Extended monsoon episodes over south Asia: assessing robustness in thermodynamic processes

Rainfall during the south Asian summer monsoon season (June through September), fluctuates between active and break phases. Extended active or break conditions (lasting 7 days or more) results in floods and droughts, respectively. Understanding the mechanisms that initiate and maintain these extended episodes, and their realistic representations in prediction models are important. Recently, Prasanna and Annamalia (2012), applied moist static energy (MSE) budget diagnostics on ECMWF-Interim reanalysis data and found that dry advection and cloud-radiative processes are, perhaps, responsible for the initiation and maintenance of extended breaks. However, large budget residuals pose limitations. The fact that monsoon region is data sparse and model prejudices influence results, the present study extends earlier results in the following ways: (1) the fact that different physical parameterizations are employed in different reanalysis system, are the identified signals robust? Is there a convergence towards “truth”; (2) what are the processes that initiate and maintain extended active monsoon conditions? and (3) from tendency terms (e.g., MERRA reanalysis), can we identify the source of model errors that account for budget residuals?

To address the above and related questions, we perform MSE diagnostics with all the available reanalysis products (ERA-INT, NCEP, JRA-25, MERRA and CFS-R) on composite as well as individual extended monsoon episodes. We identify “robust” signals in thermodynamical processes that provide precursors for the initiation and maintenance of extended break and active conditions over south Asia. Similar budget diagnostics are also performed over the equatorial Indian Ocean, tropical west Pacific and Maritime Continent regions to assess their “linkages” to occurrences of extended episodes over south Asia. Preliminary results suggest robustness in various physical processes that include (1) the temporal evolution of rainfall characteristics over central India; (2) vertical distribution of specific humidity and temperature anomalies; (3) space time evolution of dry (moist) and cold (warm) air advections and (4) cloud-radiative processes. We will discuss the quantitative contributions of individual processes and identify sources of model errors for budget residuals.

Implication of the present study lies in improving CFSv2 skill in prediction of extended monsoon episodes.