The Indian Summer Monsoon and its Sensitivity to the Mean SSTs: Simulations with the ECHAM4 AGCM at T106 Horizontal Resolution

Abstract

Seasonal mean features as well as subseasonal variations of the Indian summer monsoon are investigated on the basis of two simulations with the ECHAM4 atmospheric general circulation model at a high (T106) horizontal resolution. In the first one (period 1979-1993; AMI) observed monthly mean values of sea surface temperatures (SSTs) and sea-ice extent have been prescribed as lower boundary forcing to the model. The second one (period 1970-1999; TSL) is part of a time-slice experiment, where the atmospheric model has been forced by monthly mean values of SSTs, sea-ice extent and sea-ice thickness originating from a transient simulation with the ECHAM4/OPYC coupled atmosphere-ocean model at a low (T42) horizontal resolution. In TSL and the coupled simulation the concentrations of the important greenhouse gases have been prescribed according to observations until 1990, and after 1990 according to the IPCC scenario IS92a. In AMI the concentrations of the important greenhouse gases have been kept constant at a level close to the observed values for the period 1978-1988.
The comparison between AMI and observations indicates a quite realistic simulation of the seasonal mean characteristics of the Indian summer monsoon, as well as of its subseasonal variability, though some deficiencies remain. These are a warm bias of the seasonal mean temperature over land, leading to a too strong temperature difference between the Indian Ocean and the land areas to the north; an overestimation of the low-level monsoon flow, an underestimation of the upper-level monsoon flow; and, an underestimation of precipitation over India, in particular for the rainfall maximum near the west coast of the Indian peninsula. Other shortcomings are a northward shift and deformation of the ITCZ over the Indian Ocean and over the western Pacific. The latter is associated with an unrealistic simulation of the large-scale circulation over the western Pacific, including the East Asian monsoon. The underestimation of precipitation near the west coast of the Indian peninsula affects the patterns describing the day-to day-variability of rainfall during the monsoon season. Though the most prominent patterns of variability obtained from the simulations have structures similar to those obtained for observations, the strength of the center of action on the west coast of the Indian peninsula is considerably underestimated.
As for the seasonal mean features, most of the model deficiencies are improved in TSL. These improvements can be explained by the following mechanisms: (a) The initial warm temperature bias over the Indian Ocean acts to diminish the underestimation of precipitation over India, the warm temperature bias over India and Pakistan, and the too strong temperature difference between the Indian Ocean and the land areas to the north. (b) The El Niño-like warming pattern in the tropical Pacific helps to reverse the too strong low-level monsoon flow. (c) The erroneous SSTs in the western Pacific generated by the coupled model improve the simulation of the ITCZ and the large-scale circulation over the western Pacific, in particular the East Asian monsoon, and the ITCZ over the Indian Ocean via the Walker circulation. Little improvement is seen, however, in the simulation of the subseasonal variability.