Effect of soil hydraulic parameters on the local convective precipitation

The impact of soil hydraulic parameters upon the formation of local convective precipitation is investigated by using the Penn State-NCARMM5 Modeling System. Four soil hydraulic parameter effects are considered. These comprise the effect of two different soil datasets (US vs. Hungarian), the effects of different parameterizations of field capacity and the wilting point soil moisture contents and the subgrid-scale variability effect of soil moisture content, respectively. Precipitation from 6 convective events were simulated and verified on a spatial scale of 18 km by comparison with observed precipitation fields. The latter were analysed from a total of 657 precipitation gauges using ordinary block kriging. It is demonstrated that the simulated convective precipitation is sensitive to the used soil dataset. The rank order correlation coefficient of R-s = 0.62 (explained variance 38 %) between observed precipitation and simulations using the Hungarian soil dataset (control run) is higher than those using the US soil dataset (R-s = 0.53, explained variance 29 %). The difference of about 10 % explained variance is highly significant (p < 0.01). Simulations with alternatively field capacities result in a correlation of R-s = 0.60, which is not significantly different from the control run. Simulations with alternatively wilting points, however, result in R-s = 0.56, which is significantly lower than the correlation for the control run (p < 0.05). Finally, the application of an alternatively spatial distribution of soil moisture contents results again in a lower correlation of R-s = 0.54 (p < 0.01). Amongst the four soil effects, the soil dataset effect and the subgridscale variability of the soil moisture were greater than the parameterization effects. The results suggest that regional soil datasets should be preferred in the mesoscale models. It is also obvious that high-resolution global soil data sets in the climate models should be constructed extremely carefully because of the sensitivities obtained.