On the relationship between soil, vegetation and severe convective storms: Hungarian case studies

The effects of soil hydraulic parameter and stomatal functioning parameterization changes upon the precipitation fields of storms were compared and analyzed. The analysis was performed using results from the Penn State-NCAR MM5 Modeling System. Two sets of soil hydraulic parameters, from the USA and Hungary, were used. Stomatal functioning is parameterized as simply as possible using Jarvis' approach. The days chosen for analysis (18th April 2005 and 7th August 2006) seemed to be favourable for local storms to form when the land-surface/air interaction is the strongest. Both days were wet, however, the prevailing moisture was somewhat larger on 18th April 2005.Precipitation fields were statistically analyzed in details. First, the simulated and observed fields were compared. The observed fields were estimated from rain-gauge data applying the ordinary block kriging interpolation technique. The agreement between the simulated and observed fields was estimated using categorical and continuous verification indices. Significance tests were done to estimate how large the obtained differences were.The results obtained indicate that precipitation fields are at least as sensitive to changes in soil hydraulic parameters as to changes in stomatal functioning parameterization. The simple Student t-test hypothesis was applied to estimate how large the precipitation differences obtained were. According to the estimates, the TSS differences obtained by soil parameter and stomatal functioning parameterization changes are significant on the 10% level. The acquired differences do not depend on the initialization of soil moisture.The results suggest that all weather and climate models used for regional purposes should prefer local soil data instead of some common globally used soil datasets. This is at least as important as the parameterization of stomatal functioning.