| China is a water short country, with only 2263 m3of water resources per capita,25% of the world's average, which falls into one of the lowest countries in the world. At the same time, as an agricultural country, its agriculture has been kept as the largest water consumer that accounts for more than 65% of the total water consumption. However, under serious water shortage, waste and low efficiency of irrigation water is still popular all over the country. Therefore, necessary countermeasures are needed to build a water-saving society sustainable social-economic development. One is to implement most strict management on agricultural water use, and the second is to improve the agricultural water use efficiency. Last but not least is to promote the modernization of those large and medium-sized irrigation districts.Remote sensing, as an important space technology originated from 1960's,is currently widely applied in agriculture, water resources, environment, etc. Its roles in modernized irrigation management become more and more significant. Through use of remote sensing technology, accurate and fast monitoring of crop planting structure and irrigation area in an irrigation district is now become possible, which can provide effective spatio-temporal data for irrigation management and decision support.In this thesis, remote sensing based crop pattern mapping and irrigated area monitoring were studied in the Qinhan irrigation district in Ningxia Autonomous Region. Remote sensing models for crop planting structure and irrigated area were developed, and verified through field examination and statistical data. The main results are as follows:(1) Cropping patterns in the Qinhan irrigation district were mapped using time series of HJ satellite image data, validated with field experiment data and statistical data. Based on crop phenological characteristics analysis of different crops, the maximum likelihood method and decision tree method were applied for image classification. This paper analyzed the EVI and NDVI vegetation index, finding out EVI value of the identification information of vegetation stronger in the identification of corn and wheat. And in the identification of rice, EVI ability to grow in the early recognition of higher NDVI, late recognition capabilities gap is not, and in the late growth stage, NDVI mean larger and easier to becoming saturated. The temporal changing curves of EVI for different crops were generated and analyzed, and the maximum likelihood method was used in rice field detection, whereas the decision tree method was used in winter wheat, corn and spring wheat filed classification. Compared with local statistical results, remote sensing results is proved rather good, with an accuracy of above 92% for different crops.(2) An irrigation area monitoring model was established based on the perpendicular drought index PDI for the Qinhan irrigation district. After analysis on the Nir-Red spectrum feature space of the 2014 irrigation season of, it is found that the monthly slope M of the spatial soil baseline reduces gradually due to the increasingly weakening of soil reflectance as crop's growing up, and decrease of the average PDI is well consistent with the irrigation schedule. The proper threshold value of PDI difference to determine whether a pixel is irrigated or not is about 0.091 after comparing the monitoring results under different thresholds, taking 25 observed irrigation sites collected in June 2014.(3) It is found from the remote sensing obtained 2010 to 2014 planting structure in the irrigation district that the over cropping area is featured as corn> rice> spring wheat> winter wheat. Moreover, a clear trend shows that corn area increased rapidly in the five years, from 43.52% to 57.88%, with rice reduced from 35.84% to 29.13%, spring wheat reduced from 15.73% to 9.02%, winter wheat reduced from 4.91% to 3.96%. This undergoing adjustment of cropping structures by farmers is resulted from corn's relative higher prices, higher yield, good water saving, less labor needed, etc. |
PDF Full Text Request