The Key Technologies Research Of Cotton Monitoring Operational System Using Remote Sensing In Xinjiang

Northwest China's Xinjiang Autonomous Region is a major cotton production area in China. Its cotton production is not only very important for Xinjiang's economy, but also China's cotton market and textile industry. A cotton monitoring operational system using remote sensing is needed to collect objective, efficient and timely information of cotton growing area at a regional and national level for cotton production and trade management in Xinjiang, as well as in China. The objective of this research is to develop key technologies for cotton monitoring operational system in Xinjing using remote sensing, GIS and GPS. The research results have made the operational monitoring system more efficient, economic, and reliable. The main works and achievements are as follows.1) Cotton monitoring zoning in Xinjiang using remote sensingA background information data base was developed for the cotton monitoring zoning, and zoning index is identified such as climate, topography, crop composition and structure, crop calendar, and cotton distribution, etc. Based on the data base and zoning index, the zoning results are: (1) Landsat TM image is suitable data for cotton monitoring, (2) September is the best season to discriminate cotton on Landsat TM image, (3) Xinjing is divided to 6 sub area to remote sensing monitoring.2) Spectral Information based cotton discrimination model using Landsat TM imageCotton and other crops' spectral information were collected during cotton growth season in a field work in northern Xinjing. Analyzing spectral features of field measurements and Landsat TM image, the best time of cotton identification using remote sensing is confirmed. And a spectral information based cotton discrimination model using Landsat TM image on a large scale was developed. The model is simple, more accurate and suitable for operational work which was evaluated by mathematics analysis and field experiment.3) Linear features extraction and measurement in cotton area estimation using Landsat TM imageLinear features, such as roads and irrigation channels, make a result of about 13% errors in cotton area monitoring using Landsat TM images in Northwest China's Xinjing Province. For the operationally cotton monitoring using remote sensing in China Ministry of Agriculture, a sampling method for linear features extraction and measurement to reduce errors was developed by the authors. The theory approaches and field experiment are introduced in this paper and the results show the method can efficiently improve the results of cotton area estimation using Landsat TM.4) Sampling method research for cotton growing area estimation in XinjiangAfter analyzing several sampling methods used in China and abroad, a sampling method integrated two-stage sampling and stratified sampling method was adopted in this research. The advantages of this method are: (1) reflecting the cotton area proportion difference in different zones, (2) reducing the sampling error. The most important advantage is that the absolute cotton growing area can be measured in county level in Xinjiang Autonomous Region.5) Improving and application the operational cotton growing area monitoring system using remote sensingThe original operational cotton growing area monitoring system used by Ministry of Agriculture relies on traditional visual image interpretation technology, no standards for the processing, and the final result is two years change rate of cotton growing area. The improved monitoring system using the above key technologies is used to a case study in Shihezi in Xinjiang, and not only absolute value of cotton growing area can be estimated, but also the accuracy is better.