Fractional vegetation cover change based on vegetation seasonal variation correction: A case in Lianjiang County, Fujian Province, China

Abstract

Remote sensing change detection based on fractional vegetation cover (FVC) has become an important way in the research of vegetation and related ecosystems. It is difficult to meet the requirement for optical remote sensing in subtropical areas because of cloudy/rainy weather conditions. Using images from different seasons in the vegetation change detection will inevitably lead to errors in the change detection results due to the seasonal difference. To overcome this problem, we proposed a method for correcting vegetation seasonal variations by taking advantage of high temporal resolution advantage of MODIS remote sensing data and the high spatial resolution of remote sensing data. Based on the relationship between MODIS vegetation data in different seasons via regression analysis, we transformed the vegetation information of the high resolution images of corresponding years to the required season of the years. The method was applied in the Aojiang basin area of Lianjiang County in Fujian Province, China, with good results of vegetation information transformation. The results showed that after transforming vegetation information of the 2007 winter scene and 2013 spring scene of high resolution images to those of summer season, the FVC was enhanced from 66.5% to 79.7% for 2007, and from 58.6% to 77.9% for 2013. Our method effectively removed the seasonal difference of FVC and improved the accuracy of the FVC-based change detection results.