂, y₂ represent coordinates at which slope breaks occur along linear profiles. This geometric approach enables precise measurement of cone dimensions, while accounting for terrain inclination effects.

Favalli et al. (2009) introduced an area-based methodology that defines width from the basal area of the cone, addressing limitations of diameter-based methods when applied to irregular cone shapes, Equation 4: W c o = 4 A R E A / π (4)

This approach, which has been extensively implemented in DEM-based analyses (Grosse et al., 2012; Kervyn et al., 2012; Euillades et al., 2013; Kereszturi et al., 2013a; Grosse et al., 2020; Hunt et al., 2020; Zhang et al., 2023), reduces subjectivity in the characterisation of irregular cones by estimating width from the total basal area. The method assumes circular area equivalence, providing a standardised approach to width calculation that remains independent of cone orientation and shape complexity.

Although the area-based method reduces measurement subjectivity for irregular cones, it remains dependent on the accuracy of base delimitation procedures. The precision of this approach is fundamentally constrained by the quality of cone boundary identification, emphasising the importance of robust boundary delineation protocols.

Kereszturi et al. (2013a) used both traditional and area-based methods for widths when calculating the morphometric parameters for 61 scoria cones on Tenerife, Canary Islands. Their study revealed that the Settle (1979) method, overestimates width by an average of 1.5% compared to the Favalli et al. (2009) planimetric area method, with only 10 cones registering greater than 3% difference between methods. These findings suggest that the traditional Settle (1979) method remains acceptable for width calculation, particularly for approximately circular cone shapes.

The impact of base delimitation accuracy on width measurements, calculated using area-derived methodologies, was demonstrated by Euillades et al. (2013), who documented width differences up to 33% (average difference of 6%) between manually drawn and algorithm-derived cone outlines. This substantial variation emphasizes the critical importance of accurate cone boundary identification when deriving width measurements, particularly when employing area-based methodologies.

DEM resolution effects on width calculations were quantified by Fornaciai et al. (2012), who calculated width errors ranging from 3% to 16% for 30 m resolution ASTER DEMs across multiple volcanic fields when compared to 10 m resolution datasets. These resolution-dependent errors are likely due to the differences in determined cone basal outlines and highlight the importance of appropriate DEM selection for accurate width characterisation.

Di Traglia et al. (2014) attempted to combine both the methods of Settle (1979) and Favalli et al. (2009) to calculate a ‘mean’ width, alongside calculating width related to the circumference of the cone, Equation 5, although neither method has been used by any subsequent study: W c o = c r f π (5) where crf is the diameter of a circumference equivalent to the perimeter of the cone.

So long as the outline of the cone base is correctly and accurately identified, and an appropriate DEM or topographic map is chosen, the method chosen to calculate width is unlikely to cause significant errors. However, future studies should attempt to properly quantify the differences between each method and the impact they may have on height to width ratios, similar to Favalli et al. (2009). However, a key aspect to consider when comparing recently estimated widths to those from older studies is that they often included the debris apron of volcanic material in calculated widths (Hooper, 1995; Hooper and Sheridan, 1998; Sucipta et al., 2006). It was recommended by Grosse et al. (2012) that far reaching debris aprons should not be included in measurements.

Early morphometric studies employed relatively straightforward measurement techniques constrained by available data sources. The classical approach, first systematized by Settle (1979), defined cone height (H_co) as the elevation difference between the summit of the scoria cone and its base, where the basal elevation was calculated as the mean of the highest and lowest basal values (Equation 6; Figure 5) H c o = 1 2 A c o − A B M + A c o − A B m (6)

A_co represents the maximum altitude of the cone, A_BM the maximum altitude of the base, and A_Bm the minimum altitude of the base; this has been extensively adopted throughout the literature (Hooper, 1995; Hooper and Sheridan, 1998; Riedel et al., 2003; Aguirre-Diaz et al., 2006; Sucipta et al., 2006; Sutawidjaja and Sukhyar, 2009; Gilichinsky et al., 2010; Inbar et al., 2011; Dóniz-Páez et al., 2012; Kereszturi and Németh, 2012b; Pedrazzi et al., 2020; Vörös et al., 2021; Becerra-Ramírez et al., 2022; Vörös et al., 2022).

