Dynamic Information Monitoring In Irrigation Distract Base On Remote Sensing

The management of irrigation distract and government more and more pay attention to the dynamic information of the Irrigation Distract, especially to land use ,soil water, crop yield etc, because it is the premise of the right leading agriculture to produce and reasonable use of the current resources. It is a very meaning of lessons to monitor the dynamic information of the irrigation distract by remote sensing and offer it for management on time.This thesis make the study of land use, soil water and crop yield in Yangling and Erzhiqu area by six Landsat TM image dates from 2003 to 2005.In the method research of land use by remote image in June 16, 2004, according to the characteristics of land use in Wuquan town, the method research of divide into section was brought, and the method of Principal comp, NDVI(number degree vegetation index), unsupervised classification and tasseled cap was used to the withdraw of land use. The result is wary good, the average accuracy to attain 88% above, through the proved of the five kinds of typical models of classification result.Three models of monitoring soil water was established by the relation of six soil water dates and Landsat TM image dates from 2003 to 2005, and distinguish then 0-10cm, 0-20cm, 0-40cm, 0-60cm average soil water of contained. The methods were the wave band coefficient method, the hot indoor red monitor model, NDSWI (number degree soil waterindex).The wave band coefficient method is a monitor soil water method that each band powera coefficient, and add together. It was established by the "humidity" factor included in TM image, and according to the contribution to monitoring soil water.The hot indoor red monitor model was set up because the relativity of soil water and ground temperature is very high. After all, the ground temperature was derived by TM6 band. The result proved that estimation of soil water by TM6 band is very good, and it is less more affected by the vegetation on ground.The NDSWI method was applied to monitor soil water. The result approved that the best accuracy to monitor soil water is 0-40 cm depth, the next is 0-20cm depth, the third is 0-10 cm depth and 0-60 cm depth. Its best accuracy can attain 80% above in 0-40cm depth in actual application.Results showed that the hot indoor red monitor model is a best method than others, its accuracy can attain 90% above;the nest better are the wave band coefficient method and the NDSWI method, its accuracy can attain 80% above.Result showed that the accuracy that estimated wheat yield using GDVI (green degree vegetation index) by remote sensing is very high. And this method apt to operate, and economy, therefore, this method can use for the actual application. When the method was applied in actual, it should be further modification with perfectly, then increase its stability and functional.