Improved bioclimatic variables from regional climate models: a comparative analysis of three classic bias adjustment methods and a novel adjustment approach

Abstract

Abstract Bioclimatic variables (BCVs) play an important role in understanding ecological dynamics and species distribution under climate variability and change. This research assesses the effectiveness of three bias adjustment methods—Linear Scaling (LSC), Empirical Quantile Mapping (EQM), and Quantile Data Mapping (QDM)—on improving BCVs from Regional Climate Models (RCMs) over Europe, with ERA5-Land as reference dataset. Among the methods tested, EQM slightly stands out for its ability to accurately adjust temperature and precipitation variables, as well as BCVs. However, the different with the two other bias-correction methods is marginal. RCMs were not effective in representing interactive BCVs, especially the mean temperature of the wettest quarter (BIO8) and the mean temperature of the driest quarter (BIO9) and significant residual bias remains after the application of bias adjustment methods. A key innovation of this study was the application of a static quarter/month adjustment in conjunction with EQM, which significantly improved the representation of interactive BCVs. This novel approach effectively addressed seasonal and spatial discrepancies, aligning modeled climate patterns more closely with observed data, thereby improving the model’s ecological relevance. The finding of this study contributes to a more nuanced understanding of ecological dynamics under climate change, offering valuable insights for species distribution modelling based on BCVs.