This classical method has been implemented with various modifications across numerous studies. Kervyn et al. (2008), Kervyn et al. (2012) measures height by subtracting the average cone base elevation from the average crater rim elevation (Equation 7). H c o = H a v g − A a v g (7)

H_avg represents the average height of the crater rim and A_avg the average altitude of the base. Alternative formulations have defined height simply as the difference between the lowest altitude of the base of the cone and its summit (Equation 8) (Rodriguez et al., 2010; Guilbaud et al., 2012; Haag et al., 2019). H c o = A c o − A B m (8)

Some researchers have adopted profile-based methodologies that sample multiple elevation profiles across individual cones to obtain representative height measurements. Zarazúa-Carbajal and De la Cruz-Reyna (2021) developed approaches that analyse elevation profiles from eight different directions, four crossing the centre of the crater and four crossing the centre of the cone base, with profiles generated at 45-degree azimuthal separations. The resulting cone height is the average height along each profile, Equation 9. This method is similar to the four-profile method used by Bemis et al. (2011) and was implemented by Sieron et al. (2023). This methodology allows for correction of terrain inclination effects and provides more comprehensive characterisations of cone dimensions, particularly important for breached or irregularly shaped cones. H c o = H p 1 + H p 2 + H p 3 + H p n / n (9)

Where Hp1 is the height of the cone along profile 1 and n is the number of profiles.

The recognition that formula-based methods introduce significant errors when applied to cones situated on steep underlying slopes (>5°) led to the development of DEM-based interpolation techniques.

Favalli et al. (2009) demonstrated that traditional formula methods average approximately 22% error on dipping basal planes, prompting the development of three-dimensional basal plane interpolation methods.

The interpolation-based approach calculates maximum height as Equation 10: H m a x = Δ z max (10) where Δz_max is the maximum elevation difference between the crater rim and the pre-eruption surface.

Mean height is calculated as the mean elevation of the 3D crater rim above the 3D base surface, Figure 6.

Contemporary studies increasingly employ diverse interpolation algorithms to establish pre-eruptive basal surfaces, including natural neighbour, inverse distance weighting, and kriging techniques (Fornaciai et al., 2010; Fornaciai et al., 2012; Grosse et al., 2012; Cimarelli et al., 2013; Euillades et al., 2013; Kereszturi et al., 2013a; Di Traglia et al., 2014; Mukhopadhyay et al., 2019; Hunt et al., 2020; Grosse et al., 2020; Aguilera et al., 2022; Zhang et al., 2022; 2023; Kereszturi et al., 2025). The Triangulated Irregular Network (TIN) interpolation method has been particularly widely adopted, though it can introduce significant errors, in the case of volume calculations showing discrepancies of up to 30.6% when compared to average height methods (Zhang et al., 2023). Ideally, different methods should be tested with the error calculated for each interpolation technique to establish which method provides greater accuracy.

Few studies have compared the results of using both methods to calculate height. Favalli et al. (2009) found an average 64% difference between the Settle (1979) method and mean height using interpolation and a 27% difference when compared to maximum height (Equation 10). Hopfenblatt et al. (2021) compared the Settle (1979) and Favalli et al. (2009) methods for Stanley Patch Volcano, Antarctica, identifying a 3.8% difference, likely attributable to the shallow 3° basal plane inclination. This exemplifies the impact steep base inclinations can have on height measurement variability.

The morphometric parameters derived from different height measurement methods can vary significantly, with implications potentially cascading through subsequent analyses including volume calculations, slope angle determinations, height/width ratios, and age estimates based on morphometric degradation models.

Despite the multiple ways crater depth has been measured, no study has yet to quantify the differences between methods, therefore it is uncertain which method yields the most accurate results. As expected, a small feature such as the crater will be highly dependent on the resolution of the DEM used.

