ScholarWorks@성균관대학교

초록

Snow is a key component in hydrological cycles and climate change research, and has a significant impact on a multitude of industries and societal domains, including agriculture, energy production, and disaster preparedness. In this study, a Multi-Layer Perceptron (MLP) model was employed for the purpose of correcting Local Data Assimilation and Prediction System (LDAPS) snow depth data for South Korea and evaluating its utility. The MLP model was trained on a combination of LDAPS snow depth, surface temperature, net radiation, and NASA Digital Elevation Model (NASADEM) data. The statistical analysis between Automated Synoptic Observing System (ASOS) ground observation data and LDAPS reanalysis data demonstrated that the MLP model effectively corrected the error of LDAPS, with a correlation coefficient and root mean square error (RMSE) of 0.838 and 0.048 m, respectively. Furthermore, the MLP model effectively represented the spatial distribution of LDAPS snow depths, providing a more accurate reflection of the variations in snow depth resulting from climate change and extreme weather events. The MLP-based snow depth data produced in this study is anticipated to serve as a valuable resource for climate change analysis and disaster preparedness in South Korea. In the future, the incorporation of time series models such as Long Short-Term Memory (LSTM) is expected to enhance the predictive capabilities of the model. © 2024 Korea Water Resources Association. All rights reserved.

키워드