Kirono, Dewi G. C and Jones, Roger ORCID: 0000-0001-6970-2797 (2007) A bivariate test for detecting inhomogeneities in pan evaporation time-series. Australian Meteorological Magazine, 56 (2). pp. 93-103. ISSN 0004-9743

Abstract

Homogeneous climatological time series are necessary for studying historical climate variability and change. This paper describes the use of the bivariate test for detecting and adjusting discontinuities in Class A pan evaporation time series for 28 stations across Australia, and illustrates the benefit of using corrected records in climate studies. Ninety-two per cent of the inhomogeneities detected by the bivariate test are consistent with station metadata. Even though the test was designed to detect a single discontinuity in the mean, it can also be sensitive to multiple shifts in the mean. These show the suitability of the bivariate test as a tool for screening pan evaporation data. Having identified inhomogeneities, the adjustments were only applied to records which contained inhomogeneities that could be verified as having a non-climatic origin. The use of original and adjusted records in correlation analysis and in trend analysis produce very different conclusions. At Esperance and Woomera, for instance, unadjusted pan evaporation records do not correlate with potential evapotranspiration and are positively correlated with rainfall, whereas those of adjusted pan evaporation result in a more sensible inter-variable relationship (i.e. pan evaporation is negatively correlated with rainfall and is positively correlated with estimated potential evaporation). In a trend analysis, most unadjusted pan evaporation records show a statistically significant negative bias which, in most cases, is removed with adjustment. This is consistent with the effect of bird guard installation, early in the time series, that reduces pan evaporation. The trend in the original average of all stations was adjusted from -2.8±1.7 to -0.7±1.6 mm year-2 for 1970-2004, demonstrating the importance of screening the data before their use in climate studies.

Search Google Scholar

Repository staff login