| In recent years,with the global warming,agricultural drought has become one of the most serious natural disasters.As an important grain production base in China,the southern plain area of Hebei province has been an agricultural drought-prone area for many years.Therefore,the accurate monitoring and assessment of the occurrence law of agricultural drought events in this region is of great significance for the regional water supply and demand regulation,efficient use of water resources,timely and appropriate agricultural irrigation,drought prevention and disaster reduction.Under this situation,accurate monitoring and evaluation of the occurrence of agricultural drought in the region is of great significance for regional water resources supply and demand regulation,efficient utilization of water resources,timely and appropriate irrigation of agriculture,drought prevention and disaster reduction.Based on multi-source remote sensing data,meteorological data and field survey data,and the comprehensive application of mathematics,agricultural water,geographical information and remote sensing and other multidisciplinary technical means,this paper aims to accurately obtain the spatio-temporal information related to agricultural drought.The main research results and findings are as follows:(1)Based on generalized DEM,a new model of crop planting structure and spatial-temporal distribution was proposed.Based on GF 1,Landsat,MODIS NDVI and other multi-source remote sensing data and ground sample data,the monitoring classification,decision tree method and other crops spatial-temporal distribution information extraction models were systematically compared and analyzed.On this basis,considering the spectral information of remote sensing and phenological characteristics of crops,a new crop planting structure extraction model based on generalized DEM was proposed by referring to the construction idea of digital elevation model.The comparison and analysis of the experimental results show that the model has high extraction accuracy,simple calculation,low data requirement and easy programming under the same data source conditions.Based on this idea,this paper develops a set of software for extracting information of crop planting structure and its spatial and temporal distribution based on IDL language.(2)Based on GIS spatial analysis technology,an algorithm was proposed to extract the irrigation area and spatial and temporal distribution information from coupled land surface temperature(LST)and vegetation supply water index(VSWI).With the extraction results of agricultural planting structure(winter wheat)as the spatial constraint,based on GIS spatial superposition algorithm,coupled with the above two index extraction results,the irrigation area of Shijin irrigation area was extracted.The accuracy was verified by field investigation data and the extraction accuracy of irrigation area could reach over 80%.The irrigation water consumption was estimated according to the planting structure,irrigation quota and irrigation methods before and after regional water saving in Shijin irrigation area.(3)In order to solve the problem of the time lag effect of crop growth to the change of meteorological factors,a kind of coupled Almon time lag model and partial least square(PLS)method were proposed to detect and calculate the lag step size and the weight.The experimental results show that meteorological factors do have a time lag effect on the growth of crops,and the time lag is generally about three months.The correlation coefficient between NDVI based on temperature or precipitation regression and real NDVI was significant at P<0.01.Based on the calculation results of land type data and influence weight,the positive and negative influences were defined,and eight quantitative evaluation indexes were divided to conduct a quantitative study on the influence degree of each lag month in the spring in the research area.The research results have important guiding significance for the timely scheduling of regional crops.(4)Put forward the idea of constructing drought monitoring models considering different land types,and construct temperature vegetation drought index(TVDI)and Temperature water supply drought index(TSWDI)drought monitoring models in the plain area of south Hebei province.Base on MODIS EVI,meteorological data,soil relative humidity and other data sources,this paper studied the crop spring drought in the plain area of south Hebei province from 2000 to 2016,and the drought monitoring effects of two drought models were compared and analyzed.The results confirmed that TVDI considering the influence of land type had a higher correlation with soil relative humidity,and the results were more reliable.Based on the analysis of the advantages and disadvantages of TVDI and VSWI drought monitoring factors,the TSWDI model based on weight optimization was proposed,and the drought classification standard of the plain area in south Hebei was established by combining with soil relative humidity.Compared with TVDI model,this model has higher correlation with soil relative humidity,which is suitable for agricultural drought monitoring. |
PDF Full Text Request