Inherent bias correction is necessary for weather data of global and regional circulation models

The predicted outputs of global and regional circulation models are afflicted with biases. So to assess this biasness, 17 GCMs data available at gisweb.ciat.cgiar.org/MarkSimGCM/ was downscaled for Ludhiana. The observed annual mean Tmax was 30.35°C and the GCMs overestimated it within the range 32.11 (MIROC-ESM) to 33.08°C (IPSL-CM5A-LR), i.e. the overestimation range lied between 1.76 to 2.73°C. The observed annual mean Tmin of 18.28°C was both underestimated (17.71-18.20°C) and overestimated (18.29-18.49°C) by the GCMs. The annual mean observed RF was 542 mm and it was overestimated within the range 635 mm (GFDL-ESM2M) to 997 mm (MICROC-ESM). The annual mean observed SR was 15.29 Wm^-2 its overestimation ranged between 17.29 (MIROC5) to 18.45 Wm^-2 (GFDL-ESM2M). The validation results of the study showed that amongst 17 GCMs, four GCMs (CSIRO-Mk3-6-0, FIO-ESM, IPSL-CM5A-MR) and one Ensemble model predicted Tmax, Tmin, RF and SR satisfactorily. Hence, bias correction in projected data is necessary before its usage in impact assessment studies.