|Eric, Brown DE Colsto -|
|Ma, Ronghua -|
|Weng, Qihao -|
|Masek, Jeffrey -|
|Chen, Jin -|
|Pan, Yaozhong -|
|Song, Conghe -|
Submitted to: International Journal of Remote Sensing
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: May 5, 2012
Publication Date: July 2, 2012
Repository URL: http://handle.nal.usda.gov/10113/59917
Citation: Gao, F.N., Eric, B.U., Ma, R., Weng, Q., Masek, J.G., Chen, J., Pan, Y., Song, C. 2012. Mapping impervious surface expansion using medium resolution satellite image time series: A case study in Yangtze river delta, China. International Journal of Remote Sensing. 33:7609-7628. Interpretive Summary: For the first time in human history more people are now living in the cities than in the rural areas in the world. As a result, cities are sprawling rapidly into their surroundings. Although urban environment provides an improved quality of life for most of its dwellers and is advantageous to industrial growth, excessive expansion of urban areas can cause a series of environmental problems, including but not limited to, loss of fertile land for food production, changes to the regional climate and water cycle, loss of habitats for wildlife, and altering ecosystem goods and services. These impacts have been recognized and studied for decades. However, due to the constant expansion of urban areas (impervious surfaces), its multidimensional impacts on the environment have not been thoroughly understood. A consistent and continuous time-series map of impervious surface area is needed for this effort. In this paper, we developed an efficient mapping algorithm to map the impervious surface from a time-series of satellite images. The approach was tested in the lower Yangtze River Delta region, one of the fastest growing urban areas in China. Results from nearly four decades of medium resolution satellite data show a consistent urbanization process that agrees with economic development plans and policies. The time-series of impervious spatial extent maps derived from this study agree well with an existing urban extent polygon dataset that was previously developed independently. The overall mapping accuracy was estimated at about 92.5% for impervious type from all inputs regardless of image quality and spatial resolution. Though the mapping algorithm was developed with the lower Yangtze River Delta as a case study, the approach is not location specific, and can be used elsewhere with a similar set of moderate resolution images.
Technical Abstract: Due to the rapid growth of population and economic development in the developing countries, more people are now living in the cities than in the rural areas in the world for the first time in human history. As a result, cities are sprawling rapidly into their surroundings. A characteristic change associated with urbanization is the expansion of impervious surfaces. Mapping the dynamics of impervious surface expansion both in space and time is essential for an improved understanding of the urbanization process, land cover and land use change and its impact on air and water quality. Landsat and other medium resolution satellites provide sufficient spatial details and temporal frequency for mapping the impervious surface expansion over the past four decades. Since the US Geological Survey opened the historical record of Landsat image archive for free access in 1998, the decades old bottleneck of data limitation is gone. Remote sensing scientists are now rich with data, and the challenge is how to make best use of this precious resource. In this paper, we developed an efficient algorithm to map the continuous expansion of impervious surface area using a time series of four decades medium resolution satellite images. The objective is to extract a complete and consistent time series of impervious surface maps from a corresponding times series of images collected from multiple sensors with minimal amount of image preprocessing efforts. A supervised classification approach using decision tree based on predefined pervious and impervious training samples derived from each individual image is developed. The pervious and impervious samples from each image were then assembled as a time series image stack. Assuming irreversibility of the impervious surface during the period of study, we were able to remove inconsistent training samples, and generated a single layer training mask with each class representing the timing of conversion from pervious to impervious surface, plus a permanent impervious class and a permanent pervious class. The multitemporal images were stacked as a composite image in the same order of time as the impervious/pervious image stack, and then classified the composite image stack with the trained decision tree as an integrative whole. The approach was tested in the lower Yangtze River Delta region, one of the fastest urban growth areas in China. Results from nearly four decades of medium resolution satellite data from Landsat MSS, TM, ETM+ with SLC on and off, and CBERS show a consistent urbanization process that agrees with economic development plans and policies. The time-series of impervious spatial extent maps derived from this study agree well with an existing urban extent polygon dataset that was previously developed independently. The overall mapping accuracy was estimated at about 92.5% for impervious type from all inputs regardless of image quality and spatial resolution.