The NASA TRMM group's vertical distribution of heating datasets, developed by Dr Tao and his associates, are used in this study to improve the vertical distribution of heating from a suit of mesoscale multimodels for the monsoon region. The procedure is similar to what was done previously for a suite of global models (Krishnamurti et al 2009 JAS). Here we use a suite of mesoscsale models at a horizontal resolution of 25 km which is compatible to the resolution of the heating data sets provided by NASA. The experiments cover the summer seasons for the years 2004 and 2005. This includes daily six day forecasts for each day .We construct the multimodel consensus using our superensemble algorithm. The individual model forecast biases for the vertical distribution of heating and its amplitudes are corrected by the multimodel superenemble. The mesoscale model results appear to be much closer to the TRMM based magnitudes. These forecast improvements of heating through day 6 of forecasts are a considerable improvement compared to the same methodology applied to a global suite of models. In addition to heating this study also illustrates the much improved forecasts of precipitation and circulations for the Asian monsoon derived from the same methodology. Here we have compare these superensemble forecasts with those obtained from the member models of the ensemble suite. In summary the circulations, precipitation and the heating are all much improved by the proposed methodlogy.