The first rainfall erosivity database in Mexico: facing challenges of leveraging legacy climate data

Abstract. Soil Water Erosion (SWE) is the dominant soil degradation driver on a global scale. For quantifying SWE, erosivity is an index that reflects the potential (i.e., the energy) of rainfall to cause SWE. To enhance the assessment of the SWE process at the national scale—, the objectives of this research are a) to develop the first Mexican rainfall time series database for three climate normals CNs (1968–1997, 1978–2007, and 1988–2017) leveraging legacy climate data, and b) to estimate rainfall erosivity across continental Mexico by using daily rainfall time series. The workflow has three methodological moments: 1) development of the rainfall time series database, 2) estimation of rainfall erosivity, and 3) rainfall erosivity verification. First, we compiled and harmonized over 5000 useful rainfall time series (RTS) well distributed across the Mexican territory. We performed a quality assurance, homogeneity analysis (using the normal homogeneity test), and data gap-filling (using the proportion method). Then, we use a potential power law equation to estimate rainfall erosivity at daily resolution. Finally, we compared and verified our results with three external datasets (global, national, and local scales). The principal research product is a new database with 1370, 1678, and 1676 RTS for each CN and its corresponding rainfall erosivity. The mean values for rainfall erosivity for the three CNs were 3600, 3296, and 3461 MJ mm ha^-1 h-¹ yr^-1, respectively. The statistical distribution of the erosivity values was right-skewed for the three CNs, with high erosivity values reaching >8000 MJ mm ha^-1 h-¹ yr^-1 in all the three CNs. About the verification of erosivity values, we found that Tropical rain-forests, temperate Sierras, and the Great Plains are the ecoregions with more significant differences concerning the global database, a generalized underestimation of erosivity values concerning the national dataset, and an adjustment coefficient of 1.85 for a local condition in Michoacan state. This new database provides tools for daily climatological analysis across Mexican territory and through a multiyear period (1968 to 2017). Erosivity results trigger the study of SWE at the national scale by identifying areas with higher susceptibility to soil loss due to rainfall action and providing a more spatially dense erosivity database that follows the pattern of erosivity databases from higher time resolution. Following the FAIR principles (Findability, Availability, Interoperability, and Reproducibility) for scientific data, this database is available from a scholarly accepted repository (https://doi.org/10.6073/pasta/7479676e406aeb40127da7b096b28eb2) for public consultation.