Indian ocean warming, extreme positive Indian Ocean Dipole events, and their impact on monthly Indian Monsoon rainfall from June to November in NMME models

Abstract

Indian Ocean (IO) warming, the frequency of extreme positive Indian Ocean Dipole (EPIOD) events, and its impact on Indian monsoon rainfall (IMR) are more prominent in the recent era. A proper understanding of the relationship between the EPIOD and IMR is very important because two thirds of the people in India are dependent on agriculture and related economy. The primary focus of this study is on three main aspects: (i) analysing the monthly trend of sea surface temperature (SST) in the IO; (ii) the spatial variability of SST anomalies (SSTA) linked to the EPIOD and its impact on the IMR; and (iii) the factor responsible for the intensification of the EPIOD. These aspects are thoroughly assessed using 12 NMME models, comparing them with observations and reanalysis datasets from 1990 to 2020. The analysis reveals a warming trend in the central IO during June and July, with significant increases observed from August to November. Over the past 31 years, the central IO has experienced an approximate rise of 0.9 to 1 °C from June to November. Among the 12 models scrutinised, GFDL_FLOR_A, FLOR_B, and RSMAS_CCSM4 closely align with observed spatial trends. In contrast, models such as GFDL_SPEAR and NASA_GEOSS tend to overestimate these spatial trends. An increase in easterlies over the equatorial eastern IO was identified as a possible driver for the rising SST in the central IO during October and November. In response to EPIOD on the monthly spatial rainfall distribution over India, all NMME models often underestimate rainfall over the west coast and exhibit wet biases over the leeward side of the Western Ghats. Grid-wise skill scores for accumulated monthly rainfall demonstrated variations among NMME models, with GFDL_SPEAR, RSMAS_CCSM4, and other IC3 models showing higher skill. The interhemispheric pressure gradient (IHPG) has been identified as a precursor to EPIOD events. A positive IHPG from March to June is correlated with EPIOD occurrences from October to December. It is observed that half of the NMME models are capable of simulating a positive relationship between IOD and IHPG, consistent with the observed data.