Measurement Uncertainty and System Assessment for Weather Radar

With the widespread deployment of dual-polarization weather radars, data quality has become one of the major concerns for weather radar users worldwide. A well designed and constructed radar system is a premise to acquire high quality radar data. In practice, measurement precision and system stability are primary questions to be answered for an operational weather radar system. It is well known that weather radar measures signals from the hydrometeors, which are generally regarded as random media. Previous studies have mainly addressed the measurement uncertainty caused by the sampling effect. The uncertainty attributed to system stability and quality has not routinely been a major focus. However, the latter uncertainty is more important for assessing the overall quality of a radar system.

The major objective of this study is to quantify the measurement uncertainty of weather radar and assess the system quality. Practically, the measurement uncertainty can be discerned as the measurement variation and can be quantified as the standard deviation (SD) of radar moment estimates. Considering the random nature of the radar return from hydrometeors, the sampling effect is generally the major factor contributing to the statistical fluctuation of moment estimates. Here, this uncertainty is regarded as “sampling-induced uncertainty”. On the other hand, the hardware imperfectness (noise, instability, or other potential factors) may cause the “system-induced uncertainty”. The texture analysis method is usually used to quantify the measurement uncertainty. However, texture analysis cannot separate the contributions associated with these two types of uncertainties. Moreover, the texture analysis method requires uniform/quasi-uniform precipitation. Many factors, such as attenuation, spatial variation, and boundary effects, could degrade the quality of quantification.

Recently, Enterprise Electronics Corporation (EEC) has proposed a robust and easily-implemented method to quantify the measurement uncertainty of weather radar. It is based on the point mode scanning strategy, and more importantly, the “system-induced uncertainty” can be isolated from the total measurement uncertainty. Therefore, it is particularly helpful for assessing the overall quality of an operational weather radar system.