Anthropogenically-forced weakening of the Indian summer monsoon and enhancement of the western North Pacific tropical cyclogenesis

Abstract

This study addresses the role of human-induced climate change on the interactions of convective activities between the Indian summer monsoon (ISM) and western North Pacific (WNP) regions - an important scientific issue which has been hitherto overlooked. We have examined this problem using two numerical experiments of a high-resolution climate model, with and without anthropogenic forcing (i.e., HIST and HISTNAT) for the historical period 1951–2005, supplemented by innovative diagnostics like causal network analysis. Our findings suggest that an anthropogenically-forced weakening of the ISM circulation tends to significantly enhance the genesis potential index (GPI) of the WNP tropical cyclones by 13.5% and associated convective activities, by reorienting the large-scale flow over the tropical Indo-Pacific in a manner that is conducive for enhancement of the WNP tropical cyclogenesis. Additionally, it is found that the probability of extremely low sea-level pressure (SLP) (<995.5 hPa) around Taiwan and Chinese mainland is significantly higher by 10.3% in the anthropogenically-forced simulation as compared to the natural run. Using the model outputs from HIST and HISTNAT, we also performed a causal effect network (CEN) analysis to understand the causal connections among the three indices involved in the ISM-WNP interactions (a) Indian monsoon circulation index (IMI) which is the difference in the area-averaged 850 hPa zonal winds between the boxes (40oE-80oE, 5oN-15oN) and (70oE-90oE, 20oN-30oN) (b) WNP tropical cyclone activity expressed as the genesis potential index (GPI) averaged over the region (120oE-180oE, 5oN-30oN) and (c) Tropical Indo-Pacific wind index (IPWND) which is based on the 850 hPa zonal winds averaged over the domain (100oE-130oE, 5oN-20oN). Results from the CEN analysis indicate that an anthropogenically-forced weakening of the IMI can lead to possible strengthening of GPI and IPWND with time-lags of 5 and 7 days, respectively. In general, it is noted that the causal relationships among IMI, GPI and IPWND are associated with shorter time-lags (∼4–9 days) in HIST and longer time-lags (∼19–28 days) in HISTNAT.