Statistical quality control for local‐scale extreme temperatures. Case study: Lisbon, Portugal

• Published in: Meteorological applications Vol. 14; no. 3; pp. 275 - 290
• Main Authors: Lucio, P. S., Serrano, A. I., Deus, R. J. R.
• Format: Journal Article
• Language: English
• Published: Chichester, UK John Wiley & Sons, Ltd 01.09.2007
• Subjects: autocorrelation; homogeneity tests; moving‐range; non‐parametric tests; variance; Portugal
• ISSN: 1350-4827; 1469-8080
• DOI: 10.1002/met.30

One of the biggest constraints to the study of meteorological fields is the fact that ground‐based meteorological networks do not operate over a common period of adequate length. In general, the biggest drawback is that recorded data must be gap‐filled and quality controlled (coherent, consistent and homogeneous) to provide a reliable, continuous, reference daily/monthly/yearly series. This paper addresses procedures for statistical quality control of extremes of air temperature (daily maximum and minimum) time series. The diagnostic was accomplished based on the analysis of daily time series (1941–1996), monthly time series (1941–2001) and yearly time series (1941–2001) from 18 meteorological stations located in the Lisbon (Portugal) metropolitan region. To obtain insight into the quality of the series of the temperature data, objective statistical tests for departure from homogeneity were applied to the time series. To illustrate the procedures, one assesses the homogeneity of the reconstructed time series to check the quality of the ‘virtual’ observations and some tests were run over the annual, seasonal and monthly series of extreme temperatures. None of these assessments was able to detect a spatial non‐homogeneity in the temperature series around the tested years. As expected, the optimal interpolation approach for time series reconstruction, proposed by Lucio and Deus (2005), seems to be robust, considering a wide range of methods to test homogeneity of the reconstructed time series. Afterwards, one can apply the reconstituted monthly series of extreme air temperature anomalies, for each meteorological station, to analyse the respective trends, when they exist. Copyright © 2007 Royal Meteorological Society