Presently exploited rates of fossil fuels are expected to deplete resources within the next 40-50 years. and consequently, human beings seek alternative energy resources that are clean. friendly to the environment and sustainable. Accumulation of carbon dioxide in the lower layers of the atmosphere may cause climate change and consequent occurrence of floods, intensive rainfalls and droughts. In order to reduce such dangerous effects all countries have to try to improve their energy resources quality and, if possible, to replace fossil fuels, such as coal, with the renewable alternatives of wind. solar and solar-hydrogen energies. Among these, wind power has a priori significance for Turkey. Wind time series depend very much on meteorological measurements of wind direction and velocity. Unfortunately, in many parts of the world. it is difficult to obtain such data for wind speed time series assessments. In this study, harmonic analysis is used to model the daily wind speed values recorded at ten stations in the Marmara region, Turkey, with distinct meteorological conditions from 1993 to 1997. The coefficients. amplitude, variance and phase angle, of each harmonic are calculated for the months of January, April. July and October, leading to total variance maps for spatial interpolations. It is seen that up to the 9th harmonic more than 80 % of the total variance can be presented. The western and eastern parts of the Marmara region have different wind pattern characteristics. The contributions of each harmonic to the total variance are calculated, and then regional variance maps are evaluated. (C) 2004 Elsevier Ltd. All rights reserved.