标题：Accurate Forecasting of the Satellite-Derived Seasonal Caspian Sea Level Anomaly Using Polynomial Interpolation and Holt-Winters Exponential Smoothing
摘要：Polynomial interpolation and Holt-Winters exponential smoothing (HWES) are used to analyze and forecast Caspian Sea level anomalies derived from 15-year Topex/Poseidon (T/P) and Jason-1 (J-1) altimetry covering 1993 to 2008. Because along-track altimetric products may contain temporal and spatial data gaps, a least squares polynomial interpolation is performed to fill the gaps of along-track sea surface heights used. The modeling results of a 3-year forecasting time span (2005 - 2008) derived using HWES agree well with the observed time series with a correlation coefficient of 0.86. Finally, the 3-year forecasted Caspian Sea level anomalies are compared with those obtained using an artificial neural network method with reasonable agreement found.
