Homogenisation of relative humidity data series

The MASH (Multiple Analysis of Series for Homogenization) system was developed to complete, control and homogenize long-term monthly and daily data series. The data series homogenised by MASH sytem are the basis of other systems, since long data series represent the spatial and temporal samples for climate information. The daily average relative humidity data series are homogenised with using the MASHv3.03 software operationally at the Hungarian Meteorological Service.

In this study we present the verification statistics of the station data series and we show the representativity values of the station network. The data quality control part of MASH works efficiently for daily relative humidity values. We show the quality control results, including cases when the suspicious values are acceptable. Doubtful measurements come out due to the existence of e.g. inversion situations.

Recently we have homogenised not only the daily data but also the 6-hourly measurements. As the mathematical background of the MASH software allows us to use the homogenisation results of the past years, we homogenise these data series with using the results of the daily data. Since the hourly data series are highly inhomogeneous, moreover, their inhomogeneities are not the same as those of the daily series, therefore we cannot ignore the daily trend of the inhomogeneities.

Consequently, we decided to homogenise the 6-hourly series separately, using the usual MASH procedure. When homogenising the 6-hourly data series with the MASH system, the breakpoints detected in the daily data series can be automatically used as metadata. The results of the homogenisation of the 6-hourly values are also presented. These sub-daily data series are important for the accuracy of hazard warning, e.g. to study foggy situations.