1.3
2D Wind Field Estimation with Higher Spatial Resolution Using Dual Compact X-Band Weather Radars
Handout (1.3 MB)
1. Introduction
These days there are soaring numbers of extreme weathers in the world such as torrential rain, tornadoes and so on. And needs for observation of precipitation with high spatial and temporal resolution using dual polarimetric Doppler weather radars are growing higher and higher. Wind observation is recognized as an important part to predict precipitation phenomena. Usage of a Doppler radar is one of possibilities to observe wind directions and speed. VVP method is proposed to estimate the wind direction and speed based on an assumption that we have homogeneous winds in a sample area when one Doppler radar is available [1]. However, this method has possibilities to obtain wrong estimations and no data even though we have precipitation data.
A dual Doppler analysis using 2 Doppler radars is promised to estimate the wind with more accuracy. Furuno has successfully developed a dual-polarimetric Doppler weather radar shown in Fig. 1 [2]. Brief specifications are also summarized in Table 1. The main features of the Doppler weather radar are high precision 3-D meteorological monitoring and the smallest and lightest in market[1]. It is designed to cover 30 km range and it has very high spatial and temporal resolution, 50 m and 30 sec respectively.
In this paper, 2D wind field estimations using dual Doppler weather radars are reported. The obtained results shows usability of the wind field estimation using dual or multiple radar systems compared with usage of a single Doppler radar.
|
2. Optimized dual-Doppler analysis
The Doppler velocity is very useful for analyzing a wind-field. A horizontal wind-field can be estimated using the VVP method with a single Doppler radar. But in this method cross-beam wind cannot be calculated precisely because no Doppler information from the radar is available.
Fig. 2 shows estimated results of the wind-field with one Dual Polarization radar installed in Nishinomiya, Japan. The transmitted pulse is corresponded to 1 msec width using a pulse compression technique. Resolutions of angle and distance of measured data are 0.5 deg and 100 m respectively. In an area having gradual patterns (Area A) along with the azimuth direction, the wind-filed is well estimated. On the other hand, we have some errors in the estimation of the wind-field in other area (Area B) having complicated wind field patterns. In addition, we have much more capabilities to have errors when using the VVP method with narrower grid sizes.
Fig 2 : Estimated wind-field using the VVP at 18:00JST on June 12, 2014.
In order to overcome these issues and obtain corrected wind-fields without estimation errors, dual Doppler analyses are performed. Two dual Polarimetric radars are installed in Kobe and Nishinomiya, respectively (Radar A and Radar B), shown in Fig. 3. Horizontal wind-fields have been successfully calculated using measured data of the two radars with a small grid size of 1 km. In addition, the wind-filed can be estimated in the area where there are some errors when using the VVP methods due to effects of the cross-beam winds. We have also successfully performed good estimations of the wind-field with an extremely narrow grid size of 200 m using the measured data from 2 X-band dual Polarimetric radars with higher spatial and temporal resolutions in Fig. 4.
Fig 3 : Estimated wind-field using dual-Doppler analysis at 18:00JST on June 12, 2014.
Horizontal wind velocity and direction are overlaid on rainfall intensity.
Fig 4 : Estimated wind-field using dual-Doppler analysis with narrower grid size of 200 m
at 18:00JST on June 12, 2014.
4. Conclusions
In this paper, the estimations of the wind-field using dual Doppler analyses have been successfully demonstrated compared with results using the conventional VVP method. Data for the estimations are measured by Furuno's X-band small dual Polarimetric radars with higher spatial resolutions. The demonstration has shown that the dual Doppler analysis has advantages in estimations of the wind-field compared to the conventional VVP. In addition, we've successfully estimated the wind-filed with extremely narrow grid size of 200 m using the newly developed X-band dual Polarimetric weather radar with higher spatial resolutions.
5. Acknowledgement
The authors wish to thank to Prof. Eiichi Nakakita of Kyoto University and Prof. Satoru Oishi of Kobe University for their encouragements and useful discussions.