• Journal of Korea Water Resources Association
• /
• v.56 no.9
• /
• pp.563-574
• /
• 2023

Recently, climate change has affected functional responses of river basins to meteorological variables, emphasizing the importance of rainfall-runoff simulation research. Simultaneously, the growing interest in machine learning has led to its increased application in hydrological studies. However, it is not yet clear whether machine learning models are more advantageous than the conventional conceptual models. In this study, we compared the performance of the conventional GR6J model with the machine learning-based Random Forest model across 38 basins in Korea using both gauged and ungauged basin prediction methods. For gauged basin predictions, each model was calibrated or trained using observed daily runoff data, and their performance was evaluted over a separate validation period. Subsequently, ungauged basin simulations were evaluated using proximity-based parameter regionalization with Leave-One-Out Cross-Validation (LOOCV). In gauged basins, the Random Forest consistently outperformed the GR6J, exhibiting superiority across basins regardless of whether they had strong or weak rainfall-runoff correlations. This suggest that the inherent data-driven training structures of machine learning models, in contrast to the conceptual models, offer distinct advantages in data-rich scenarios. However, the advantages of the machine-learning algorithm were not replicated in ungauged basin predictions, resulting in a lower performance than that of the GR6J. In conclusion, this study suggests that while the Random Forest model showed enhanced performance in trained locations, the existing GR6J model may be a better choice for prediction in ungagued basins.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2023.05a
• /
• pp.29-29
• /
• 2023

To mitigate the damaging impacts of floods, accurate prediction of runoff, streamflow and flood inundation is needed. Conventional approach of simulating hydrology and hydraulics using loosely coupled models cannot capture the complex dynamics of surface and sub-surface processes. Additionally, the scarcity of data in ungauged basins and quality of data in gauged basins add uncertainty to model predictions, which need to be quantified. In this presentation, first the role of integrated modeling on creating accurate flood simulations and inundation maps will be presented with specific focus on urban environments. Next, the use of machine learning in producing streamflow predictions will be presented with specific focus on incorporating covariate shift and the application of theory guided machine learning. Finally, a framework to quantify the uncertainty in flood models using Hierarchical Bayesian Modeling Averaging will be presented. Overall, this presentation will highlight that creating accurate information on flood magnitude and extent requires innovation and advancement in different aspects related to hydrologic predictions.

• Journal of the Korean Association of Geographic Information Studies
• /
• v.11 no.3
• /
• pp.52-67
• /
• 2008

Recently, frequent occurrence of flash floods caused by climactic change has necessitated prompt and quantitative prediction of precipitation. In particular, the usability of rainfall radar that can carry out real-time observation and prediction of precipitation behavior has increased. Moreover, the use of distributed hydrological model that enables grid level analysis has increased for an efficient use of rainfall radar that provides grid data at 1km resolution. The use of distributed hydrologic model necessitates grid-type spatial data about target basins; to enhance reliability of flood runoff simulation, the use of visible and precise data is necessary. In this paper, physically based $Vflo^{TM}$ model and ModClark, a quasi-distributed hydrological model, were used to carry out flood runoff simulation and comparison of simulation results with data from Imjin River Basin, two-third of which is ungauged. The spatial scope of this study was divided into the whole Imjin River basin area, which includes ungauged area, and Imjin River basin area in South Korea for which relatively accurate and visible data are available. Peak flow and lag time outputs from the two simulations of each region were compared to analyze the impact of uncertainty in topographical parameters and soil parameters on flood runoff simulation and to propose effective methods for flood runoff simulation in ungauged regions.

• Journal of Korea Water Resources Association
• /
• v.57 no.3
• /
• pp.195-207
• /
• 2024

This study selected major drought events that occurred in the Jeonnam region from 1991 to 2023, examining both meteorological and hydrological drought occurrence mechanisms. The daily drought index was calculated using rainfall and dam storage as input data, and the drought propagation characteristics from meteorological drought to hydrological drought were analyzed. The characteristics of the 2022-23 drought, which recently occurred in the Jeonnam region and caused serious damage, were evaluated. Compared to historical droughts, the duration of the hydrological drought for 2022-2023 lasted 334 days, the second longest after 2017-2018, the drought severity was evaluated as the most severe at -1.76. As a result of a linked analysis of SPI (StandQardized Precipitation Index), and SRSI (Standardized Reservoir Storage Index), it is possible to suggest a proactive utilization for SPI(6) to respond to hydrological drought. Furthermore, by confirming the similarity between SRSI and SPI(12) in long-term drought monitoring, the applicability of SPI(12) to hydrological drought monitoring in ungauged basins was also confirmed. Through this study, it was confirmed that the long-term dryness that occurs during the summer rainy season can transition into a serious level of hydrological drought. Therefore, for preemptive drought response, it is necessary to use real-time monitoring results of various drought indices and understand the propagation phenomenon from meteorological-agricultural-hydrological drought to secure a sufficient drought response period.