A Comparative Study between Meteorological Data from Conventional and Automatic Weather Stations in Espírito Santo, Brazil

Abstract

Meteorological variables are mainly monitored by conventional and automatic weather stations. Presently, conventional weather stations are now being replaced by automatic weather stations or being installed to complement and improve observations in areas where there is little or no observation. In order for this permanent replacement to take place, it is necessary that the exposure conditions of the sensors and the methodologies used to obtain meteorological data remain standardized. This study aims to carry out a comparative study of meteorological data from the conventional and automatic weather stations in two cities, São Mateus and Vitória, located in the State of Espírito Santo, Brazil. Daily meteorological data series of maximum and minimum air temperatures, average relative humidity, rainfall and atmospheric pressure were used simultaneously from 2007 to 2016. The data from the respective stations were compared using frequency histogram, linear regression analysis, a coefficient of determination, Willmott index of agreement, bias (systematic error), relative root mean square error, confidence coefficient, and Pearson correlation coefficient. From the results, it was observed that the best data adjustments were found for maximum and minimum air temperature and atmospheric pressure, as for the other meteorological variables, there was a need for adjustment coefficients so as to ensure that the current historical series continue to exist in order to consequently replace conventional weather stations with automatic ones.