Publications

As new observation systems are developed and deployed, new and presumably more precise information is becoming available for weather forecasting and climate monitoring. To take advantage of these new observations, it is desirable to have schemes to accurately retrieve the information before statistical analyses are performed so that statistical computation can be more effectively used where it is needed most. The authors propose a sequential variational approach that possesses advantages of both a standard statistical analysis [such as with a three-dimensional variational data assimilation (3DVAR) or Kalman filter] and a traditional objective analysis (such as the Barnes analysis). The sequential variational analysis is multiscale, inhomogeneous, anisotropic, and temporally consistent, as shown by an idealized test case and observational datasets in this study. The real data cases include applications in two-dimensional and three-dimensional space and time for storm outflow boundary detection (surface application) and hurricane data assimilation (three-dimensional space application). Implemented using a multigrid technique, this sequential variational approach is a very efficient data assimilation method.

One of the largest sources of error in the current automated convective weather forecast systems is due to its inability to accurately account for new convective storm development. In many situations the initiation of new convection is preceded by low altitude convergence in the horizontal winds. These regions of low altitude convergence, often referred to as boundaries, are typically associated with synoptic scale fronts, drylines, and thunderstorm outflows. Gridded wind analyses that utilize Doppler weather radar, surface, and aircraft measurements are one of the best sources of low altitude winds that can be used to identify wind boundaries over large domains. This study summarizes the preliminary results of a study which examined the feasibility of using gridded wind analyses from operational wind analysis systems to make automated detections of wind boundaries. The analysis focused on two operational wind analysis systems both capable of producing high update, and high spatial resolution wind analyses over a domain that covers the eastern half of the Continental United Sates (CONUS), the Space Time Mesoscale Analysis System (STMAS) and the Corridor Boundary layer wind analysis system (CBOUND). Wind analyses from both systems were first processed with a Lagrangian temporal filter and then passed through an automated boundary detection algorithm based on the Terminal Doppler Weather Radar (TDWR) Machine Intelligent Gust Front Algorithm (MIGFA). The results indicate that the temporal filter improves the boundary signal to noise ratio such that it is technically feasible to make fully automated boundary detections with image processing techniques.

As part of the Aviation Weather Development Program of the Federal Aviation Administration, a high resolution winds analysis system was demonstrated at Orlando International Airport (MCO) in the summer of 1992. The purpose of this demonstration was to illustrate the winds analysis capability possible from operational sensors in the mid '90s. An important part of the design of this system was the development of a procedure for the assimilation of Doppler data from multiple radars. This procedure had to be able to automatically handle regions with missing data from one or more radars, as well as avoid baseline instability. The two operational radars scanning the analysis region were the National Weather Service WSR-88D (NEXRAD) radar located approximately 65 km east and slightly south of MCO, and the MIT prototype Terminal Doppler Weather Radar (TDWR) located 7 km due south of the airport. The base data from these two Doppler radars were the major information component for the analysis system. Our system includes the most recent improvements in the winds analysis portion of the Local Analysis and Prediction System (LAPS) developed by the Forecast Systems Laboratory (McGinely et al., 1991). LAPS is designed to run locally on systems affordable for operational weather offices and takes advantages of all sources of local data at the highest possible resolution. Our implementation for the airport terminal region id called the Terminal-area LAPS (T-LAPS). LAPS formerly had a technique for the assimilation of data from a single Doppler radar. We have modified that technique for the assimilation of data from the two available radars. Our approach, using a Multiple Single Doppler Analysis (MSDA) technique, is more suited for unsupervised operational analysis than traditional Dual Doppler Analysis (DDA), because it is able to handle such problems as incomplete data and baseline instability. We will describe the T-LAPS analysis, with particular attention to our implementation of ASDA, and give some examples from our demonstration.