Evaluation of Rainfall Measurements from the WXT510 Sensor for Use in the Oklahoma City Micronet

ABSTRACT

A network of 40 real-time, automated atmospheric monitoring stations was deployed in Oklahoma City and officially commissioned on 8 November 2008: the Oklahoma City Micronet (OKCNET). The Oklahoma City Micronet includes 36 stations mounted on traffic signals which utilize the Vaisala WXT510 sensor. As part of the design of the WXT510, an impact sensor is utilized for the collection of rainfall observations. Prior to deployment in Oklahoma City, an array of 33 WXT510 sensors were deployed at the OKCNET intercomparison facility and compared with traditional instruments used to measure rainfall including tipping bucket rain gauges and a Geonor weighing gauge. The results of the comparison revealed that a consistent, linear bias was present between the WXT510 sensors and the traditional gauges whereby, on average, the traditional gauges measured approximately 26% less precipitation than the WXT510 sensors. In addition, the variation in recorded rainfall between WXT510 sensors was consistent with that recorded by the tipping bucket gauges. As such, a correction was developed using the WXT510 and tipping bucket data. This correction was applied to the WXT510 rainfall observations and cross-verified using the Geonor gauge data. The overall result of the study yielded a bulk correction that can be applied to rainfall observations recorded by the WXT510 to greatly improve the accumulated rainfall values. This correction is designed to improve the overall accuracy of the observations without specifically calibrating each individual WXT510 sensor and is valid regardless of rainfall intensity, length of the precipitation event, seasonal characteristics of the rainfall, or rainfall type (i.e., stratiform, convective, etc.)