Assessment of Indian Summer Monsoon in Seasonal Hindcasts From Next‐Generation GFDL and IITM Models

Abstract

The complex interactions among atmosphere, ocean, land and cryosphere make it challenging to predict Indian summer monsoon rainfall (ISMR) accurately. This study evaluates two next‐generation seasonal prediction systems—GFDL‐SPEAR (Geophysical Fluid Dynamics Laboratory‐Seamless System for Prediction and Earth System Research) and the MMCFSv2 (Monsoon Mission Coupled Forecast System version 2), which incorporate updated model physics and improved dynamical cores. By assessing model hindcasts during the summer monsoon seasons (June–September) for the period 1991–2020, we demonstrate a 2%–16% improvement in ISMR prediction skill–measured as the anomaly correlation between the ensemble mean model ISMR and observed ISMR–compared to their predecessor models. In addition to ISMR prediction skill, we examine the models' ability to represent tropical sea surface temperature (SST) mean states, variability, and their teleconnections with ISMR. Our analysis reveals that while GFDL‐SPEAR accurately represents the SST mean state, MMCFSv2 demonstrates relatively better skill in capturing the interannual variability of the El Niño–Southern Oscillation (ENSO) and Indian Ocean Dipole. A dry rainfall bias (1.25 mm/day) is noted in MMCFSv2, while GFDL‐SPEAR exhibits a comparatively smaller wet bias (0.75 mm/day) over the Indian landmass. MMCFSv2 also shows improved Indo‐Pacific SST–ISMR teleconnections, contributing to its enhanced ISMR skill (0.58 in MMCFSv2 and 0.47 in GFDL‐SPEAR). However, the ISMR prediction skill in both models exhibits considerable decadal variability, with challenges in capturing the decadal fluctuations of the ENSO–ISMR relationship. Our findings emphasise that improved representation of tropical SST teleconnections, rather than mean‐state biases alone, is critical for achieving better ISMR prediction skill. This process‐level understanding provides insights for the continued development of reliable seasonal prediction systems and climate services over South Asia.

Bibliographic Information