Leaf Area Index（LAI） is one of the most important parameters in describing the dynamics of vegetation on land surfaces.LAI products have been produced from data of many remote sensing satellite sensors

such as the Moderate Resolution Imaging Spectroradiometer（MODIS）.In this paper

we used the Dynamic Harmonic Regression（DHR） model to analyze the LAI time series products.The model can decompose the trend

seasonal and residuals components from the original time series

and predict the short-time LAI values.We use the DHR model to extract the time change information from the MODIS LAI time series products.The results show this method to be very effective in predicting the short-term LAI on the pixel basis.