Study By Multi-temporal Remote Sensing Technology On The Change Of Pollution Source Of Indigenous Coke-Making In Historical Period

Although indigenous coke-making has been frequently forbidden in historical period, it is never out of sight because of its modest investment and high profit. As indigenous coke-making sites are always in small scale and rest in obscure position, which makes it difficult to detect the exact location by traditional detecting device and technology. Remote sensing technology, with its multi-temporal, multi-scale, multi-spectral as well as dynamic detection, has been applied to various relevant fields by researchers. Traditional detecting device cannot effectively, properly and immediately gain the information-the high temperature generated by indigenous coking which can cause heat anomaly around the sites, but remote sensing technology can get it owing to its features of microscopic, immediate and dynamic.The dissertation, with the help of RS and GIS technology, probes how to discern pollution source of indigenous coke-making by remote sensing, extracts the relevant information and analyzes the number and location of indigenous coking sites in experimental area according to the remote-sensing data on Landsat-7ETM+in1999and2002. Through using Change Detection Technology to indigenous coke-making, it is easy to find the detecting result in two phases in experimental area in1999and2002. The key conclusions are as following:(1) Hot spot of indigenous coke-making is in the band of ETM+7,5. Although smoke plume is obvious in ETM+6thermal infrared band which usually used to recognize fire spot, the result is inferior because of its poor image spatial resolution.(2) It is not true that the higher the value of OIF is, the better of the image recognition on multispectral composite image. Since there is abundant information in the bands of ETM+7,5, so its false color composite image has a perfect visual effect. And DN value of fire spot on single-band spectrum curve has a difference compared with other projects.(3) An integrated method of extracting information of indigenous coking fire spot has been developed by the way of analyzing OIF on false color composite image as well as various visual effects and comparing DN value on spectrum curve of different projects. According to the field research toward typical area, the accuracy can reach90%on the basis of this method.(4) By the aid of Change Detection Technology and GIS Spatial Analysis, the information of indigenous coke-making sites on different phases in experimental area has been extracted and read. The finding is that the number of indigenous coke-making sites increased to forty-six and sixty-eight in experimental areas in1999and2002respectively.The study is a technical support for environmental protection department to research contamination and is a scientific basis for agricultural food safety in indigenous coke-making...