Monitoring Grassland Desertification around the Built-Up Area of the City Based on Multi-Temporal Remotely Sensed Images

• Published in: 2009 International Conference on Management and Service Science pp. 1 - 4
• Main Authors: Guang-jun Wang, Mei-chen Fu, Qiu-ping Xiao, Zeng Wang
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.09.2009
• Subjects: Artificial neural networks; Cities and towns; Control systems; Educational technology; Geology; Paper technology; Remote monitoring; Remote sensing; Satellites; Vegetation mapping
• ISBN: 1424446384; 9781424446384
• DOI: 10.1109/ICMSS.2009.5303133

This paper presents the spatial extent, magnitude and temporal behavior of grassland desertification caused by urban expansion. A time-series of Landsat TM images covering a period of 20 years from 1987 to 2006 was used. The data sets were geometrically and radiometrically pre-processed in a rigorous fashion, followed by a linear spectral mixture unmixing approach to derive the vegetation fraction images and sand fraction images. Through combing the vegetation fraction images, sand fraction images, and NDVI images, the grassland desertification information on the extent and spatial distribution around the built-up area of the city was extracted based on ANN (Artificial Neural Networks) algorithm. The results of our studies indicate that there is a gradual increase in urban expansion from 1987 to 2006, and the similar trend was observed in the temporal magnitude behavior of the severe grassland desertification around the built-up area of the city; the severe grassland desertification is always in the shape of a triangle around the built-up area of the city, and developed to the east quickly with the help of the monsoon; there is a positive correlation between the built-up area of the city and the area of severe desertification, so the built-up area of the city can serve as an important assessment and prediction indicator for grassland desertification due to urban expansion.