Graduate Exam Abstract
Matthew MartinezM.S. Final
February 16, 2011, 3:00 pm
Description and Evaluation of the CASA Dual-Doppler System
Abstract: Long range weather surveillance radars are designed for observing weather events for hundreds of kilometers from the radar and operate over a large coverage domain independently of weather conditions. As a result a loss in spatial resolution and limited temporal sampling of the weather phenomenon occurs. Due to the curvature of the Earth, long-range weather radars tend to make the majority of their precipitation and wind observations in the middle to upper troposphere, resulting in missed features associates with severe weather occurring in the lowest three kilometers of the troposphere. The spacing of long-range weather radars in the United States limits the feasibility of using dual-Doppler wind retrievals that would provide valuable information on the kinematics of weather events to end-users and researchers.
The National Science Foundation Center for Collaborative Adapting Sensing of the Atmosphere (CASA) aims to change the current weather sensing model by increasing coverage of the lowest three kilometers of the troposphere by using densely spaced networked short-range weather radars. CASA has deployed a network of these radars in south-western Oklahoma, known as Integrated Project 1 (IP1). The individual radars are adaptively steered by an automated system known as the Meteorological Command and Control (MCC). The geometry of the IP1 network is such that the coverage domain of the individual radars in the network are intertwined.
A dual-Doppler system has been developed for the IP1 network which takes advantage of the overlapping coverage domains. The system is comprised of two subsystems, scan optimization and wind field retrieval. The scan strategy subsystem uses the DCAS model and the number of dual-Doppler pairs in the IP1 network to minimizes the normalized standard deviation in the wind field retrieval. The scan strategy subsystem also minimizes the synchronization error between two radars. The retrieval itself is comprised of two steps, data resampling and the retrieval process. The resampling step map data collected in radar coordinates to a common Cartesian grid. The retrieval process uses the radial velocity measurements to estimate the northward, eastward, and vertical component of the wind. The error in the retrieval is related to the beam crossing angle. The best retrievals occur at beam crossing angles greater than 30 degrees.
During operations statistics on the scan strategy and wind field retrievals are collected in real-time. For the scan strategy subsystem statistics on the beam crossing angels, maximum elevation angle, number of elevation angles, maximum observable height, and synchronization time between radars in a pair are collected by the MCC. These statistics are used to evaluate the performance of the scan strategy subsystem. Observations of a strong wind event occurring on April 2, 2010 are used to evaluate the decision process associated with the scan strategy optimization. For the retrieval subsystem, the normalized standard deviation for the wind field retrieval is used to evaluate the quality of the retrieval. Wind fields from an EF2 tornado observed on May 14, 2009 are used to evaluate the quality of the wind field retrievals in hazardous wind events.
Two techniques for visualizing vector fields are available, streamlines and arrows. Streamlines are a family of curves that are instantaneously tangent to the vector field. Arrows are derived from the magnitude and direction of the vector. Each visualization technique is evaluated based on the task of visualizing small and large scale phenomenon. Streamlines are favored over arrows when visualizing small scale phenomenon, where as arrows are preferred over streamlines for large scale phenomenon.
Applications of the wind field retrievals include the computation of the vorticity and divergence fields. Vorticity describes the amount of rotation a fluid has about a central axis. In the northern hemisphere cyclonic rotation is in the counterclockwise direction, and is associated with positive vorticity. On the other hand, anti-cyclonic rotation is in the clockwise directions and is associated with negative vorticity. Divergence is the measurement of how a vector field behaves like a sink or a source at any given point. In the atmosphere divergence is the outflow of air. Vorticity and divergence for an EF2 tornado observed on May 14, 2009 are evaluated against vorticity and divergence for other observed tornadoes.
Adviser: Dr. V Chandrasekar
Non-ECE Member: Dr. Paul W. Mielke, Jr. Statistics
Member 3: Dr. Branislav Notaros
Addional Members: N/A,N/A
Chandrasekar, V., M. Martinez, and S. Zhang, 2010: The CASA dual-doppler sys-
tem. Proc. IEEE IGARSS 2010., IEEE, Honolulu, Hawaii, 4138-4141.
Chandrasekar, V., Y. Wang, N. Bharadwaj, S. Zhang, M. Martinez, D. Mclaughlin, J. Brotzge, M. Zink, and B. Phillips, 2010b: Observation of an ef2 tornado winds using networked radar test-bed. Proc. IEEE Radar Conference 2010., IEEE,Washington,
Program of Study: