Scholarlink

<p>Climate input datasets are a major source of bias and uncertainty in hydrological modelling. The present study investigates the extent to which bias correction of the COordinated Regional Downscaling Experiment (CORDEX) dataset can improve the quality of hydroclimatic variables and drought simulations over South Africa’s Western Cape–a region known for recurring multi-year severe droughts. For the study, we applied three bias-correction techniques on the CORDEX dataset: one univariate bias-correction method (Quantile Delta Mapping; QDM) and two multivariate bias-correction methods (MBC; Spearman rank correlation dependence structure and N-dimensional probability density function transform; hereafter MBCr and MBCn, respectively). We calibrated and used the improved version of the Soil Water Assessment Tool Plus (SWAT +) to simulate hydrological variables over the Western Cape river basins. The Global Meteorological Forcing Dataset (GMFD) was used as a reference dataset for the bias correction of the CORDEX dataset and for forcing SWAT+ during the calibration. We quantify the extent to which the three bias-correction methods improve the quality of simulated hydroclimatic variables and droughts over the region. Our results show that the SWAT+ simulation forced with GMFD data provides credible simulations of streamflow over Western Cape river basins when compared with station observations. The simulations with the original CORDEX dataset capture the spatial distribution and the annual cycle of hydroclimate variables over the basin well. Still, there are some biases and substantial uncertainties in the simulations. The bias-correction methods reduce these biases and uncertainties, but the MBCn method performs best. Due to the standardisation of the hydroclimatic variables in the computation of the hydroclimatic drought indices, the biases in the drought characteristics (intensity and frequency) from the original CORDEX simulations are not substantial, and the bias-correction improvement in drought characteristics is minimal. Hence, although bias correction is essential for the correction of hydrological variables, it is less essential when assessing drought characteristics obtained using standardisation indices. The study results have applications for improving the simulation of hydrological variables and droughts in the Western Cape.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Effect of bias correction on CORDEX simulations of hydrological droughts in the Western Cape region of South Africa

  • 期刊论文

  • 发表日期：

    2026年1月2日

查看全文
  • Myra Naik,
  • Babatunde J. Abiodun

摘要

Climate input datasets are a major source of bias and uncertainty in hydrological modelling. The present study investigates the extent to which bias correction of the COordinated Regional Downscaling Experiment (CORDEX) dataset can improve the quality of hydroclimatic variables and drought simulations over South Africa’s Western Cape–a region known for recurring multi-year severe droughts. For the study, we applied three bias-correction techniques on the CORDEX dataset: one univariate bias-correction method (Quantile Delta Mapping; QDM) and two multivariate bias-correction methods (MBC; Spearman rank correlation dependence structure and N-dimensional probability density function transform; hereafter MBCr and MBCn, respectively). We calibrated and used the improved version of the Soil Water Assessment Tool Plus (SWAT +) to simulate hydrological variables over the Western Cape river basins. The Global Meteorological Forcing Dataset (GMFD) was used as a reference dataset for the bias correction of the CORDEX dataset and for forcing SWAT+ during the calibration. We quantify the extent to which the three bias-correction methods improve the quality of simulated hydroclimatic variables and droughts over the region. Our results show that the SWAT+ simulation forced with GMFD data provides credible simulations of streamflow over Western Cape river basins when compared with station observations. The simulations with the original CORDEX dataset capture the spatial distribution and the annual cycle of hydroclimate variables over the basin well. Still, there are some biases and substantial uncertainties in the simulations. The bias-correction methods reduce these biases and uncertainties, but the MBCn method performs best. Due to the standardisation of the hydroclimatic variables in the computation of the hydroclimatic drought indices, the biases in the drought characteristics (intensity and frequency) from the original CORDEX simulations are not substantial, and the bias-correction improvement in drought characteristics is minimal. Hence, although bias correction is essential for the correction of hydrological variables, it is less essential when assessing drought characteristics obtained using standardisation indices. The study results have applications for improving the simulation of hydrological variables and droughts in the Western Cape.