Evaluating temperature bias corrections in regional climate models: Insights from Rajkot's historical and future data

This study evaluates bias correction of the Regional Climate Model (RCM) outputs from the CORDEX South Asia dataset for Rajkot district of Gujarat, using data from the Indian Meteorological Department (IMD). It focused mainly on daily maximum and minimum temperatures during the historical period (1951-2005) and future projections (2006-2100) under RCP 2.6, RCP 4.5, and RCP 8.5. The bias correction method used was Gaussian distribution mapping, and statistical parameters such as monthly mean, coefficient of variation (CV), skewness (Cs), and kurtosis (Ck) were analysed. During the control period (1951-2005), raw RCM data was found underestimated than the observed IMD temperature data. After bias correction, monthly means closely matched the observations, with an R² of 1.00 for both maximum and minimum temperatures during calibration (1951-1995) and validation (1996-2005). The CV improved significantly after correction, with values aligning with IMD. However, the Gaussian method failed to correct higher-order moments such as skewness (Cs) remained negative during calibration but became positive for some months during validation, and kurtosis (Ck) stayed positive for all months, indicating the persistence of heavy-tailed distributions. For future scenarios (2006-2100), bias-corrected RCM data was found overestimated for monthly mean temperatures across all RCPs. The CV was consistently lower than in raw data. Skewness (Cs) and kurtosis (Ck) showed varied results: under RCP 2.6 and RCP 4.5, Cs was mostly positive in winter months, while Ck was positive in both summer and winter. Under RCP 8.5, skewness and kurtosis were positive for most months. Overall, Gaussian distribution mapping effectively corrected the mean temperature values and provided a solid foundation for temperature projections, making it a valuable method for adjusting climate models, though further refinement may be needed to address the extreme temperature events for enhanced climate resilience planning.