11:25 AM - 11:40 AM
[ACG48-08] River Discharge Analysis Using the Japan Ocean-River-RunOff Model (JORRO) for the Entire Shirakawa River Basin and Comparison with a Hydrological Model
Keywords:hydrological model, ocean model, Shirakawa
The analysis of water dynamics in rivers and oceans is essential for assessing ecosystems in brackish and coastal areas. Conventional analyses have employed a combination of river models and ocean models; however, they have not adequately captured the flow dynamics at the boundary region. The ocean-river-runoff integrated layer model “JORRO” used in this study is characterized by its ability to analyze the flow of land and ocean as a unified system. However, the accuracy of land-based analysis has not been sufficiently validated, and the effects of land use and groundwater are not considered. Therefore, this presentation evaluates the characteristics of JORRO’s flow analysis results by comparing them with the hydrological model CDRMV3.3.1 and observational data.
JORRO adopts a model where freshwater and seawater are treated as separate layers based on the ocean layer thickness model, with a single-layer model for land areas and a two-layer model for ocean areas. On the other hand, CDRMV3.3.1 is a hydrological model that divides soil into three layers and considers interflow, improving the reproducibility of flow rates through calibration. However, since it relies on parameter adjustments for optimization, its physical interpretability is challenging, and its application to different environments is constrained. In contrast, JORRO’s simple structure facilitates its application to other regions and enables logical evaluation.
This study examines the runoff processes during the heavy rainfall event in September 2015 in the Shirakawa River basin in Aso, Kumamoto Prefecture, a region that includes diverse physical processes such as flatlands, steep slopes, and meandering rivers, making it suitable for evaluating the importance of each process. The JORRO analysis used 30m resolution digital elevation data, but when integrated with river network data, unnatural reservoirs were generated, causing unintended flow divergence. To resolve this, the terrain data were adjusted to ensure a consistent water flow from the upper to the lower reaches of the Shirakawa River.
Flow analysis was conducted at three locations: upstream at Tateno, midstream at Jinnai, and downstream at Yotsugibashi. While the time lag between the JORRO model and observed peaks was only a few hours at Tateno and Jinnai, a difference of about one day was observed at Yotsugibashi, with greater deviations downstream. In terms of absolute flow values, the initial peak at Tateno was underestimated, while the second peak at Jinnai tended to be overestimated. Observational data showed that the flow at Jinnai was lower than at Tateno, but in JORRO, the flow was reversed. This discrepancy may be attributed to the fact that JORRO does not account for groundwater effects and that the adjusted terrain fixes the water flow paths. Other factors, such as geology and land use, may also play a role. To improve reproducibility, the model needs to be enhanced by incorporating groundwater dynamics.
The CDRMV3.3.1 analysis showed some differences in peak flow values, but it more accurately reproduced flow changes at Yotsugibashi and captured the pattern where Jinnai’s flow was lower than Tateno’s, unlike JORRO. This is because CDRMV3.3.1 is calibrated to match actual flow data. However, since parameters are adjusted for each observation point, the accuracy is not necessarily consistent across the entire basin. In contrast, JORRO considers physical characteristics such as changes in Froude numbers along flow paths, and by incorporating groundwater effects, it has the potential to reproduce realistic flows without calibration. Future improvements may further develop JORRO into a more versatile hydrodynamic model.
JORRO adopts a model where freshwater and seawater are treated as separate layers based on the ocean layer thickness model, with a single-layer model for land areas and a two-layer model for ocean areas. On the other hand, CDRMV3.3.1 is a hydrological model that divides soil into three layers and considers interflow, improving the reproducibility of flow rates through calibration. However, since it relies on parameter adjustments for optimization, its physical interpretability is challenging, and its application to different environments is constrained. In contrast, JORRO’s simple structure facilitates its application to other regions and enables logical evaluation.
This study examines the runoff processes during the heavy rainfall event in September 2015 in the Shirakawa River basin in Aso, Kumamoto Prefecture, a region that includes diverse physical processes such as flatlands, steep slopes, and meandering rivers, making it suitable for evaluating the importance of each process. The JORRO analysis used 30m resolution digital elevation data, but when integrated with river network data, unnatural reservoirs were generated, causing unintended flow divergence. To resolve this, the terrain data were adjusted to ensure a consistent water flow from the upper to the lower reaches of the Shirakawa River.
Flow analysis was conducted at three locations: upstream at Tateno, midstream at Jinnai, and downstream at Yotsugibashi. While the time lag between the JORRO model and observed peaks was only a few hours at Tateno and Jinnai, a difference of about one day was observed at Yotsugibashi, with greater deviations downstream. In terms of absolute flow values, the initial peak at Tateno was underestimated, while the second peak at Jinnai tended to be overestimated. Observational data showed that the flow at Jinnai was lower than at Tateno, but in JORRO, the flow was reversed. This discrepancy may be attributed to the fact that JORRO does not account for groundwater effects and that the adjusted terrain fixes the water flow paths. Other factors, such as geology and land use, may also play a role. To improve reproducibility, the model needs to be enhanced by incorporating groundwater dynamics.
The CDRMV3.3.1 analysis showed some differences in peak flow values, but it more accurately reproduced flow changes at Yotsugibashi and captured the pattern where Jinnai’s flow was lower than Tateno’s, unlike JORRO. This is because CDRMV3.3.1 is calibrated to match actual flow data. However, since parameters are adjusted for each observation point, the accuracy is not necessarily consistent across the entire basin. In contrast, JORRO considers physical characteristics such as changes in Froude numbers along flow paths, and by incorporating groundwater effects, it has the potential to reproduce realistic flows without calibration. Future improvements may further develop JORRO into a more versatile hydrodynamic model.