Spatio-temporal outlier detection and analysis in pressure and wind speed data to study climate change

Detecting outliers is one of the most important steps in data analysis. In climate data, an outlier can be an extreme event or an error due to measurement, observation and recording. If outliers which are error are not identified and deleted, they will be recorded as extreme data and cause the bias in result of the climate change studies. In this paper, the outliers of wind speed and pressure data for meteorological stations in the normal climatic period 1991-2020 were analyzed. For this purpose, first the spatial outliers were determined using the CLIMATOL algorithm and second, errors were identified by temporal and meteorological analyzes. In the first step, for the parameters of station pressure, sea level pressure, vapor pressure, wind speed and maximum wind speed were identified as 40, 42, 93, 52 and 41 outliers, respectively. In the second step, 20, 10, 56, 20 and 27 of those data were recognized error, respectively. These results have been reported by station and date, to be used by researchers in other studies, especially climate change studies.