A multi-model analysis of changing climate in the Hindu Kush-Himalayan region using CMIP3 projections for temperature and precipitation

The Hindu Kush-Himalayan (HKH) region epitomizes regions where cryospheric processes coupled with hydrological regimes are under threat owing to a warming climate. Information from multi-model ensembles of changes in temperature and precipitation in the HKH region can not only provide uncertainty in projections as represented by model consensus, but also regional and interannual variability of these changes to identify critical areas. We analyze global climate models in the Coupled Model Intercomparison Project (CMIP3) archive to quantitatively evaluate the changes occurring in temperature and precipitation. Specifically, we examine and evaluate multi-model, multi-scenario (B1, A1B, and A2) climate change projections and seven extreme temperature and precipitation indices over the HKH for the 21st century. The probability density functions (PDF) of all temperature and precipitation indices are outside their 20th century ranges by the end of the 21st century. PDFs of observed climate indices for available meteorological stations with long term data are also analyzed to provide late 20th century distributions of climate indices. PDFs of simulated intra-annual extreme temperature range indicate negative shifts during the 21st century, which is an indication of rapidly increasing minimum temperatures than the maximum temperatures. The time evolution of changes in simulated annual extreme indices is also evaluated by calculating multimodel average time series for each scenario covering the period 1901-2099. Time series of multimodel average temperature and precipitation indices exhibit increases in consecutive dry days, 5-day precipitation, fraction of total annual precipitation due to events exceeding the 1961-1990 95th percentile, and precipitation intensity, and decreases in intra-annual extreme temperature range and total number of frost days. These results provide a detailed assessment of coarse-scale, multi-model ensembles of global climate model scenarios of future changes in climatic variables known to affect both regional hydrology and glacier mass balance.