Reconstruction of Hydrometeorological Data in Lake Urmia Basin by Frequency Domain Analysis Using Additive Decomposition

Abstract

Frequency domain analysis using an additive decomposition method is proposed to reconstruct the missing hydrometeorological data of selected sites in Lake Urmia basin in Iran. Precipitation, evaporation, streamflow and groundwater time series are used for this aim. Trends, within- and multi-year cycles, and randomness are taken into account to reconstruct each of the time series for which models are developed, calibrated and validated separately. Statistical similarity between the observed and reconstructed time series is checked. Statistical characteristics including the average, standard deviation, skewness, and the first-order autocorrelation coefficient are well preserved at the reconstructed time series. A conceptual water budget model is also established to check for the consistency between the reconstructed and the observed datasets. The water budget model is taken as a quantitative way to confirm that the frequency domain analysis using the additive decomposition is an effective method for the reconstruction of the missing hydrometeorological data based on the case study performed for the Lake Urmia basin in Iran.