Diabatic heating model of the Indian monsoon

Abstract

In Part I of this paper, influence functions are derived for the response of a quasi-geostrophic atmosphere to transient heat sources and sinks, assuming that the effects of horizontal advection can be neglected and assuming a fairly reasonable vertical distribution of static stability. The influence is studied for diabatic heating of different horizontal wavelengths and for two different types of the vertical distribution. In Type I, heating is largest at the ground, decreasing to zero at p = 0. In Type II, heating is maximum in the middle atmosphere and decreases parabolically to zero at p = 0 and at the ground. It is shown that, in both types, the horizontal wavelength L of the heating function is very important in determining not only the intensity of pressure fall in the lower levels and of pressure rise aloft in the region of heating, but also the level of maximum pressure effect. It is seen that wavelengths of the order of 15,000 km produce maximum geopotential variations around the 150-mb level. Introduction of Ekman layer friction decreases the intensity of pressure fall in the lower layers, increases the intensity of pressure rise aloft, lowers the level of phase reversal, and introduces a phase lag between the high pressure wave aloft and the low pressure wave below. Part II deals with the application of theoretical results obtained in Part I to the problem of the Indian monsoon. It is visualized that the 12-monthly monsoon oscillation in southeast Asia is a linear perturbation on the annual mean flow pattern, the perturbation being essentially forced by differential diabatic heating in the horizontal plane; the perturbation is materially affected by low-level friction, while advection is assumed to be only of secondary importance. In the first instance a 2-dimensional model of the monsoon in the y, p plane is constructed along the meridian 77.5°E, where observed annual mean conditions are taken as a basic state. On this is superimposed a linear perturbation forced by diabatic heating, sinusoidal in y and t and incorporating a combination of heating of Types I and II. The resulting total patterns of zonal wind in different months are presented. It is very encouraging to find that such a simple model, with only one wavelength in the y direction, is able to reproduce quite a few observed features of the zonal wind pattern in all months, including the westerly jet stream in winter and the easterly, jet stream in summer.