Evaluating Statistical Bias Correction Techniques to Enhance Precipitation Projections in the Cachi Basin, Peruvian Andes

Introduction: To assess the impact of climate change on watersheds, accurate meteorological data are required. Regional Climate Model (RCM) data sets are widely used to feed ecological and hydrological models. However, in Andean regions such as the Peruvian Andes such as the Cachi basin, the scarcity of observation stations generates biases in the RCMs. This study evaluates the performance effect of a simple (Linear Scaling) and complex (Power Transformation and Distribution Mapping) statistical bias correction technique applied on the set of 10 RCM simulations with and without bias correction (BC), using historical climate data (1981-2004). The suitability of each bias correction method was evaluated by analyzing statistical indices, Taylor diagrams and error distributions that were calculated considering the observed and projected rainfall data. Based on our findings, it was observed that the methods can reduce the strong precipitation bias inherent to RCM simulations, evidenced by positive correlations of 0.8 and 0.7 compared to PISCO and RAIN4PE gridded observations, respectively. However, the data set presents an overestimation and underestimation of rainfall in the uncorrected RCM simulations. On the other hand, the power transformation (PT) exhibits solid agreement with the observed data, standing out over the DM and LS method by adjusting both the mean and the variance. In summary, different correction methods may influence the simulated change signals differently. Furthermore, BC methods should be used with caution taking into account regional climatic characteristics and specific research topics. Objective: The objective of this study is to investigate the performance of statistical bias correction techniques applied to regional climate models (RCMs) in the Cachi River Basin, with the aim of improving the accuracy of precipitation projections. The study seeks to enhance climate impact assessments and hydrological modeling, specifically focusing on the selection of appropriate correction methods to mitigate climate change uncertainties in Andean regions. Theoretical Framework: In this topic, the main concepts and theories that underpin the research are presented. Theories such as bias correction in climate models (Teutschbein & Seibert, 2013), regional climate modeling (CORDEX project), and precipitation projection techniques (Fang et al., 2015) stand out, providing a solid basis for understanding the context of the investigation. These frameworks highlight the importance of statistical corrections to reduce the discrepancies between observed and simulated climate data, which is essential for accurate hydrological forecasting in mountainous regions. Method: The methodology adopted for this research comprises a quantitative study design, which includes the application of various statistical bias correction techniques, such as Linear Scaling (LS), Power Transformation (PT), and Distribution Mapping (DM). Data were collected from gridded observational datasets (PISCO and RAIN4PE) and RCM simulations from the CORDEX-SAM domain for the period 1981-2004. The analysis was conducted using performance metrics such as root mean square error (RMSE), Kling-Gupta efficiency (KGE), and correlation coefficients, to evaluate the effectiveness of the bias corrections. Results and Discussion: The results obtained revealed that all bias correction methods improved the alignment between simulated and observed precipitation data. The Power Transformation (PT) method demonstrated superior performance in adjusting both the mean and variance of the precipitation data, especially in high-altitude areas. In the discussion section, these results are contextualized in light of the theoretical framework, highlighting the implications for hydrological modeling and water resource management in the Peruvian Andes. Discrepancies between the models' performances at different altitudes and the limitations of each correction method are also considered in this section. Research Implications: The practical and theoretical implications of this research are discussed, providing insights into how the results can be applied or influence practices in the field of climate change adaptation and hydrological modeling. These implications could encompass areas such as water resource management, agricultural planning, and disaster risk mitigation in mountainous regions affected by climate change. Originality/Value: This study contributes to the literature by providing a detailed evaluation of bias correction techniques for regional climate models in the Andean region, an area with limited observational data. The originality of this research lies in its focus on a high-altitude basin and its application of multiple correction methods, improving the reliability of precipitation projections. The relevance and value of this research are evidenced by its potential to enhance the accuracy of climate impact assessments and inform water management strategies in vulnerable regions.