Intercomparison of Univariate and Joint Bias Correction Methods in Changing Climate From a Hydrological Perspective

Olle Räty,Thomas Bosshard,Jouni Räisänen,Chantal Donnelly

doi:10.3390/cli6020033

Olle Räty, Thomas Bosshard + Show 2 more

Open Access

https://doi.org/10.3390/cli6020033

Copy DOI

Abstract

In this paper, the ability of two joint bias correction algorithms to adjust biases in daily mean temperature and precipitation is compared against two univariate quantile mapping methods when constructing projections from years 1981–2010 to early (2011–2040) and late (2061–2090) 21st century periods. Using both climate model simulations and the corresponding hydrological model simulations as proxies for the future in a pseudo-reality framework, these methods are inter-compared in a cross-validation manner in order to assess to what extent the more sophisticated methods have added value, particularly from the hydrological modeling perspective. By design, bi-variate bias correction methods improve the inter-variable relationships in the baseline period. Cross-validation results show, however, that both in the early and late 21st century conditions the additional benefit of using bi-variate bias correction methods is not obvious, as univariate methods have a comparable performance. From the evaluated hydrological variables, the added value is most clearly seen in the simulated snow water equivalent. Although not having the best performance in adjusting the temperature and precipitation distributions, quantile mapping applied as a delta change method performs well from the hydrological modeling point of view, particularly in the early 21st century conditions. This suggests that retaining the observed correlation structures of temperature and precipitation might in some cases be sufficient for simulating future hydrological climate change impacts.

Highlights

In recent years, bias adjustment has become the de facto standard for preprocessing global (GCM) and regional (RCM) climate model simulations for climate change impact studies, hydrological modeling being no exception
In this data set the monthly means of daily mean temperature and precipitation have been adjusted for biases in relation to the large scale gridded observations produced by the Climatic Research unit (CRU) and the Global Precipitation Climatology Centre (GPCC) with subsequent elevation corrections applied to both variables
When first concentrating on the results shown for the years 2061–2090, it is seen that bias correction methods M2–M4 slightly outperform M1 in adjusting the temperature distribution in terms of both Cramér–von Misés (CM) and mean absolute error (MAE)

Summary

Introduction

Bias adjustment has become the de facto standard for preprocessing global (GCM) and regional (RCM) climate model simulations for climate change impact studies, hydrological modeling being no exception. Due to systematic errors in climate model simulations with respect to the observed climate, GCM and RCM output usually cannot be directly used in impact modeling, as impact models require unbiased, high-resolution information as their input. This is because of non-linear and threshold processes within impact models. A cold bias in forcing data to a hydrological model could lead to an impact result indicating no change in snow depths if the cold bias kept temperatures below 0 degrees

Methods

Results

Conclusion