Author(s): Kazuaki Yorozu; Atsushi Kambo; Yasuto Tachikawa

Linked Author(s):

Keywords: Bias correction; Discrete Fourier transform; Runoff; Streamflow

Abstract: Typically, General Circulation Models (GCMs) are used for future climate projections. Since the GCM outputs contain biases from various sources, their correction is necessary when utilizing them for impact assessments. A previous study proposed a bias correction method that applied the Quantile-Quantile Mapping (QQM) method to runoff outputs from the MRI-AGCM3.2S model. While QQM is a widely used method for bias correction, it does not directly account for the temporal continuity of the variable. Meanwhile, bias correction methods focusing on the characteristics of time-series data of target variable have been proposed. In this study, we attempted to apply discrete Fourier transform to the bias correction of runoff data generated by the MRI-AGCM3.2S model. Comparing the results with the QQM method used in a previous study, it was found that newly proposed method demonstrated a correction accuracy comparable to that of QQM.

DOI:

Year: 2026