Background
The authors write that "because of its variability," the northeast monsoon (Oct-Nov-Dec) "is very important" in that "the rainfall during this period is of huge societal significance because it supports the main cultivation season, known as Rabi, over southern peninsular India," citing in this regard the study of Zubair (2002).

What was done
Working with eight different general circulation models (GCMs), Nair et al. analyzed their outputs (forecasts for the whole season issued in September) by comparing them with high-resolution observed and gridded rainfall data obtained from the India-Meteorological Department for the period 1982-2010.

What was learned
The five Indian researchers quickly discovered that "the predictive power of the models," as they phrased it, "was usually very low." For example, they report that although the models were able "somewhat" to represent "some" of the characteristic features revealed in the observations, (1) the correlation coefficients between model-simulated rainfall and observed rainfall "suggested the incapability of GCMs to model the observations." They additionally found that (2) "signal-to-noise ratio and external variance to error variance for the models was also unsatisfactory." And they discovered that (3) "the models are not able to capture the interannual variability present in northeast monsoon rainfall."

What it means
"It can be concluded," in the words of the scientists who conducted the study, that "the models are not able to capture the inter-annual variability present in northeast monsoon rainfall." And they say that this failure arises "because of deviations of these patterns from those of their observed counterparts, which suggests that the models tend to either overestimate or underestimate the observations."

Reference
Zubair, L. 2002. El Nino-Southern Oscillation influences on rice production in Sri Lanka. International Journal of Climatology 22: 1309-1313.