A retrieval and validation method for shelterbelt vegetation fraction

Abstract

Shelterbelts are important in defending against natural disaster and maintaining ecological balances in farmland. Understanding of the shelterbelt vegetation fraction is fundamental to regional research of shelterbelts using remote sensing. We used SPOT5 imagery with 10×10m spatial resolution in combination with knowledge of the characteristics of shelterbelts to develop a method for retrieval of the vegetation fraction of shelterbelts by the pixel un-mixing model. We then used the method to retrieve values for shelterbelts in study area. By combining the parameters of photographic images with characteristics of shelterbelts, we developed a method for measuring the vegetation fraction of shelterbelts based on an advanced photographic method. We then measured the actual values to validate the retrieval result. The multiple correlation coefficients between the retrieved and measured values were 0.715. Our retrieval and measuring methods presented in this paper accurately reflect field conditions. We suggest that this method is useful to describe shelterbelt structure using remote sensing.

Keywords

shelterbelt / vegetation fraction retrieval / vegetation fraction measuring / remote sensing