Abstract

Gridded runoff product at the sub-basin level is pivotal for effective hydrologic modeling and applications. Although reanalyses can overcome the lack of traditional stream gauge networks to provide reliable geospatial runoff data, the inherent uncertainties associated with single products are still a problem. This study aims to improve the single products’ limitations over the heterogeneous Republic of Korea region by merging three common global reanalysis datasets to generate a high-quality and long-term gridded runoff product at a high resolution. The merging method relies on triple collocation (TC) analysis, which requires no reference runoff dataset, with a modification that was applied separately to wet and dry seasons (seasonal merging). A comparison between the merged runoff and its parent products at 0.10° grid, on a daily basis, and using the entire 10-year period (2011–2020) against an independent ground-based sub-basin runoff product generally indicated a superior performance of the merged product even at the national scale of Republic of Korea. Moreover, a slight improvement obtained with the seasonal merging compared to the traditional all-time merging highlighted the potential of this modification to address several drawbacks in the TC assumption, especially the non-stationary runoff pattern caused by seasonal rainfall effects in the Republic of Korea. Despite the need for further improvement such as bias correction, the results of this study encourage making a reliable benchmark runoff product at a regional scale, which is beneficial for flood/drought monitoring and artificial intelligence-based hydrologic model training.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call