A formula-based method includes calculating volume from the height, width, crater width, and crater depth parameters. Hasenaka and Carmichael, 1985 calculated the volume of a scoria cone as a symmetrical truncated cone, Equation 11: π H c o 12 × W c r 2 + W c r × W c o + W c o 2 (11) where W_cr represents crater width, W_co represents cone width, and H_co represents cone height. This approach assumes complete crater infilling, which may not accurately represent young cones with open craters.

Riedel et al. (2003) therefore defined volume by subtracting the volumes of the inverse crater cone from the volume of the whole-cone, Equation 12: π H co 3 ⁢ W co − W cr 2 H co + W cr 2 2 H co 2 + 3 W cr W co − W cr 4 H co 2 or 1 2   ⁢ π ⁡ W co 3 H co W co − W cr − W cr 3 ⁢ H co W co − W cr + D cr W cr (12)

Subsequent analysis led to simplified height-dependent volume relationships, Equations 13, 14; (Riedel et al., 2003; Kervyn et al., 2012): 11.5 H c o 3 (13) 11.31 H c o 3 (14)

These empirical relationships suggest that cone volume scales with the cube of height, providing simplified estimation methods for morphometric studies.

For breached or open cones, Dóniz-Páez et al. (2012) applied volume corrections by reducing calculated volumes by 50% when structural collapse was evident. For cones lacking distinct craters, volume was calculated using oblique cone geometry, Equation 15; (Becerra-Ramírez et al., 2022): 1 3 π H c o × R c o 2 (15)

Where R_co is the radius of the cone base or W_co/2.

The advent of DEMs enabled direct volume calculations through surface interpolation and integration techniques. Carmichael et al. (2006) calculated the volume of a scoria cone by the difference of the surface topography and an interpolated base determined by the surrounding topography for scoria cones in Colima, Mexico. Favalli et al. (2009) formalised this approach as the volume enclosed between DEM surfaces and three-dimensional basal surfaces derived from Delaunay triangulation (also known as Triangulated Irregular Network (TIN)) of cone base coordinates. Studies have also calculated volume from the present-day surface (Inbar et al., 2011; Fornaciai et al., 2012), inverse distance weighting (IDW) (Grosse et al., 2014), or continuous curvature splines (Hunt et al., 2020). The impact of the various interpolation techniques on cone volume has yet to be examined, however it is likely that the choice of interpolation method will have a significant impact on the resulting volume.

The pre-eruptive surface can also be approximated from modification of contour lines, slope angles of the surroundings, or interpolations of surrounding elevations, where volume can be calculated for every grid feature of a DEM, Equation 16 (Kereszturi et al., 2013b). ∑ Δ Z i x   y (16)

Where ΔZ_i is the height difference between the DEM and basal surface, and x, y represent pixel dimensions.

Rodriguez-Gonzalez et al. (2010), Rodriguez-Gonzalez et al. (2011) developed comprehensive geomorphological reconstruction techniques for volcanic units, incorporating field investigations to develop pre-eruptive, post-eruptive, and current Digital Terrain Models (DTMs). Total original volume (V_O) was calculated from differences between post-eruption and pre-eruption DTMs, while the actual volume (V_R) represented differences between present-day and pre-eruption DTMs. Eroded volume (V_D) was expressed as Equation 17: V D = V O − V R (17)

