High spatial and temporal resolution Normalized Difference Vegetation Index (NDVI) data can be used to describe vegetation dynamics and provide the variation of surface for monitoring phenology and land cover change quantitatively. This paper presents a method using MODIS Land Cover data with 30m LULC map calculates the percentage of every class in the MODIS pixel. And the mean MODIS NDVI can be got through the average value of pure pixels using MODIS NBAR product from 2004 to 2010. Then the logistic model is fitted to the average MODIS NDVI to simulate the variation in NDVI time series. At last, the simulated NDVI time series of all vegetation types are extracted as background values and the HJ-1 CCD NDVI is used to adjust the curve of time-series NDVI to estimate the NDVI at high spatial and temporal resolution. The method is applied to the Heihe River basin and the region growing two crops a year. The results are compared with some filed measured data, which shows the high feasibility of the method to generate accurate and reliable data. It is proved that the method can be used in small scales to lager regions and the results can be a kind of fundamental data in other studies.
Based on the Aster LAI estimation, the main object of this paper is to generate the high spatial and
high temporal resolution LAI product. One method is proposed to get high spatial and temporal resolution LAI
product by fusing MODIS LAI product and Aster LAI. In this method, the LULC data is used to register with
MODIS data, then the percentage of classes of PFT classification in the MODIS pixel can be calculated. And the
multi-year mean MODIS LAI values are the background data, the Aster LAI is used to adjust this curve of
multi-year mean MODIS LAI. And we validate LAI with high spatial and high temporal resolution using the
measured data that is not to be used as the training data. The results is good and can meet our study needs.