Space-time analysis of radar rainfall fields

Since the first conception of using the radar as a meteorological tool the importance has been to process the radar measurements to obtain a meaningful estimate of precipitation (or specifically rainfall). Typically, radar rain rate algorithms are expressed in formula representations that are deduced through experimental means or by simulation methods. Ultimately the goal has been to relate the radar estimates of precipitation to ground truth measurements of rainfall taken at the surface. Making radar rain gauge comparisons is a difficult task that is complicated by various error sources inherent between the two instruments. In this dissertation three techniques are introduced that relate the space-time measurements of the radar and rain gauge instruments to each other. Each method produces an estimated optimal area of rainfall at a specified radar elevation level that is correlated to rain gauge measurements on the ground. A radar data time series profile is generated along with a time delay of rainfall that can be used to accurately compare the radar measurements to the gauge measurements. A Surface Projection Method is introduced that uses only radar data to estimate rainfall at a specified point on the ground.; In the dissertation several convective rainfall events are presented that encompass a variety of meteorological conditions. Each of the three techniques developed here are applied to the rainfall events in order to illustrate the details of the procedures and to demonstrate the performance of the techniques versus surface measurements. Polarimetric radar observations from the CSU-CHILL and S-POL radars were taken simultaneously with conventional rain gauges, a capacitance rain gauge, and a 2D video distrometer. Comparisons are made between various radar rain rate algorithms against these surface instruments. Direct comparisons of the radar measurement parameters are also made against the distrometer.