Forecasting of Air Temperature Based on Remote Sensing

Abstract

The aim of this research is to forecast air temperature based on remote sensing data. So, land surface temperature and air temperature values which were measured by Republic of Turkey Ministry of Forestry and Water Affairs (Turkish State Meteorological Service) during the period 1995-2001 at seven stations (Adana, Ankara, Balikesir, Dzmir, Samsun, Sanliurfa, Van) were compared. The monthly land surface temperature and air temperature were used to have correlation coefficients over Turkey. An empirical method was obtained from equation of correlation coefficients. Separately, Price algorithm was used for the estimation of land surface temperature values to get air temperatures. Then as statistical, air temperature values, belongs to meteorological data in Turkey (26-45 degrees E and 36-42 degrees N) throughout 2002, were evaluated. The research results showed that accuracy of estimation of the air temperature changes from 2.453 degrees K to 2.825 degrees K by root mean square error.