Articles | Volume 8, issue 3
Research article
28 Sep 2017
Research article |  | 28 Sep 2017

A method to preserve trends in quantile mapping bias correction of climate modeled temperature

Manolis G. Grillakis, Aristeidis G. Koutroulis, Ioannis N. Daliakopoulos, and Ioannis K. Tsanis

Abstract. Bias correction of climate variables is a standard practice in climate change impact (CCI) studies. Various methodologies have been developed within the framework of quantile mapping. However, it is well known that quantile mapping may significantly modify the long-term statistics due to the time dependency of the temperature bias. Here, a method to overcome this issue without compromising the day-to-day correction statistics is presented. The methodology separates the modeled temperature signal into a normalized and a residual component relative to the modeled reference period climatology, in order to adjust the biases only for the former and preserve the signal of the later. The results show that this method allows for the preservation of the originally modeled long-term signal in the mean, the standard deviation and higher and lower percentiles of temperature. To illustrate the improvements, the methodology is tested on daily time series obtained from five Euro CORDEX regional climate models (RCMs).

Short summary
We present a methodology to adjust the systematic errors of climate-modeled temperature with a simultaneous long-term trend preservation. The method considers the normalization of the temperature towards a reference period modeled temperature and the estimation of a residual signal, in order to apply adjustment only to the former. The skill of the methodology is compared to other methods while also assessed on the European scale.
Final-revised paper