Forecast Models
Phased Array Radar Data Assimilation
An initial effort has been made in assimilating phased array radar data and the goal is to improve numerical analyses and predictions of severe storms. To achieve this goal, a comprehensive approach is being taken to attack problems in three important aspects: (i) velocity dealiasing in radar data quality control, (ii) error covariance estimation, and (iii) radar data assimilation using the estimated error statistics. This approach is demonstrated by applying the upgraded 3.5-dimensional variational (3.5dVar) radar data assimilation package to phased array radar data collected for the squall line on 2 June 2004. First, the three-step radar velocity dealiasing method (Gong et al. 2003) is simplified and integrated into the 3.5-dimensional variational (3.5dVar) radar data assimilation package. The simplified dealiasing performs only two steps: reference check and continuity check. This first step uses reference velocities from the predictions of the Navy's Coupled Ocean/Atmosphere Mesoscale Prediction System (COAMPS). This step is computationally more efficient than the reference check in the three-step method, and the two-step dealiasing is found to be effective (as shown in Fig. 1). Then, since the fast phased array radar scans allow radial-velocity innovation (observation minus forecast background) data to be accumulated rapidly, the innovation data can be used to estimate the radar radial-velocity observation error and background wind error covariances nearly real time by using a newly-reformulated and efficient innovation method. The 3.5dVar radar data assimilation package is upgraded to use the estimated error covariances. This leads to an integrated approach in phased array radar data assimilation. This approach was applied to phased array radar data collected for a squall line on 2 June 2004 to improve COAMPS analyses and predictions. The results are highlighted in Fig. 2, and the details are presented in Xu et al. (2005, 32nd Conference on Radar Meteorology, Amer. Meteor. Soc.).
(a)  |
(b)  |
Fig.1. Phased array radar observed radial velocity (a) and dealiased radial velocity (b) at 21:47 UTC on 2 June 2004. The aliased velocity areas are highlighted by yellow circles in (a). |
|
(a) |
(b) |
Fig. 2. (a) RMS differences between the model-produced (analyzed or predicted) and observed radial-velocity fields. (b) Spatial correlation coefficients between the model-produced and observed reflectivity fields. The red curves are for the COAMPS control run without using radar data. The blue curves are for the test-1 run with three phased array radar volume scans assimilated through one cycle from 2108 to 2112 UTC. The green curves are for the test-3 run with nine phased array radar volume scans assimilated through three cycles from 2108 to 2120 UTC on 2 June 2004. |
|
