Marek Kosowski

A comparison of the radiosonde data and the data set basing on the NCEP-NCAR reanalysis

The purpose of this study is to determine the usefulness of the NCEP-NCAR reanalysis for studying the geopotential height of isobaric levels, air temperature, wind speed and wind direction in the troposphere over the Central Europe. The study compares the values of selected meteorological elements obtained from the NCEP-NCAR data with the values measured at the aerological stations in Łeba, Legionowo, Lindenberg and Poprad. The results regarding geopotential height of isobaric levels reveal a very strong correlation between the NCEP-NCAR and radiosonde data at all the pressure levels studied. However, the analysis of the mean error and R M S E indicates less concordance, especially in the U T L S. The seasonal variability is very small and is noticeable only in Łeba, where the mean error and R M S E both increase in winter. The analysis carried out for the air temperature indicates a high concordance, especially up to 300 hPa. Above, a decrease in correlation is observed, as well as an increase in the values of both errors. It is most evident in summer and autumn. The comparison carried out for the wind speed shows that the correlation coefficients reach the highest values in the U T L S and decrease with decreasing altitude. High values of R M S E are noted at the level of 300 hPa. The values of the mean error and R M S E accompanying wind direction are the lowest in the U T L S and middle troposphere and increase at the 850 hPa pressure level. The differences between the frequencies of wind directions, calculated on the basis of the two datasets are very small ( 0–4 % ). It does not apply to the level of 850 hPa in Poprad, where substantial discrepancies are found. The results regarding wind parameters show very little seasonal variability. It is concluded that the NCEP-NCAR reanalysis is a valuable source that can be used to supplement or even replace radiosonde data. This applies to the entire troposphere, except for its lower layers over mountain areas, where reanalysis data should be verified every single time.