DEM-based volume calculations are fundamentally dependent on base delimitation accuracy, interpolation methods, and DEM resolution. Fornaciai et al. (2012) demonstrated that the ASTER 30 m DEM results in volume errors decreasing from approximately 60% for volumes ∼10 × 10⁶ m³ to <30% for volumes ∼30 × 10⁶ m³, while TINITALY 10 m DEM errors decrease from 40% to 10% for similar volume ranges. Zhang et al. (2022) documented average volume errors of 2.8%–4.5% for cones <5 × 10⁶ m³, emphasizing the importance of edifice size considerations in comparative analyses. Furthermore, the method does not consider positive or negative topography beneath the edifice (Grosse et al., 2012). However, extensive fieldwork to reconstruct the pre-eruptive terrain could improve this, such as the method of Rodriguez-Gonzalez et al. (2010), Rodriguez-Gonzalez et al. (2011). Overlapping edifice complications require subjective methodological decisions, with Grosse et al. (2012) employing enlarged outlines to encompass multiple edifices, while Hunt et al. (2020) applied formula-based methods for individual overlapping cones. It is subjective as to what method should be used when cones are overlapping, however, the formula-based method based on an ellipsoidal shape for each of the cones can be calculated, where the volume of each overlapping cone can be subtracted from the bottom cone to produce a volume estimate.

Formula-based volume methods struggle to capture morphological diversity and obtain precise volumetric measurements. Volume calculations are subject to 'scaling' issues where successive errors in height, width, and crater measurements propagate to produce substantial over- or underestimates (Bemis and Farencz, 2017). Sieron et al. (2023) documented anomalous volumes for approximately 50% of scoria cones in the Los Tuxtlas Volcanic Field, Mexico, likely due to dense vegetation coverage affecting LiDAR data corrections, necessitating reversion to formula-based methodologies.

O’Hara et al. (2020) compared volumes of composite volcanoes between their work with that of Hildreth (2007) and Grosse et al. (2014), reporting a mean absolute difference of 183% and 342% respectively, reduced to 32.1% and 92.3% respectively when outliers are removed. These differences are likely caused by a combination of differences in basal outlines and methods to obtain volume, with O’Hara et al. (2020) opting for the MBOA method compared to MORVOLC of Grosse et al. (2014). These substantial discrepancies highlight the need for systematic methodological comparisons and standardisation protocols, extending the study to consider scoria cones across different volcanic and environmental settings, and attempting to identify the main cause of the discrepancies, something we have tried to address in this review.

Comprehensive comparative studies between formula-based and DEM-based volume calculation methods remain limited, representing a critical research gap in morphometric methodology. The identification of primary sources of discrepancies between methodological approaches requires systematic investigation across diverse volcanic settings and cone morphologies. Additionally, investigation of environmental factors affecting measurement accuracy, including vegetation effects, surface roughness influences, and terrain complexity impacts, would enhance understanding of error sources and improve interpretation of morphometric analyses.

Throughout the literature, slope calculation methodologies have evolved significantly, paralleling developments in height measurement techniques with the advent of high-resolution DEMs. Hasenaka and Carmichael, 1985 developed a method to obtain average flank slope angles through trigonometric modelling of the cone’s basal widths, crater widths, and height, Equation 18, which has been widely used (e.g. Hooper and Sheridan et al., 1998; Riedel et al., 2003; Aguirre-Diaz et al., 2006; Sucipta et al., 2006; Guilbaud et al., 2012). S l o p e = ⁡ tan − 1 2 H c o W c o − W c r (18)

And simplified to Equation 19 for cones without a crater. S l o p e = ⁡ tan − 1 2 H c o W c o (19)

This trigonometric method continues to be implemented in contemporary studies due to its computational simplicity and applicability to topographic map-based measurements (e.g. Bemis and Farencz, 2017; Jaimes-Viera et al., 2018; Haag et al., 2019; Benamrane et al., 2022; Sieron et al., 2023).

A similar method can be used to calculate the inner crater slope assuming a vent or conduit width, Equation 20 (Kervyn et al., 2012; Bemis and Ferencz, 2017). Inner Slope = ⁡ tan − 1 2 D cr W c r − W v (20)

Where W_v is width of the vent, often assumed to be 0 m (Bemis and Ferencz, 2017).

