3:45 PM - 4:00 PM
[ACG42-08] Toward an improvement of river runoff in the Japanese Coastal Ocean Monitoring and Forecasting System
★Invited Papers
Keywords:coastal model, operational oceanograpy, river runoff, Japanese coastal seas
The Japan Meteorological Agency (JMA) started the operation of the "Japanese Coastal Ocean Monitoring and Forecasting System" (JPN System) in October 2020 for the purpose of grasping ocean states around Japan in detail. This system was developed by the Meteorological Research Institute (MRI) and consists of an OGCM that covers a wide area around Japan with a horizontal resolution of about 2 km (JPN model) and a four-dimensional variational data assimilation system. The JPN model incorporates major coastal processes such as tidal and pressure response, enabling ocean simulations that can be directly compared to the real ocean. With the start of operation of the JPN system, new marine information such as detailed coastal current fields are being provided by the JMA, and we hope that it will be widely utilized by the ocean community across industry, government, and academia. The JPN quick analysis and prediction data can be obtained from the Japan Meteorological Business Support Center, and, in addition, a long-term reanalysis data from 2008 is also being provided by the MRI. (https://mri-ocean.github.io/mricom/mri.com-user_jpn_start.html)
Although the JPN system has incorporated major coastal processes, there is still room for improvement in the system specifications. One of them is that daily climatology of Japanese first-class rivers is used for river runoff in the model, not taking the actual time variations into consideration. As a result, low-salinity plumes formed in estuaries during freshets due to heavy rain cannot be reproduced by the JPN system. Therefore, we conducted an experiment to introduce real-time river runoff data for the purpose of improving the JPN system in the future. Specifically, following Urakawa et al. (2016, Weather Service Bulletin), we made an hourly runoff dataset of 3986 rivers nationwide, based on the JMA Runoff Index, which has been used for flood warnings by the JMA, and we ran the JPN model under the new dataset. First, we executed a long-term free-run experiment, and confirmed that this dataset realistically reproduces seasonal changes in sea surface salinity in the Seto Inland Sea. Next, a hindcast experiment was conducted using an initial field by data assimilation, in order to reproduce coastal states after a typhoon passed in September 2011. In the inner part of the Osaka Bay, a low-salinity plume developed due to a freshet of the Yodo River, and surface salinity decreased by more than 5 psu there. This result is consistent with salinity observations along the coast of the Osaka Bay. These results indicate the possibility of improving river runoff in the JPN system. Now, we are planning an experiment to reproduce the low-salinity plume at the time of freshet using an Osaka Bay model, which is downscaled to a resolution of several hundred meters.
Although the JPN system has incorporated major coastal processes, there is still room for improvement in the system specifications. One of them is that daily climatology of Japanese first-class rivers is used for river runoff in the model, not taking the actual time variations into consideration. As a result, low-salinity plumes formed in estuaries during freshets due to heavy rain cannot be reproduced by the JPN system. Therefore, we conducted an experiment to introduce real-time river runoff data for the purpose of improving the JPN system in the future. Specifically, following Urakawa et al. (2016, Weather Service Bulletin), we made an hourly runoff dataset of 3986 rivers nationwide, based on the JMA Runoff Index, which has been used for flood warnings by the JMA, and we ran the JPN model under the new dataset. First, we executed a long-term free-run experiment, and confirmed that this dataset realistically reproduces seasonal changes in sea surface salinity in the Seto Inland Sea. Next, a hindcast experiment was conducted using an initial field by data assimilation, in order to reproduce coastal states after a typhoon passed in September 2011. In the inner part of the Osaka Bay, a low-salinity plume developed due to a freshet of the Yodo River, and surface salinity decreased by more than 5 psu there. This result is consistent with salinity observations along the coast of the Osaka Bay. These results indicate the possibility of improving river runoff in the JPN system. Now, we are planning an experiment to reproduce the low-salinity plume at the time of freshet using an Osaka Bay model, which is downscaled to a resolution of several hundred meters.