Estimation of the accuracy of GHCN-CAMS reanalysis in calculations of the intra-annual distribution of air temperature in the territory of the Sakha Republic (Yakutia)

Abstract

The accuracy of GHCN-CAMS reanalysis reproduction of the intra-annual air temperature distribution was assessed for the historical (1961-1990) and current (1991-2020) climatic periods, on the scale of climatic norms of seasons and individual months. Data on 86 meteorological stations on the territory of the Sakha Republic (Yakutia), for which climatic series in both time periods were available, were used. The climatic norms of meteorological stations for both periods, as well as their change between periods, were compared with the GHCN-CAMS reanalysis data in corresponding pixels. The root mean square error, Pearson correlation coefficient, and Lin's concordance coefficient were calculated as similarity metrics. Climatic norms for meteorological stations and calculated in the GHCN-CAMS reanalysis model are closely related linearly, with a correlation coefficient above 0.9, and the highest accuracy is observed for spring months. At the same time, the quality of reanalysis reconstruction of change in climatic norms for months and seasons between periods is significantly lower; the best reanalysis accuracy is for fall and winter, and significantly lower for summer. In general, the GHCN-CAMS model can be used for calculation and forecasting applications that require intra-annual air temperature distribution.