The Skill of Extended-Range ERPAS and GloSea5 Models in Capturing Extratropical–Tropical Interactions During the North Indian Summer Monsoon

Abstract

The transient eddies in the atmosphere are short-lived, moving disturbances prominent over mid-latitude. Transient eddy transport enables the exchange of mass, energy, and moisture between extratropical and tropical regions. Based on observations, about 40% of the rain that falls in northern India during the summer monsoons is influenced by transient eddy heat and momentum fluxes. During these rainfall cases, the four-stage cycle of transient eddy heat and a momentum feedback process exists. Global-scale circulation anomalies are generated due to their forcing on the mean flow. This impact of transient eddies on mean flow is referred to as eddy–eddy feedback, commencing in around 22 days. On a daily scale, the enhancement of rainfall over the Western Ghats and north-western India is linked with upper tropospheric poleward transient eddy heat flux transport and equatorward transient eddy momentum flux transport. On a quasi-biweekly scale, however, the transport direction reverses. Additionally, this rainfall pattern is governed by the meridional passage of monsoon intraseasonal oscillation phases (MISO, tropical mode) and the zonal passage of wave number 7–8 patterns (Rossby wave, extratropical mode). Interactions of tropical–extratropical modes are associated with this eddy –eddy feedback that drives the hemispheric upper-tropospheric circulation patterns. This includes wave generation, propagation, and the dissipation of waves away from the source region. The hindcast skill analysis of subseasonal to seasonal scale models, namely Extended Range Prediction Application to Society (ERPAS) and Global Seasonal Forecast version 5 (GloSea5), shows that the models can predict northern Indian rainfall associated with eddy–mean flow interactions at 1-week lead times. After a week, the skill of both models diminishes under the influence of transient eddy transport. The monsoon circulation is more consistent and predictable in both models when transient eddies are absent. A theoretical understanding of the dynamical feedback of upper-tropospheric transient eddies is crucial for improving rainfall prediction.