2.4.3. Weather Data: Forcing on-Field Weather Data to a Meteorological Model

Weather data was obtained from two sources, on-field weather station data and Météo France's Système d'Analyse Fournissant des Renseignements Adaptés à la Nivologie (SAFRAN) model [74]. The SAFRAN model was used since it contains records for the period 1970–2018 while the weather stations, located next to a pond network, were used to account for the climatic regulating effect reported by Nedjai et al. (2018) [36]. SAFRAN was used in order to provide a large time span for the study. Moreover, the SAFRAN was validated by test-correlation with weather station data using Pearson's correlation coefficient "r" [75]. Prior to correlation, harmonization of both datasets was performed since the weather stations record parameters at the hourly time step, while SAFRAN simulates at the daily time step. Results of the correlation are presented in Table 2.


**Table 2.** Validation testing of the SAFRAN model against measured on field-data for revealing SAFRAN's validity for use.

For the CORINE model, the rainfall and temperature parameters are used for computation of the MFI and BGI indices following their respective formulas (Figure 2); these indices were then used to calculate erosivity following the workflow of Figure 2. In SWAT, the weather database was built using the WGN maker macro-tool [76] and input into the SWAT database.
