PNNL

Abstract

The DOE ARM Program operates three independent surface radiation measurement systems co-located within a few meters at the Southern Great Plains Central Facility (SGP CF) site. This redundancy affords a unique opportunity for producing a high quality estimate of the actual continuous irradiance record. The Best Estimate Radiation Flux Value Added Product (VAP) currently being developed for ARM (beflux1long VAP) is attempting to determine the best estimate value for each radiation field from these multiple measurements as an operational product. In the development of this VAP, it is necessary to assess the nominal long-term unattended operational accuracy (as opposed to accuracy assessments based on calibrations or short term attended operation) to screen the data for quality assessment. We will present statistical results of this assessment, including our estimates of nominal operational accuracies, and the amount of data that pass the resultant data quality testing. Central to data quality assessment is the notion that having three pieces of information allows one not only to detect measurement problems, but to identify which of the three similar measurements is likely to be in error. We will discuss the techniques we have developed to use similar, but often differing, measurement data as comparison tools for operationally detecting measurement errors. We will also present statistical analyses of the resultant best estimate radiation climatology for the SGP CF.