Forecastability of a heavy precipitation event at different lead-times using WRF model: the case study in Karkheh River basin

Since insufficient warning time limits the possibility of taking precautionary actions by managers of water resources, forecast lead time in the hydro-meteorological warning considers as a crucial index. The aim of this study, first, is to investigate the sensitivity of heavy precipitation forecast toward lead time and, second, to identify forecast lead time, best suited for predicting the heavy precipitation event, on 31 March 2019, on Karkheh River basin in Iran. By applying Weather Research and Forecasting (WRF) model, a total of 12 experiments were designed via combinations of three microphysics (MPS) and two cumulus (with and without for nest domain) parameterization schemes (CPS) over two interactively nested domains. Finally, to achieve the aims, 24-h accumulated precipitation forecasts through the designed experiments at different lead times (up to 198 h) were examined by comparing against observations. The results showed that the 4-km domain has an advantage over the 12-km domain at lead-times shorter than 102 h, while the sensitivity to the use of CPS for the 4-km domain is positively increased at lead-times longer than 102 h. Based on the assessed lead-times, the performance of Grell–Freitas CPS was better than that of Kain–Fritsch CPS. The WSM6 MPS also showed advantages over the Thompson and Goddard MPSs at lead-times shorter than 78 h. The maximum amount and the spatial average of precipitation tend to be underestimated, and the extent of the underestimation increases with lead-time. Taken together, these results suggest that a forecast lead-time of 78–102 h was appropriate for issuing warnings for the targeted heavy precipitation event.