Run-based multi-model interannual variability assessment of precipitation and temperature over Pakistan using two IPCC AR4-based AOGCMs

摘要

The interannual variability of precipitation and temperature is derived from all runs of the Intergovernmental Panel on Climate Change (IPCC) fourth Assessment Report (AR4)-based two Atmospheric Oceanic General Circulation Model (AOGCM) simulations, over Pakistan, on an annual basis. The models are the CM2.0 and CM2.1 versions of Geophysical Fluid Dynamics Laboratory (GFDL)-based AOGCM. Simulations for a recent 22-year period (1979-2000) are validated using Climate Research Unit (CRU) and NCEP/NCAR datasets over Pakistan, for the first time. The study area of Pakistan is divided into three regions: all Pakistan, northern Pakistan, and southern Pakistan. Bias, root mean square error, one sigma standard deviation, and coefficient of variance are used as validation metrics. For all Pakistan and northern Pakistan, all three runs of GFDL-CM2.0 perform better under the above metrics, both for precipitation and temperature (except for one sigma standard deviation and coefficient of variance), whereas for southern Pakistan, third run of GFDL-CM2.1 perform better expect for the root mean square error for temperature. A mean and variance-based bias correction is applied to bias in modeled precipitation and temperature variables. This resulted in a reduced bias, except for the months of June, July, and August, when the reduction in bias is relatively lower.