ABSTRACT
Accurate and reliable hydrological forecasts play a pivotal role in ensuring water security, facilitating flood preparedness, and supporting agriculture activities. This study investigates the potential of hydrological forecasting in South Korea, focusing on medium-range lead times ranging from 1 to 10 days. The methodology involves leveraging a Transformer neural network, a model entirely based on attention mechanisms. Specifically, our study introduces the Dualformer, a dual-encoder-based transformer model capable of accommodating two distinct datasets: historical and forecast meteorological data. The performance of this proposed model, along with its variants designed to test specific structural aspects, is evaluated in predicting daily streamflow across 473 grid cells covering extensive regions within the study area. Furthermore, the proposed model is assessed against the performance of a recently developed approach aiming for the same objective. These models are trained using historical meteorological variables and geographic characteristics, alongside the Global Ensemble Forecast System, version 12 (GEFSv12) reforecasts, in addition to historical runoff. The results indicate that our proposed model performs competitively, especially for relatively short lead times while effectively managing information from two distinct data sources. For instance, the mean Nash-Sutcliffe efficiency for 473 grids is 0.664 for the first one-day lead when using the Dualformer, whereas the benchmark model achieves a score of 0.535. Additionally, we observe an additional enhancement in Dualformer's performance when utilizing a larger dataset. Finally, we conclude this paper with a discussion regarding potential improvements to the forecast model through the incorporation of additional input and modeling structures.
ABSTRACT
Heat waves can provide detrimental impacts on human society and the environmental system, and thus have received substantial attention in scientific research. Since heat waves are relevant to a wide range of stakeholders, definitions for heat wave events vary in terms of threshold values, durations, and utilized variables. While there is a value in this diversity of perspectives, the various definitions often complicate the assessment of heat wave risk, thereby underscoring the improved utility of a unified definition. In this study, we examine the interannual variability of heat wave patterns by using a proposed copula-based framework. From five observed temperature-related variables, this study first evaluates the individual changes of fifteen previously published heat wave indices focused on heat wave events across the Korean Peninsula for the last 49 years (1973-2021). We then extract the integrated signals to understand the overall trend patterns using the multiple heat wave indices. Results indicate that different heat wave definitions present distinctive attributes (e.g., in the magnitude of temporal changes) depending on the locations, implying that the diversity of heat wave definitions leads to potentially inconsistent conclusions. Using the integrated analysis, we identify that the expected heat wave day has increased across the majority of the regions in the Korean Peninsula. To be specific, substantial increases are shown in North Korea, while rapid increases in heat wave events have been observed after the year 2010 over South Korea. Finally, through the sensitivity analysis, we demonstrate the importance of choosing the heat wave definition in the integrated analysis.
