From cumulative NWP precipitation data to small scale rain intensity distribution: Assessment of a procedure

Numerical Weather Predictions (NWPs) may be used as global meteorological input to several models for the estimation of radiowave atmospheric impairments. This contribution extends the evaluation of the effectiveness of SPET (Spatial P(R) Estimation Technique), a methodology recently proposed to estimate the spatial distribution of point rain rate, P_S(R), from cumulative rain precipitation data, namely the total rain amount, M_t, and the convective rain amount, M_c, fallen in a reference time period T within a given area A on a snapshot basis. The performance of SPET in estimating P_S(R) is assessed against weather radar data collected by the NIMROD network over the UK and, in addition, according to the well established quasi ergodicity properties of the rainfall process, against the long term point rainfall statistics, P_T(R), collected by raingauges in some sites. Results confirm the good accuracy of SPET in estimating P_S(R), which improves as T and A increase. Preliminary tests against the DBSG3 database of measured P_T(R)s indicate encouraging results, as well as a tight dependence of the performance of SPET on the correctness of the M_t value provided by NWPs used as input to the estimation procedure.