Multi-Sensor Measurements of Raindrop Size Distribution at NASA Wallops Island

Disdrometer is an instrument that measures the raindrop size distribution (DSD) at a point on the ground. Disdrometers also measures the fall velocity and shape of falling hydrometeors. While long-term (> year) disdrometric measurements are required to have a sufficient sample in studying the physical aspects of the DSD, the accuracy of the measurements are equally important in determining the error covariance of remote sensing of rainfall. Knowledge of shortcomings of disdrometers is essential in determining the accuracy of the measurements and requires coincident measurements of the same and different types of disdrometers. In this study, we examined the performance of impact-type, optical, and radar disdrometers through a field campaign at NASA Wallops Island where two of each type of disdrometers and two tipping bucket rain gauges were operated. Although all three types of disdrometers were able to capture 30 major rain events between May and August 2004, there were significant differences between the same and different type of disdrometers rain accumulation and raindrop spectra. The gauges had a very good agreement throughout the experiment and are considered to be reference. Among the different sources of errors, wind played an important role, particularly on radar disdrometer. The radar disdrometers measured spurious small drops in windy conditions, while background noise from winds seems to suppress the number of small drops in impact-type disdrometers. The radar disdrometers did also undergo recycling mode of four minutes in intense rainfall where lighting is considered to be cause. The optical disdrometers, on the other hand, showed the systematic differences in raindrop fall speed in windy and non-windy cases. While optical disdrometers recorded more small drops than the impact-type disdrometers, the agreement between the two disdrometers was very good in mid-size range that typically contributes most of the rainfall. The optical and impact-type disdrometers showed some significant differences at large size drops where the sampling is limited and the impact-type disdrometer cannot determine the size of very large drops. It was also interesting to note that one of impact-type, optical, and radar disdrometer recorded systematically more rainfall than their counterpart units. One of the impact-type and of the optical disdrometers had very good agreement with the rain gauges and their measurements were considered to be near the truth in measuring rainfall.