Technological advancements like increase in computational power have led to high-resolution simulations of climate variables by Global Climate Models (GCMs). However, significant biases exist in GCM outputs when considered at a regional scale. Hence, bias correction has to be done before using GCM outputs for impact studies at a local/regional scale. Six bias correction methods, namely, delta change (DC) method, linear scaling (LS), empirical quantile mapping (EQM), adjusted quantile mapping (AQM), Gamma-Pareto quantile mapping (GPQM) and quantile delta mapping (QDM) were used to bias correct the high-resolution daily maximum and minimum temperature simulations by Meteorological Research Institute-Atmospheric General Circulation Model Version 3.2 (MRI-AGCM3–2-S) model which is part of Coupled Model Intercomparison Project Phase 6 (CMIP6), of Netravati basin, a tropical river basin on the south-west coast of India. The quantile-quantile (Q–Q) plots and Taylor diagrams along with performance indicators like Nash–Sutcliffe efficiency (NSE), the Root-Mean Square Error (RMSE) or Root-Mean Square Deviation (RMSD), the Mean Absolute Error (MAE), the Percentage BIAS (PBIAS) and the correlation coefficient (r) were used for the evaluation of the performance of each bias correction method in the validation period. Considerable reduction in the bias was observed for all the bias correction methods employed except for the LS method. The results of QDM method, which is a trend preserving bias correction method, was used for analysing the trend of future temperature data. The trend of historical and future temperature data revealed an increasing trend in the annual temperature. An increase of 0.051 °C and 0.046 °C is expected for maximum and minimum temperature annually during the period 2015 to 2050 as per RCP 8.5 scenario. This study demonstrates that the application of a suitable bias correction is needed before using GCM projections for climate change studies.

技术进步，例如计算能力的提高，已导致通过全球气候模型（GCM）对气候变量进行高分辨率模拟。但是，在区域范围内考虑，GCM产出存在重大偏差。因此，在将GCM输出用于本地/区域规模的影响研究之前，必须进行偏差校正。共有六种偏差校正方法，分别是增量变化（DC）方法，线性缩放（LS），经验分位数映射（EQM），调整分位数映射（AQM），伽马-帕累托分位数映射（GPQM）和分位数增量映射（QDM）。由气象研究所-大气通用环流模型版本3.2（MRI-AGCM3-2-S）模型（用于耦合模型比较项目第6阶段（CMIP6）的一部分）用来校正高分辨率的每日最高和最低温度模拟，Netravati盆地，印度西南海岸的热带河流域。分位数（Q–Q）图和泰勒图以及性能指标，如纳什–苏克利夫效率（NSE），均方根误差（RMSE）或均方根偏差（RMSD），平均绝对误差（ MAE），百分比BIAS（PBIAS）和相关系数（r）用于评估在验证期内每种偏差校正方法的性能。对于所有采用的偏差校正方法（除了LS方法），都可以观察到偏差的显着降低。QDM方法的结果是趋势保留偏差校正方法，用于分析未来温度数据的趋势。历史和未来温度数据的趋势表明年温度呈上升趋势。增加0。根据RCP 8.5情景，预计2015年至2050年期间，每年的最高和最低温度分别为051°C和0.046°C。这项研究表明，在将GCM预测用于气候变化研究之前，需要应用适当的偏差校正。