Comparison of bias correction methods for climate change projections in the lower Shivaliks of Punjab

Abstract

Abstract The study evaluates the performance of bias correction techniques by dividing the observed climate data period into calibration and validation sets. For this purpose, the daily data of temperature, rainfall, and solar radiation from 2010–2095 for lower Shivaliks of Punjab (Ballowal Saunkhri) were downloaded from Marksim weather generators using outputs of CSIRO-Mk3-6-0 climate model under four RCP scenarios (RCP 2.6, RCP 4.5, RCP 6.0, and RCP 8.5). The bias correction of model data (temperature, rainfall, and solar radiation) was done by developing correction functions (using a model and observed data from 2010 to 2015) from different bias correction methods (difference method, Leander and Buishand method, modified difference method, linear scaling, variance scaling, and quantile mapping). The corrected model data for the year 2016–2020 were validated against the observed data. The difference method was found to be best for bias correction due to low error and high efficiency. The corrected future model data (2021–2095) analysis on an annual and seasonal basis predicted a rise in maximum temperature and minimum temperature by 1.3–2.8 °C and 0.5–3.0 °C, respectively, under different scenarios. The study predicted more increase in rainfall and solar radiation under RCP 8.5 followed by RCP 6.0, RCP 4.5, and RCP 2.6 scenarios.