Study on the precision of 1-minute X-Band MP radar rainfall data in a small urban watershed

Abstract

In Tokyo metropolitan area, the flood risk is increasing due to social and environmental conditions including concentration of population and industry etc. In small urban watersheds, there exists a high risk of inundation by river flooding and/or inner water induced by heavy rainfall in a short time. To estimate river water level accurately in urban small rivers, it is critically important to obtain accurate rainfall data with high resolution of spatio-temporal distribution. X-band MP radar (X-band multiparameter radar), providing high resolution of spatio-temporal distribution rainfall data, is expected to be made effective use of its advanced detailed data. However, few studies on the precision evaluation of 1-minute X-band MP radar data were performed. In this study, a precision evaluation of X-band MP radar data was conducted by comparing radar data to 1-minute ground observation data, which is densely deployed in Tokyo Metropolitan area, provided by Tokyo Integrated Information System on Water Disaster Management. In this study, 10 heavy rainfall events in 2013 were selected as target events in the upper Kanda River, one of the typical urban small rivers in Tokyo. As for rainfall intensity, 1-minute data of X-band MP radar expressed a better peak accuracy compared with 10-minute data. The peak values of 10-minute data were underestimated about 30% in some event. One-minute data of X-band MP radar were observed a few minutes earlier than ground observation data. The falling time of rain drops is considered as the major factor of the difference. This study showed that it is possible to obtain detailed rainfall characteristics such as precise rainfall intensity and its time of occurrence by utilizing 1-minute X-band MP radar rainfall, which are not sufficiently provided by 10-minute data and its extended resolution is anticipated to be practically useful in smaller urban watersheds.

Point-Based Rainfall Intensity Information System in Mt. Merapi Area by X-Band Radar

“…Because of this reason, the original radar data could not be compared with the rain gauge data directly. The tendency of spatial rainfall data overestimation at specific conditions had been discussed in other studies too [6,14,21]. Those studies indicated that spatial rainfall data may present spatial rainfall pattern properly.…”

“…As an example, scenario R/15 means that the radar data were divided by 15. In this scenario, if the original rainfall intensity recorded by radar is 75 mm/h, the adjusted rainfall value will be 5 mm/h. In statistical analysis, the divisor is equivalent to index of systematic error because it represent ratio between estimated rainfall value (radar data) and observed rainfall value (rain gauge data) [6,14,21]. The divisor adjusted radar recorded rainfall intensity but maintained its spatial pattern.…”

“…Rainfall radar is typically used when the ground rain gauge data cannot be quickly transferred to the monitoring office during a heavy rainfall event [5]. Radar can detect rainfall events at a higher temporal resolution of up to 60 s measurement interval [5,6]. Radar rainfall data are commonly used as bases for validating weather forecast models because they contain not only rainfall intensities, but also rainfall distribution profiles [7].…”

“…Merapi eruption [2]. Nowadays, researchers started to explore the applicability of radar data for early warning system, although radar data are still generally low-resolution [5,6]. The radar data provided by the Indonesian Meteorological, Climatological, and Geophysical Agency (BMKG), as presented in Fig.…”

Urban Flood Runoff Modeling in Japan: Recent Developments and Future Prospects

“…For localized torrential rainfall events in the upper Kanda River watershed (Figure 1), the USF model has also been used to evaluate hydrograph reproducibility using XRAIN rainfall data and the FCIS rain gauge data [99]. At first, 1-minute XRAIN mesh rainfall data were compared to 1-minute FCIS rain gauge data of the same mesh to check the precision of XRAIN [100,101]. The results showed that 1-minute XRAIN values corresponded well with the 1-minute ground observation rainfall values, but the former were observed 1 to 3 minutes earlier than the ground observation, which was considered mainly due to the falling time of the raindrops.…”