The Key Role of Temporal Stratification for GCM Bias Correction in Climate Impact Assessments.

Characterizing climate change impacts on water resources typically relies on Global Climate Model (GCM) outputs that are bias‐corrected using observational data sets. In this process, two pivotal decisions are (a) the Bias Correction Method (BCM) and (b) how to handle the historically observed time series, which can be used as a continuous whole (i.e., without dividing it into sub‐periods), or partitioned into monthly, seasonal (e.g., 3 months), or any other temporal stratification (TS). Here, we examine how the interplay between the choice of BCM, TS, and the raw GCM seasonality may affect historical portrayals and projected changes. To this end, we use outputs from 29 GCMs belonging to the CMIP6 under the Shared Socioeconomic Pathway 5–8.5 scenario, using seven BCMs and three TSs (entire period, seasonal, and monthly). The results show that the effectiveness of BCMs in removing biases can vary depending on the TS and climate indices analyzed. Further, the choice of BCM and TS may yield different projected change signals and seasonality (especially for precipitation), even for climate models with low bias and a reasonable representation of precipitation seasonality during a reference period. Because some BCMs may be computationally expensive, we recommend using the linear scaling method as a diagnostics tool to assess how the choice of TS may affect the projected precipitation seasonality of a specific GCM. More generally, the results presented here unveil trade‐offs in how BCMs are applied, regardless of the climate regime, urging the hydroclimate community to carefully implement these techniques. Plain Language Summary: Global Climate Models (GCMs) are useful tools to characterize the historical and future evolution of the Earth's climate and its impacts on water resources. Because these models contain errors and their horizontal resolution is too coarse for local impact assessments, spatial downscaling, and bias correction are required steps. In particular, bias correction methods can be trained and applied using all the available historical data or by splitting the time series (e.g., by season or months). Since there are no guidelines for selecting a temporal stratification (TS), we analyze bias‐corrected GCM outputs obtained using three types of strategies (entire period, seasons, and months) and seven bias‐correction techniques over continental Chile. We show that the choice of BCM and the TS applied can modify the projected precipitation signal and seasonality. We also propose using a simple statistical technique to identify if the TS may be a relevant decision for climate impact assessments for a given climate model. Key Points: The choice of temporal stratification (TS) for GCM bias correction is crucial for removing biases, even for GCMs with good raw seasonalityDifferent temporal stratifications used for GCM bias correction may yield different future seasonalities and signals in projected changesThe linear scaling approach can be used to easily identify GCMs whose projections are sensitive to the choice of TS [ABSTRACT FROM AUTHOR]