Sensitivity of Asian Summer Monsoon precipitation to Tropical Sea Surface Temperature Anomalies

2020-05-12113

Title: Sensitivity of Asian Summer Monsoon precipitation to Tropical Sea Surface Temperature Anomalies

Journal: Climate Dynamics, 47: 2501-2514

Authors: FAN L.*, S-I Shin, Z. -Y. Liu, and Q. -Y. Liu

Abstract:Sensitivity of Asian Summer Monsoon (ASM) precipitation to tropical sea surface temperature (SST) anomalies was estimated from ensemble simulations of two

atmospheric general circulation models (GCMs) with an array of idealized SST anomaly patch prescriptions. Consistent sensitivity patterns were obtained in both models. Sensitivity of Indian Summer Monsoon (ISM) precipitation to cooling in the East Pacific was much weaker than to that of the same magnitude in the local Indian–western Pacific, over which a meridional pattern of warm north and cold south was most instrumental in increasing ISM precipitation. This indicates that the strength of the ENSO–ISM relationship is due to the large-amplitude East Pacific SST anomaly rather than its sensitivity value. Sensitivity of the East Asian Summer Monsoon (EASM), represented by the Yangtze–Huai River Valley (YHRV, also known as the meiyu–baiu front) precipitation, is non-uniform across the Indian Ocean basin. YHRV precipitation was most sensitive to warm SST anomalies over the northern Indian Ocean and the South China Sea, whereas the southern Indian Ocean had the opposite effect. This implies that the strengthened EASM in the post-Niño year is attributable mainly to warming of the northern Indian Ocean. The corresponding physical links between these SST anomaly patterns and ASM precipitation were also discussed. The relevance of sensitivity maps was justified by the high correlation between sensitivity-map-based reconstructed time series using observed SST anomaly patterns and actual precipitation

series derived from ensemble-mean atmospheric GCM runs with time-varying global SST prescriptions during the same period. The correlation results indicated that sensitivity maps derived from patch experiments were far superior to those based on regression methods.