Desert pavements, characterized by a stone layer over fine eolian materials, are key geomorphological features in arid regions and play a critical role in the global dust cycle. Despite their importance, quantitative evidence regarding their global distribution remains limited. This digital object contains raster files and R scripts providing a preliminary assessment of desert pavement distribution using a GIS-based multiple-criteria decision analysis (GIS–MCDA) framework. Key environmental factors, including climate, topography, vegetation, soil texture, and anthropogenic disturbance, were incorporated into a global favorability index at a 1 km × 1 km resolution. Validation against 20 documented desert-pavement research sites revealed a significant association with high index values, with 15% of sites exhibiting an index ≥0.90 and 80% ≥0.75. The model estimates that up to 25.7 million km², or 19.0% of the Earth’s land surface, holds potential for desert pavements. These results provide a foundation for future studies, emphasizing the need for localized, high-resolution research and the integration of geomorphometric and remote-sensing techniques with machine-learning models. This initial global assessment underscores the utility of GIS–MCDA in geomorphic distribution modeling and highlights the importance of refining global environmental datasets for arid landscapes.

Files: dpindex_mod.tif: Main results file with global desert pavement potential index (DPPI); areas with DPPI < 0.75 have been set to zero, and terrestrial no-data regions filled with 0. dpindex_raw.tif: Raw output of index calculation, with values ranging from 0 to 1 and possible small gaps. *_index.tif: Intermediate results of subindex calculations. *.R: R scripts used to process the raw data, calculate subindices, combine them into the final index, generate statistics and plot maps for the associated paper. The methodology of this analysis is described in the paper by Brenning et al. (2025), https://zenodo.org/records/15014982, where additional results and maps can also be found. Information on raw input data (not provided due to size and copyright restrictions) is available in the 01_subindex_calculation.R script file.