Paper Reviewed
Nair, A., Mohanty, U.C., Robertson, A.W., Panda, T.C., Luo, J.-J. and Yamagata, T. 2014. An analytical study of hindcasts from general circulation models for Indian summer monsoon rainfall. Meteorological Applications 21: 695-707.

In introducing their study of monsoonal rainfall over India, Nair et al. (2014) note that atmospheric general circulation models (GCMs) have historically had "serious difficulties in correctly reproducing monthly or seasonal precipitation," citing the findings of Wang et al. (2008); and, therefore, they set out to further explore what advances may have been made in this area of climatological research over the intervening six years, in a study where they say that "precipitation outputs from retrospective seasonal forecasts made by nine GCMs are used to investigate historical Indian summer monsoon seasonal rainfall variability and predictability over India." And what did they thereby learn?

Nair et al. report that (1) "most of the models are not able to simulate the climatological mean rainfall structure," that (2) over North India "predicted magnitudes are also underestimated," that (3) the inter-annual variability (IAV) of JJAS seasonal precipitation ranges from 0 to 10 mm/day in reality, whereas in the models "it is from 0 to 2 mm/day," such that all of the tested models (4) "underestimate the observed all-India precipitation," along with (5) "its variability," that (6) "all these models tend to exaggerate the ENSO teleconnection pattern," that (7) they "are not able to simulate the weakening of ENSO's influence on the Indian monsoon in the recent decades," that (8) "the monsoon meridional circulation and its related convective activity over the monsoon region need to be modelled in a better way," that (9) "there is still inadequacy in GCM performance and understanding of the monsoon dynamics," that (10) "there is a need to diagnose the GCM's outputs, in order to bring out their flaws," that (11) the models' ensemble mean or MME "is not able to simulate the climatology, IAV, temporal evolution and trend of observation," that (12) "the models are overly biased towards stronger ENSO's influence, with (13) "characteristic ENSO SST anomaly patterns much stronger than those observed."

Last of all, in the concluding sentence of their paper, Nair et al. say their results suggest that "the air-sea coupled processes associated with the DMI [dipole mode index], moisture fluxes and convection associated with monsoon meridional circulation need to be improved in the models in order to better simulate/predict the monsoon rainfall over India, which sounds like a reasonable conclusion, and which simultaneously provides a great rationale for FFF (further future funding) of the climate modelling enterprise.

Reference
Wang, B., Lee, J.Y., Kang, I.S., Shukla, J., Kug, J.S., Kumar, A., Schemm, J., Luo, J.J., Yamagata, T. and Park, C.K. 2008. How accurately do coupled climate models predict the leading modes of Asian-Australian monsoon interannual variability? Climate Dynamics 30: 605-619.