5:15 PM - 6:45 PM
[AAS08-P08] Toward investigation of a Potential Impact of an Atmospheric Observation at Upstream Area of Heavy Rainfall by Compact IoT Instruments Network
★Invited Papers
Keywords:Observing System Simulation Experiment (OSSE), Water vapor observation, Heavy rainfall, IoT sensor
In recent years, heavy rainfall events have been increasing, and they often cause hazardous disasters to the human lives. To prevent and mitigate them, there is a need for accurate and early weather forecast information, and various national efforts are underway [2, 3]. Especially in Kyushu region, a wide network of radar-based rain cloud observations and water vapor observations using radio wave delays has been established. These systems observe the formation processes of rain clouds and the inflow of water vapor amount from sea to land areas [4].
However, conventional direct oceanic observational stations are installed on land or ships, thus they lack extensive and steady coverage. Therefore, our aim is to construct an expansive and continuous direct observation infrastructure in sea areas using small IoT sensors (referred to as NTT observation). This will contribute to the forecasting of heavy rainfall events using our observation data [5]. Nevertheless, an actual construction requires a huge economical cost. Therefore, it is crucial to construct efficiently. This can be achieved by evaluating the impact of the NTT observations on forecasting heavy rainfall events. It is also important to provide specifications for sensor requirements such as the areas of observation, points, and factors.
This research verifies the NTT observation through a series of Observing System Simulation Experiments (OSSEs). This study aims to investigate an impact of NTT observation on heavy rainfall forecasts, building on the previous study [6] showed the impact of a rich Phased Array Weather Radar (PAWR) network covering the entire Kyushu region by OSSE.
The virtual true state, so-called the Nature run, comes from Maejima et al. (2022), which well reproduced the July 2020 heavy rainfall event. The synthetic observation data is created from the Nature run, assuming numerous small IoT sensors deployed in the lower atmosphere on the sea area. The observation area is set as 200 km vertically and 100 km horizontally off the west coast of Kyushu, and below 1000 m in height. The observation factors are limited to every 10-minute water vapor and horizontal wind direction/speed, This observation data is assimilated into the control run in Maejima et al. (2022).
The results are evalutated by the rainfall area such as the root mean square error of precipitation. This is done to determine the extent to which the NTT observation contributes to improve forecast accuracy for the target case. Additionally, we also compare the results with those of the previous study which assimilated the PAWR observations. The NTT observation is expected to improve the accuracy of the inflow of water vapor amount. Moreover, it leads to accurate the forecast of the generation and maintenance processes of the the heavy rainfalls.
[1] Kato T., “Past 45 years’ long-term trend of the occurrence frequency of heavy rainfall events in Japan extracted from three-hourly AMeDAS accumulated precipitation amounts” (in Japanese), Tenki, 2022, volume 69, issue 5, p. 247-252
[2] Japan Meteorological Agency, “Linear Precipitation Zone Forecasting Accuracy Improvement Working Group” (in Japanese), https://www.jma.go.jp/jma/kishou/shingikai/kondankai/senjoukousuitai_WG/senjoukousuitai_WG.html, accessed on February 7, 2023
[3] National Research Institute for Earth Science and Disaster Resilience, “V. Announcement regarding the development of a linear precipitation zone observation and forecasting system” (in Japanese), https://www.nied-sip2.bosai.go.jp/research-and-development/theme_5.html, accessed on February 7, 2023
[4] National Research Institute for Earth Science and Disaster Resilience, “Press release ’Deploying a water vapor observation network for linear precipitation zones - Challenge to improve the accuracy of short-term rainfall forecasting -‘“, https://www.bosai.go.jp/info/press/2022/20220629.html, accessed on February 7, 2023
[5] NTT, “Super-wide area atmosphere and ocean observation technology”, https://www.rd.ntt/se/technology/iot_satellite.html, accessed on February 7, 2023
[6] Maejima Y., and co-authors, “Observing System Simulation Experiments of a Rich Phased Array Weather Radar Network Covering Kyushu for the July 2020 Heavy Rainfall Event”, SOLA, 2022, Volume 18, Pages 25-32
Keywords
Observing System Simulation Experiment (OSSE), Water vapor observation, Heavy rainfall, IoT sensor
However, conventional direct oceanic observational stations are installed on land or ships, thus they lack extensive and steady coverage. Therefore, our aim is to construct an expansive and continuous direct observation infrastructure in sea areas using small IoT sensors (referred to as NTT observation). This will contribute to the forecasting of heavy rainfall events using our observation data [5]. Nevertheless, an actual construction requires a huge economical cost. Therefore, it is crucial to construct efficiently. This can be achieved by evaluating the impact of the NTT observations on forecasting heavy rainfall events. It is also important to provide specifications for sensor requirements such as the areas of observation, points, and factors.
This research verifies the NTT observation through a series of Observing System Simulation Experiments (OSSEs). This study aims to investigate an impact of NTT observation on heavy rainfall forecasts, building on the previous study [6] showed the impact of a rich Phased Array Weather Radar (PAWR) network covering the entire Kyushu region by OSSE.
The virtual true state, so-called the Nature run, comes from Maejima et al. (2022), which well reproduced the July 2020 heavy rainfall event. The synthetic observation data is created from the Nature run, assuming numerous small IoT sensors deployed in the lower atmosphere on the sea area. The observation area is set as 200 km vertically and 100 km horizontally off the west coast of Kyushu, and below 1000 m in height. The observation factors are limited to every 10-minute water vapor and horizontal wind direction/speed, This observation data is assimilated into the control run in Maejima et al. (2022).
The results are evalutated by the rainfall area such as the root mean square error of precipitation. This is done to determine the extent to which the NTT observation contributes to improve forecast accuracy for the target case. Additionally, we also compare the results with those of the previous study which assimilated the PAWR observations. The NTT observation is expected to improve the accuracy of the inflow of water vapor amount. Moreover, it leads to accurate the forecast of the generation and maintenance processes of the the heavy rainfalls.
[1] Kato T., “Past 45 years’ long-term trend of the occurrence frequency of heavy rainfall events in Japan extracted from three-hourly AMeDAS accumulated precipitation amounts” (in Japanese), Tenki, 2022, volume 69, issue 5, p. 247-252
[2] Japan Meteorological Agency, “Linear Precipitation Zone Forecasting Accuracy Improvement Working Group” (in Japanese), https://www.jma.go.jp/jma/kishou/shingikai/kondankai/senjoukousuitai_WG/senjoukousuitai_WG.html, accessed on February 7, 2023
[3] National Research Institute for Earth Science and Disaster Resilience, “V. Announcement regarding the development of a linear precipitation zone observation and forecasting system” (in Japanese), https://www.nied-sip2.bosai.go.jp/research-and-development/theme_5.html, accessed on February 7, 2023
[4] National Research Institute for Earth Science and Disaster Resilience, “Press release ’Deploying a water vapor observation network for linear precipitation zones - Challenge to improve the accuracy of short-term rainfall forecasting -‘“, https://www.bosai.go.jp/info/press/2022/20220629.html, accessed on February 7, 2023
[5] NTT, “Super-wide area atmosphere and ocean observation technology”, https://www.rd.ntt/se/technology/iot_satellite.html, accessed on February 7, 2023
[6] Maejima Y., and co-authors, “Observing System Simulation Experiments of a Rich Phased Array Weather Radar Network Covering Kyushu for the July 2020 Heavy Rainfall Event”, SOLA, 2022, Volume 18, Pages 25-32
Keywords
Observing System Simulation Experiment (OSSE), Water vapor observation, Heavy rainfall, IoT sensor