NDVI Maps Research Articles

NPP distribution related to the terrains along the North-South Transect of Eastern China

The time series NDVI distribution maps of the study area were calculated with AVHRR images acquired from May 1999 to April 2000 on the North-South transect of Eastern China. Based on the analysis of the characteristics and their contributions to NPP on the transect, the RS NPP distribution maps were created by zones and the quantitative models were established between NDVI maps and field measured NPP sample data collected from Central and South China for forest and cultivated land. Furthermore, the spatial distribution and quantitative result were obtained through an overlapping and fitting analysis between the NPP maps and DEM. The result shows that the NPP distribution features are not only distinguished on detail zonation of natural vegetation in vertical, latitudinal and longitudinal distribution but also reflected an obvious effect from the crop growing period of agricultural area. According to the results above, an opposite opinion was given that the NPP values for the forest and agricultural areas vary greatly in different regions, with the forest areas having a higher value than the agricultural areas in Southern China. In the Huang, Huai and Hai Plains, agricultural regions have higher NPP values than the surrounding low mountains. In Northeastern China, the NPP values form a gradient across the mid-to-high mountain forest zone to plains and low mountain areas. Further, for most of the agricultural areas, NPP values range between 25–35 t · hm2 · a−1 and do not show significant areal differentiation.

Extraction of vegetation cover in an arid area based on satellite data

An attempt is made to derive a vegetation index applicable for extraction of vegetation cover in an extremely arid area such as Taklimakan Desert from Landsat TM image data. In this method a ratio of vegetation coverage in each pixel is computed from the ratio of Landsat TM band 5 and 7 digital values. Then if the value of the ratio is larger than the threshold NDVI (Normalized Vegetation Index, Rouse et. al., 1974) is selected while if the value is less than the threshold SAVI is selected thus the vegetation index consisting of combination of NDVI and SAVI is made which can be defined as NDSAVI. The threshold is decided as follows. The reflectance in the spectral bands corresponding to TM band 5 and 7 is measured for various places in Taklimakan Desert using spectral radiometer which are compared with NDVI and SAVI and an appropriate threshold is decided. It has been confirmed that NDSAVI map shows better correspondence with actual vegetation cover in an arid area than single NDVI or SAVI maps.