Quality control of the series of daily temperature for the stations of the INIA-Venezuela in the period 1950-2005

Abstract

The data quality control as an initial phase in any study that involves calculation, it is a vital step that ensures the accuracy of the results and interpretations of them. The most notable problem with the quality control of climate data is the presence of outliers due that, removing them or keep them can affect the analysis of events leading to overestimates or underestimates. In the present work was implemented a methodology for quality control of temperature data through the integration of various analysis that includes the identification of outliers with robust statistical measures in 15 climatic stations of INIA. It was found that the majority of the comments fell dawn within the expected thresholds. In the analysis of the temporal series, some cases were detected with Tmax>50 and Tmax °C<Tmin (OTD< 0 °C), which were eliminated. In the station CENIAP-Maracay it was found very high values of FD10 (10 days in 1959 with Tmin< 10 °C) and SU35 (183 days in 1959 with Tmax>35 °C). It was demonstrated that through the construction of intervals with mean and deviation bi-weighted, it is possible to detect series with questionable data, as the temperature serie may-november/1959 of CENIAP-Maracay where Tmax> than the upper limit and Tmin<than the lower limit. The questioned values were reported in the refined data base.