Supplementary material to "Exploring the applicability of Censored Shifted Gamma Distribution (CSGD) error model to radar based rainfall nowcasts: A UK case study"

This supplement outlines the strategy for identifying optimal interpolation models to estimate CSGD (Censored Shifted Gamma Distribution) parameters at ungauged locations.Two widely-used techniques are compared: Ordinary Kriging (OK) and Radial Basis Function (RBF).The selection criteria and key intermediate results are described below.For OK interpolation, an appropriate variogram model must first be identified, as interpolation quality depends on how well the variogram represents the spatial dependence of parameter variability with distance.Twelve commonly-used theoretical 5 models were tested for fitting variograms of the climatological, linear, and non-linear conditional CSGD parameters.As shown in Figure S1, most fitted variograms plateau within 500 km, supporting the decision to truncate the distance threshold at 500 km to reduce outlier influence.The corresponding goodness-of-fit statistics, including R 2 and weighted least squares (WLS), are summarised in Tables S1 and S2.For each parameter, the variogram models with the highest R 2 and the lowest WLS were shortlisted as candidates for OK interpolation. 10