Raw signals collected by radar are expressions of the echoes received from the transmit pulse. To transform these first-level data into valued information for meteorology and other applications require a set of steps whose bases lie at the interface between engineering and meteorology. These include:
As we try to extract increasing amount of information from radars, these techniques undergo constant evolution. Examples of some of our work on the subject include, but are not limited to:
Error characterization: The quantitative use of radar products requires a much improved characterization of the errors in each product than we have ever had. Three separate efforts have been launched in that direction. First, we are continuing a thorough evaluation of the error structure of rainfall products: how does the variability of drop size distribution affect rainfall estimates; how does the variability in the vertical structure of reflectivity bias rainfall; how correlated are those various errors (Berenguer and Zawadzki 2008, and references therein). Second, we are trying to identify and quantify the sources of errors affecting refractive index measurements by radar (Park and Fabry 2010). Finally, we are starting an effort to determine and archive the error on each of the second-level radar quantities for each radar pixel. In the process, we are also attempting to improve the way we identify the type of targets generating the observed echoes and clean radar data from unwanted contamination.
Sensitivity characterization and enhancement: Our radar has never been very sensitive for a variety of reasons. As a result, we have been forced to develop innovative ways to characterize our sensitivity for the purpose of identifying what were the limiting factors. In parallel, we are also working on signal processing approaches to increase our ability to detect very weak echoes (Fabry 2010, ERAD).
Integrated display system (Profilers): We have many radar sensors, each providing a piece of puzzle. Most processing and visualization systems display data from one sensor. We have developed a processing and display system called Profilers that allows us to combine information from our surface sensors (disdrometers, POSS, …), vertically pointing radars (UHF, VertiX, MRR…), and scanning radar.