An examination of precipitation variability with respect to frontal boundaries
Date
2006
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
University of Delaware
Abstract
Accurate, real-time estimates of rainfall are critical to many different aspects of society, including emergency management, water resource management, and agriculture. Weather radar is currently the most widely used tool for estimating rainfall in real-time. However, weather radar does not directly measure rainfall. Instead, it measures the surface area of raindrops, which are related to the volume of rain through a logarithmic relationship called the Z-R (Reflectivity-Rainfall Rate) relationship. Several methods have been developed which attempt to improve the estimation of rainfall by examining the statistical relationship between the radar and rain gage networks and adjusting the variables in the Z-R relationship. Although these methods are useful, they fail to examine or utilize the relationship between rainfall and the physical mechanisms that form rainfall. ☐ This study examined the relationship between radar estimated rainfall and surface frontal boundaries using only surface data available in real-time. The frontal boundary plays a physically significant role in the formation of rainfall, since it delineates regions with large differences in energy and moisture. These differences affect the drop-size distribution of the rainfall, and thus the frontal boundary allows for different Z-R relationships to be determined for regions of rainfall separated by a frontal boundary. An objective frontal analysis method was developed using edge detection and pattern recognition techniques to locate the frontal zone in a surface equivalent potential temperature (θe) field. Using the analyzed frontal zones, prefrontal and postfrontal radar estimated rainfall was segmented. ☐ A nonparametric statistic was calculated to examine the similarity between the prefrontal and postfrontal radar estimated rainfall distributions for each event. In cases where there was a large, easily discernable &thetas; e gradient (i.e., strong frontal boundary), the prefrontal and postfrontal radar estimated rainfall distributions were very dissimilar. This relationship demonstrates that the frontal zone is useful in delineating rainfall formed by different precipitation processes. In other cases where the θe gradient was not as large, the segmented radar estimated rainfall distributions were not as dissimilar. ☐ The spatial variability and intensity variability of the segmented radar estimated rainfall distributions were also examined. In general, there were only slight differences between the mean intensity of the prefrontal and postfrontal radar estimated rainfall regions. An average area statistic was slightly more useful in demonstrating large differences in average area of rainfall in each segmented region. Thus areas where rainfall consisted of a few large, continuous regions were moderately well segmented from regions of numerous small areas of rainfall. ☐ In general, the objective frontal analysis method developed in this study performed reasonably well in accurately determining the location and orientation of the frontal zone. Several statistics were examined and used to demonstrate the utility of the frontal zone in effectively delineating radar estimated rainfall into regions with different spatial and intensity characteristics.