17:15 〜 19:15
[AAS02-P13] Evaluation of Data Assimilation Using Sea Surface Observation Data during Typhoons with OSSEs
キーワード：現場観測データ、データ同化、観測システムシミュ―レーション実験
Simultaneous observation data of the atmosphere and ocean at the sea surface during a typhoon are crucial for typhoon forecasting. These data are also valuable for understanding the interaction between typhoons and the ocean and for improving the performance of various physical models. The authors have previously used the Wave Glider (WG), an unmanned surface vehicle (USV), to observe typhoons in areas with strong winds and storms [1]-[3]. In previous observation operations, when a typhoon track forecast was announced, the WGs were positioned as close as possible to the forecast track. To clarify the positioning of the WG, which is effective for typhoon forecasting, observation positions will be set based on typhoon forecasts in this experiment.
First, a typhoon coordinate system is defined, as shown in Fig. 1, based on the typhoon's path direction and its positional relationship with the center of the typhoon. Next, the number of observation positions within the strong wind area was divided into stages based on wind characteristics, as shown in Fig. 2. An Observation System Simulation Experiments (OSSEs) are applied in this experiment to compare and evaluate differences in observation conditions, such as positions and parameters, in strong wind areas. The experiment used the Weather Research and Forecasting Model (WRF, version 4.5.1) with WRFDA-3DVAR for data assimilation. The initial conditions for the control run were set to 06 UTC on August 29, 2022, and for the nature run to 18 UTC on August 30, 2022. The data assimilation time was set to 12 UTC on August 31, 2022, and the forecast values at 12 UTC on September 1, 2022 (24 hours later) were used for comparison and evaluation. The horizontal resolution was set to 2 km, with vertical 50 layers. Initial and boundary values were sourced from NCEP GDAS/FNL, and the background error was based on NCEP’s CV3 provided in WRF.
The results of assimilating sea level pressure (SLP) and temperature are shown in Fig. 3a, and wind speeds, wind direction and temperature in Fig. 3b. The only variables in the experiment are the SLP and wind speeds of the assimilated data. It was confirmed that the values were corrected to align more closely with the nature run, depending on the number of data assimilation points. One notable point is that there was no significant difference in the prediction results between conditions (4) and (5) in Fig. 2, despite the difference in the number of observation points. Additionally, there were periods when the results for condition (5) were better.
In this experiment, we evaluated the impact of differences in sea surface observation conditions on typhoon forecast. We confirmed that variations in observation locations and the number of observation points, based on the characteristics of sea surface winds under a typhoon, significantly affect prediction results. The results are intended to guide future observation strategies and enhance observation conditions. In the future, we plan to reduce the number of observation positions to levels closer to actual observation conditions and implement cycling runs over time to further assess the effects.
References:
[1] N. Kosaka, et al., “Synchronous observations of atmosphere and ocean directly under typhoons using autonomous surface vehicles,” SOLA, 19, 116−125, 2023. doi:10.2151/sola.2023-016
[2] N. Kosaka, et al., “Improving air-sea observations of typhoons using wave gliders,” SOLA, 20, 347−356, 2024. doi:10.2151/ sola.2024-046
[3] N. Kosaka, et al., “Sea surface typhoon observations using autonomous surface vehicles in 2024,” IWTRC-2, 2024.
First, a typhoon coordinate system is defined, as shown in Fig. 1, based on the typhoon's path direction and its positional relationship with the center of the typhoon. Next, the number of observation positions within the strong wind area was divided into stages based on wind characteristics, as shown in Fig. 2. An Observation System Simulation Experiments (OSSEs) are applied in this experiment to compare and evaluate differences in observation conditions, such as positions and parameters, in strong wind areas. The experiment used the Weather Research and Forecasting Model (WRF, version 4.5.1) with WRFDA-3DVAR for data assimilation. The initial conditions for the control run were set to 06 UTC on August 29, 2022, and for the nature run to 18 UTC on August 30, 2022. The data assimilation time was set to 12 UTC on August 31, 2022, and the forecast values at 12 UTC on September 1, 2022 (24 hours later) were used for comparison and evaluation. The horizontal resolution was set to 2 km, with vertical 50 layers. Initial and boundary values were sourced from NCEP GDAS/FNL, and the background error was based on NCEP’s CV3 provided in WRF.
The results of assimilating sea level pressure (SLP) and temperature are shown in Fig. 3a, and wind speeds, wind direction and temperature in Fig. 3b. The only variables in the experiment are the SLP and wind speeds of the assimilated data. It was confirmed that the values were corrected to align more closely with the nature run, depending on the number of data assimilation points. One notable point is that there was no significant difference in the prediction results between conditions (4) and (5) in Fig. 2, despite the difference in the number of observation points. Additionally, there were periods when the results for condition (5) were better.
In this experiment, we evaluated the impact of differences in sea surface observation conditions on typhoon forecast. We confirmed that variations in observation locations and the number of observation points, based on the characteristics of sea surface winds under a typhoon, significantly affect prediction results. The results are intended to guide future observation strategies and enhance observation conditions. In the future, we plan to reduce the number of observation positions to levels closer to actual observation conditions and implement cycling runs over time to further assess the effects.
References:
[1] N. Kosaka, et al., “Synchronous observations of atmosphere and ocean directly under typhoons using autonomous surface vehicles,” SOLA, 19, 116−125, 2023. doi:10.2151/sola.2023-016
[2] N. Kosaka, et al., “Improving air-sea observations of typhoons using wave gliders,” SOLA, 20, 347−356, 2024. doi:10.2151/ sola.2024-046
[3] N. Kosaka, et al., “Sea surface typhoon observations using autonomous surface vehicles in 2024,” IWTRC-2, 2024.