Assessing the numerical weather prediction (NWP) model in estimating extreme rainfall events: A case study for severe floods in the southwest Mediterranean region, Turkey

• Published in: Journal of Earth System Science 2023-07, Vol.132 (3), p.125, Article 125
• Main Author: Ozkaya, Arzu
• Format: Article
• Language: English
• Subjects: Accuracy; Correlation; Datasets; Earth and Environmental Science; Earth Sciences; Emergency communications systems; Emergency warning programs; Extreme weather; Flood forecasting; Flood predictions; Flood warning systems; Floods; Gauges; Mathematical models; Numerical weather forecasting; Precipitation; Predictions; Rain; Rain gauges; Rainfall; Rainfall correlations; Rainfall distribution; Severe floods; Space Exploration and Astronautics; Space Sciences (including Extraterrestrial Physics; Warning systems; Weather; Weather forecasting
• ISSN: 0973-774X; 0253-4126; 0973-774X
• DOI: 10.1007/s12040-023-02137-7

In flood warning systems, numerical weather predictions (NWP) are important complementary tools as they increase the forecast lead times required to provide timely warnings. However, these predictions may include non-predictable uncertainties that decrease the correlation with gauge observations and ascend the number of missed or false warnings. Investigation of the potential of such products that are generated one-day apart can reveal the success of lead times in hydrological applications. In this study, two periodically generated products of the Weather Research and Forecasting (WRF) model are assessed by considering 15 flood events experienced in the southwest Mediterranean region of Turkey with a network of 26 rain gauges. The general results show that the rainfall distribution of both WRF datasets is found to be similar with slight differences in magnitude. However, the rainfall product generated earlier shows greater agreement in the observational data for 1-hr interval time but the latter one shows less bias and less cumulative error as interval time increases. In season- and elevation-based analyses, both WRF datasets show the highest correlation value in the winter season. With this study, it is revealed that the rainfall correlations results are superior in the former data whereas the rainfall accumulations are better represented with the posterior data.