Testing the homogeneity of short-term surface solar radiation series in Europe

Abstract

Non-climatic factors, such as changes in instruments or the relocation of meteorological stations, can cause sudden shifts or gradual biases in a climate data time series. The use of such inhomogeneous time series in data analysis might lead to false conclusions about climate variability and change. In this work, we test the homogeneity of 172 surface solar radiation (SSR) monthly series over Europe available in the Global Energy Balance Archive (GEBA) during the period 2000-2007. Four absolute homogeneity tests are applied to each series, and a classification of inhomogeneous and homogeneous stations is given. The results show that 20 out of 172 series (11.6% of the total) are inhomogeneous at the 99% significance level. The mean average time series of both data sets, the original and the one with only the homogeneous series, show positive linear trends (0.59 and 0.70 Wm-2yr-1). The omission of the inhomogeneous series increases the original trend by 0.11 Wm-2yr-1 or 1.1 Wm-2decade-1. Our results highlight the importance of testing the homogeneity of SSR time series before any trend analysis is performed. © 2013 AIP Publishing LLC.