Delayed Impacts of ENSO on the Frequency of Summer Extreme Hot Days in the Asian Monsoon Region. Part II: Implication for Seasonal Prediction

Abstract

Abstract Understanding and predicting extreme weather and climate are ultimately important for adaptation and resilience. Here, we assess the prediction skills of frequency of summer extreme hot days (SEHDs) in the Asian monsoon region (AMR) by using the 1981–2014 hindcasts of the Predictive Ocean Atmosphere Model for Australia version 2 (POAMA-2) subseasonal-to-seasonal prediction system. Generally, the good prediction skills of SEHD frequency appear in the southern AMR south of 30°N including the Indian subcontinent, Indo-China peninsula, and the Philippines; the anomaly correlation coefficients between the observed and predicted region-mean SEHD anomalies are 0.72, 0.72, 0.70, 0.60, 0.61, and 0.61 at 1–6 months lead, respectively, all statistically significant at the 99.9% confidence level. The high prediction skills in POAMA-2 are due to its capacity on reproducing the observed delayed ENSO impacts on the large-scale atmospheric circulation, the seasonal surface air temperature, and thus the SEHD frequency in the southern AMR via provoking the Indo–western Pacific Ocean capacitor effect. Since POAMA-2 only provides the prediction at most half a year in advance, we also conduct the hindcasts in the SINTEX-F that can reproduce well the robust ENSO–SEHD relationship and has high prediction skill of ENSO itself. Results show that skillful prediction of the region-mean SEHD frequency can be up to 14 months in advance. These results are complementary to Part I of this study that ENSO can impact the SEHDs in the AMR and provide the sources of predictability.