Parrot (2007) advocated for the utilisation of high-resolution DEMs to enable automated parameterisation of volcanic cones, including direct slope calculations from elevation data. This approach represents a significant advancement over formula-based methods, as mean dipping angles can be calculated directly from DEM surface derivatives. Contemporary software implementations, such as ENVI 4.6 topographic modelling procedures and ArcGIS/QGIS Spatial Analyst tools, provide standardised approaches for slope calculations. This approach is frequently applied in contemporary literature (e.g. Pedersen et al., 2020; Uslular et al., 2021; O’Hara and Karlstrom, 2023; Pedrazzi et al., 2024). Average and median slope angles can be derived from a DEM on the flanks and within the inner crater, including at different height intervals within the crater (Grosse et al., 2012).

The accuracy of DEM-based slope calculations is directly proportional to DEM resolution/type and base/crater delimitation precision. Coarser resolution DEMs systematically smooth steep slope angles, with Root Mean Square Error values more than doubling for slope angles exceeding 10° compared to gentler slopes (Kervyn et al., 2006; Gilichinsky et al., 2010; Zhang et al., 2022). The relationship between DEM resolution and slope measurement accuracy has been extensively documented, with Gilichinsky et al. (2010) demonstrating that scoria cone YZN in Tolbachik, Kamchatka, was underestimated by 9.8° when using the SRTM 90 m DEM compared to a 5 m contour digitised map-based DEM. The base and crater delimitations are crucial as they determine the slope values that are included within the slope histogram; inclusion of a flat-lying base or crater rims may skew average slope angles or generate high standard deviations (Kereszturi and Nemeth, 2012a). As suggested for height measurements, lower edifice sizes also lead to higher errors in slope angles (Bemis et al., 2011; Zhang et al., 2022).

Due to the potential formation of complex internal architectures of scoria cones, significantly reshaping the morphology, the slope angles of scoria cones can be misinterpreted and more complicated than generally assumed (Kereszturi et al., 2012). Therefore, Kereszturi et al. (2012) developed a method that splits the outer flanks of scoria cones into three types, ‘uphill’, ‘downhill’, and ‘other’, allowing for a more robust estimate of flank slope angles in the presence of complex cone architecture and steep underlying surfaces, with differences in slope up to 12° on a flat basal slope and 30° on steep basal slopes.

A similar approach was taken by Vörös et al. (2021) who implemented a ‘sectorisation’ of scoria cones to reflect asymmetry. A scoria cone is split into sectors of ∼15° (depending on cone size), omitting the crater, resulting in ∼24 ‘cut outs’ of the cone, each with their own calculation of average slope angle. This methodology enables quantification of cone asymmetry and accounts for directional variations in slope characteristics that may result from wind effects during the eruption or preferential erosional processes.

Inbar et al. (2011) compared slope angles calculated using the 30 m ASTER DEM, where slope angles represented averages of pixel slope values situated along the steepest profile with greatest elevation difference, to map-based methods using spacing between contours. The largest discrepancy of slope angles between the two methods, for cones in Tolbachik, Kamchatka, was an overestimation of the map-based method by 3.2° (10.5% difference). The study noted uncertainty regarding whether variations resulted from DEM resolution differences compared to topographic maps, or from methodological differences in slope angle calculation procedures (through either error in the formula or subjectivity of the analyst), and it was recommended that only a single source of elevation data should be used in future studies (Inbar et al., 2011).

Kereszturi and Nemeth (2012a) calculated slope angles using both manual and DEM-based methods employing identical input data from a 1:10,000 topographic map with 5 m contour intervals (rasterised for the DEM-based methods using linear interpolation). Formula-based slope angles were calculated using trigonometric relationships, while average, median, mode, and maximum slope angles were directly derived from pixels within delimited areas (not the method of Kereszturi et al., 2012). The largest difference in mean slope angle between the methods was 9.5° (132% difference), likely attributable to cone morphological complexity, with formula-based methods consistently underestimating average slope angles for each measured cone.

