Solar radiation data validation

This paper describes and applies a procedure for the quality control and validation of solar radiation data for two independent co-located measurement systems based at Loughborough University, United Kingdom. An assessment of the measurement error of simultaneous data from four pyranometers was undertaken over a range of averaging periods. A data filter of 0-1500W/m2 was found to reduce measurement errors by a factor of between 2 and 4 with observed hourly, daily and monthly errors of approximately 9%, 5% and 3.5% respectively for all sensors. These errors were greater than those found in the literature, indicating the possible presence of a systematic component of error. Analysis of the temporal variation of measurement error and its relationship with incident irradiance indicated the presence of an inter-system discrepancy in sensor offset. The close proximity of the two systems indicated that this was due to differences in system calibration, maintenance or response rather than environment and the results can therefore be used for future system re-calibration and to improve data accuracy. This paper demonstrates that straightforward validation procedures can yield meaningful results and greater emphasis on data validation is recommended for the solar community as a whole.