A Study On Extraction Of Plastic-Mulched Landcover And On Retrieval Of Mulched Land Surface Temperature Using Remote Sensing Data

The scale of global plastic films covered farmland, which has become an important agricultural landscape, is increasing rapidly at a rate of 20% every year. As a superpower of the plastic films covered cultivation, the scale of the covered farmland had risen to 20.79 million hectares by the year of 2010 in our country, which accounts for 17% of the country’s arable land. Such large-scale landcover change should impact the climate and environment not only regionally but also globally. This dissertation chosed Xinjiang cotton region, which is the largest mulch-covered area in our country, as the study area. Based on four scenes of Landsat TM images, it focused on extraction of plastic-mulched landcover information and on retrieval of mulched land surface temperature. The contents are listed as follows:(1)By analyzing two periods of spectral curves between plastic-mulched landcover and other land use types, and by using other remote sensing characteristics, the feautres of extraction of plastic-mulched landcover information were identified.(2)By defining platic-mulched landcover classification rules, a decision tree model was proposed for platic-mulched landcover information extraction.(3)By comparing several main Land Surface Emissivity(LSE) estimating methods, Zhihao Qin’s model was adopted to retrieve land surface temperature(LST), and then Mono-window algorithm was used to estimate land surface temperature.The results showed as follows:(1)The characteristics of plastic-mulched landcover can be interpreted visually. Due to the differences of subsurface soil conditions, two types of plastic-mulched landcover were identified. Their reflection rate difference about 0.1 on each spectral band. However, their spectral curves are similar to those of fallow land and bare land, whereas distincting from those of saline-alkali soil, vegetation, and other land types. Moreover, the reflectivity of plastic-mulched landcover is lower than fallow land and bare land on band 5 and 7, which is similar to those in band 1,2 and 3.(2)The proposed model can solve issues like same object with different spectrums(e.g plastic-mulched landcover) and different objects with same spectrum(e.g fallow and bare land) in the studied area, and can be used to extract plastic-mulched landcover effectively with an overall classification accuracy above 90%.(3)The surface temperature of plastic-mulched landcover is lower than that of fallow land and bare land. Greenhouse effect of mulched films is not obvious by retrieving of LST using the satellite images.The results can offer a useful reference for extracting plastic-mulched landcover, and the scientific basic for the government with film mulching cultivation and management for plastic film residuing, and the research basic for the influence of plastic-mulched landcover.