Estimating sunshine duration from global solar radiation for Antananarivo

The aim of this paper is to find the best algorithm, among the existing, to estimate sunshine duration from global solar irradiation for the town of Antananarivo-Madagascar. Five methods in the literature based on Pyranometric Methods have been proposed for four years data as the Step Algorithm, the Carpentras or Meteo France Algorithm, the Slob & Monna Algorithm, the revised Slob & Monna Algorithm and the Linear Algorithm. Methods comparison showed that the best method is the Meteo France Algorithm for Antananarivo and the suitable coefficients are [Formula: see text] and [Formula: see text]. Sunshine duration from solar global radiation results has been discussed. Data analysis has been performed at the parameters statistical significance and in residual analysis for the model fit quality. Results showed that the mean daily differences between the reference and the Meteo France algorithm in hour are 0.0006, the standard deviation of the daily differences in hour is 0.55, the four years total cumulative differences in hour are 0.83 and the Normalized Root Mean Square Error is 8.54%. This study is a part of a global project to study the sunshine duration from global irradiation for the islands in Indian Ocean.