Search

Study of the ground clutter problem at the Norman site indicates that clutter cross sections of 40 to 50 dB below 1 sq. m/sq. m are typical. Doppler spectrum width, Tau sub c, of the clutter signal is between 0.1 and 1 m/s with the mean value of 0.25 m/s at an antenna rotation rate of 10 deg/s. The design of clutter filters is very much dependent on this width and for proper cancellation the passband cutoff velocity v sub p must be larger or equal to 4.8 Tau sub c. Several canceling schemes are investigated, and the most efficient one is with a third order recursive filter. This filter achieves a 50 dB rejection in the stop band with total annihilation of DC. One dB ripple is in the passband, and the ratio of passband cutoff v sub p to stop band cutoff v sub s is about 3.5. The filter operates best in steady state, but it can also be made to operate in transient by properly initializing its memory elements. Performance on 64 simulated and real time samples show that no more than 10 dB loss in effective canceling over steady state occurs if the filter is initialized. To minimize weather signal loss, canceling is recommended only in regions that contain clutter. Therefore, the range extent over which canceling is performed should vary with elevation. Also the antenna rotation at the two lowest angles should be slower.

An algorithm to automatically detect mesocyclonic shear from radial velocity data of a Doppler radar is developed. The algorithm searches for azimuthally increasing velocities. Runs of increasing velocities are stored and the resulting shears are compared to mesocyclonic shear thresholds. Also, products of velocity differences are multiplied with the corresponding arc lengths and then compared to momentum thresholds. Those runs that pass a set of such thresholds and are spatially close are identified as belonging to a region of mesocyclonic shears or a shear line. The algorithm requires simultaneous storage and operation on two consecutive radials of data, and thus is suitable for real time implementation. Flow charts and performance on two data sets are presented.

Effects of the ground clutter ring in the second trip area for velocity estimation are investigated. Besides unfavorable effects on weather signal due to range square advantage, the clutter poses a new problem when interlaced samples for assigning correct ranges to velocity estimates are used. The problem arises because signals that generate reflectivity estimates for power comparisons cannot be filtered in the same manner as signals from which velocity estimates are generated. Therefore, comparison of powers in the reflectivity and velocity channels is required. Performance of a third order recursive elliptic filter is analyzed. The filter operates best in steady state, but it can also be made to operate in transient by properly initializing its memory elements. Performance on 8 simulated time samples shows that about 10 dB of clutter to signal margin is lost for power and 20 dB for mean velocity estimation; with such a small number of samples, the width estimate is useless. (Author)

This report presents the methodology and results of a study to estimate the lifetimes of significant features in typical storm phenomena. These results are expected to help in deciding the scan strategy of NEXRAD radars. In particular, the question of the adequacy of a 5-minute information update rate for NEXRAD in its aviation weather surveillance role has been addressed. Two methods are used for the lifetime study: photo-interpretive and computer-based correlation. The basis and program steps of the correlation method are discussed. Several storms are studied using both methods. It is found that the storms studied contain no feature that might have been missed by a 5-minute scan cycle provided that data from all the three-moment fields (reflectivity, radial velocity, and Doppler spectrum width) at several elevations are utilized in the detection of hazardous phenomena. (Author)