Sub-Seasonal-to-Seasonal Convective-Permitting Forecasts of Extreme Events over Arabian Peninsula

In the past years, we have seen several extreme events over the Arabian Peninsula (AP). Among them are extreme precipitation over Jeddah from the 24th to 25th November 2022 and extreme temperature during the Hajj Season over Mecca from the 26th June to 1st July 2023. The former caused flooding and casualties in Jeddah, reminiscent of the 25th November 2009 extreme precipitation event and flood in Jeddah. The latter resulted in a concerning situation for the Hajj pilgrims as they made their activities outdoors during the Hajj Season. In this research we tested our Sub-seasonal-to-seasonal (S2S) convective-permitting model (CPM) capability as shown in Risanto et al (2022) to hindcast the former and forecast the latter. We utilized the Advanced Research version of the Weather Research and Forecasting (WRF-ARW) model with the S2S ECMWF forecasts with 51 ensemble members as its boundary forcing. The extreme precipitation event was hindcast with 2-week, 3-week, and 4-week lead times whereas the extreme heat events was forecast with 4-week lead time. To evaluate the extreme precipitation, statistical analyses such as Critical Success Index and Probability of Detection were used relative to the radar observation. Several precipitation thresholds were applied in these analyses. To evaluate the extreme heat, daily forecast probability over Mecca and the AP was calculated based on all 51 ensemble members with several temperature thresholds greater than 40 degrees Celsius. The temperature forecasts were also evaluated using the temperature of the ECMWF Reanalysis (ERA5). This research compared the skill of the S2S-CPM WRF hindcasts / forecasts with its S2S ECMWF forecasts. It shows that the S2S-CPM WRF predictions have higher forecast skills of extreme precipitation and extreme heat in time, space, and intensity than the S2S ECMWF forecasts.