The method of Vörös et al. (2021) documented similar results with strong overestimations of slope angle (exceeding 10° in some cases) using the formula-based methods compared to DEM-derived sectorization approaches. These studies collectively emphasise the limitations of formula-based methods and the potential inaccuracies they introduce when interpreting scoria cone morphology, particularly for morphometric-based dating applications. It is worth noting that in Vörös et al. (2021) formula-based methods overestimated slope angle compared to DEM-based methods, however Kereszturi and Nemeth (2012a) found underestimations of the formula-based method. The discrepancies between the two findings outline the complexities in measuring flank slope angles.

The accuracy of slope measurements can be influenced by various environmental factors beyond DEM resolution, including vegetation cover effects and surface roughness variations. Dense vegetation can affect DEM surface detection capabilities, potentially introducing systematic errors that vary between different slope calculation methodologies. These effects remain poorly quantified in existing literature but may contribute significantly to measurement uncertainties in heavily vegetated volcanic fields. Furthermore, the impact of boundary delineation also remains a present challenge in slope calculations, with Van Wees et al. (2024) finding an RSD of 6.12% between NETVOLC and manually drawn boundaries for stratovolcanoes when calculating average slope using the DEM-based method.

Here, we attempt to analyse some of the variations that can exist between different studies that use contrasting methods for morphometric analysis, showing the variations in results that can appear. It is uncertain which results are closest to the ‘actual’ morphology of a scoria cone, however this outlines the challenge that can exist interpreting morphometric data.

Due to the complexities of scoria cone morphology, which are largely dependent on the context of the study, it may not yet be appropriate to suggest a complete standardised protocol. Instead, a series of recommendations can be made to improve accuracy and comparability going forward.

DEM Selection and Use:

• AW3D30 DEM offers the best overall trade-off between accuracy, coverage, and accessibility for global comparative studies, particularly at large regional scales. The TanDEM-X DEMs are also viable options for such analysis

Scoria Cone Boundary Delimitation:

• Hybrid approaches (automated detection + manual refinement) yield the most robust results.

Estimation of Pre-eruptive Surface, Cone Heights, and Cone Volumes:

• Favalli et al. (2009) interpolation-based method is universally recommended for future studies, due to its robust error assessment linked to DEM vertical accuracy. Formula-based methods can result in errors exceeding 20%, particularly for irregular or complex scoria cones

Slope Analysis:

• Prefer DEM-based slope calculations over formula-based methods, especially for detailed morphological studies

The volcanological community requires coordinated efforts to develop standardised delimitation protocols and comprehensive methodological frameworks. Priority should be given to large-scale comparative studies that systematically evaluate different measurement approaches using identical high-resolution datasets (preferably <10 m resolution DEMs) across diverse volcanic settings. Such studies should compare morphometric parameters under varying conditions including DEM resolutions, manual versus automatic base delimitations, and formula-based versus DEM-based methods, while grouping cones by shape, underlying slope angle, age, and composition, expanding on the initial results presented within this study. Research should extend to further volcanic regions, such as the Ethiopian Rift Valley or Indonesian volcanic fields, where research appears to be limited yet represent diverse tectonic, environmental, and volcanic settings.

Critical research gaps include systematic evaluation of environmental factors affecting boundary detection accuracy (vegetation effects, surface roughness variations, climatic influences) and comprehensive quantification of methodological uncertainties across various environmental, geomorphological, and volcanic settings. The development of criteria-based selection frameworks is essential to support identification of the most appropriate methods for specific applications, thereby limiting errors in results and interpretation.

Following this review, it is crucial to evaluate the application of morphometric measurements, such as morphometric dating, shape classification, and inferring process from shape. Given the errors and inaccuracies that can appear with morphometric measurements, as outlined in this study, it is possible unknown errors have appeared within the applications of the measurements. Therefore, it is crucial to analyse how the different methods to obtain morphometric parameters may impact the results of morphometric dating (e.g. height/width ratios).

The development of standardised protocols for calculation methods, boundary delineation procedures, and comprehensive error quantification remains crucial for advancing the field of volcanic morphometry. Only through such methodological standardisation can the volcanological community develop reliable, reproducible approaches to morphometric analysis that can support robust volcanic hazard assessment and process understanding across diverse volcanic fields worldwide.