Observed, stochastically simulated, and projected precipitation variability in Pakistan

Abstract

Stochastic simulation of daily precipitation at 46 different weather stations across Pakistan is assessed for the 36-year baseline period during 1976–2011 using a stochastic precipitation generator, both on seasonal and annual basis. The precipitation projections of 15 intergovernmental panel on climate change (IPCC) fourth assessment report (AR4)-based Atmospheric Oceanic General Circulation Models (AOGCMs) are assessed for three 20-year projected time periods centered at 2011–2030, 2046–2065, and 2080–2099, which are embedded in the stochastic precipitation generator. The slope difference method is used to delineate the three precipitation regimes in Pakistan: arid, semi-arid, and humid. The change in (i) amount of precipitation, (ii) number of wet days, (iii) amount of precipitation per wet day, and (iv) area of precipitation regimes are assessed for all three projection time periods relative to the baseline period. Pakistan has received 487.48 mm of average annual precipitation, with an average of 43.54 wet days year−1 during 1976–2011, whereas simulated average annual precipitation is 513.70 mm with 43.60 wet days in a year. The stochastic precipitation generator has less than 2% error in simulating the mean annual numbers of wet days during the baseline period, whereas simulated mean annual precipitation is 11.78 mm wet day−1 versus the observed 11.70 mm wet day−1. There is an underestimation of ~ 1% in simulated mean wet spell length during the baseline period. The AOGCMs display varying increase (decrease) in humid and semi-arid (arid) areas in Pakistan in all the three projected time periods.