| Drought is the major natural disaster as well as one of the most severe meteorological disasters throughout the world. Aridity seriously threatens to do harm to national economic, social development and national production and life. With the rapid development of satellite remote sensing technology, which has the characteristics of high timeliness, wide space range, objectivity, accuracy, low-cost and so on, its application to drought monitoring becomes trends and hot spots. So far, many drought monitoring models based on the visible, near infrared, far infrared and microwave remote sensing technology have been presented at home and abroad. Although each method has its advantages, some problems still exist:(1) Firstly, as the methods, using vegetation index and surface temperature, which are supported by the data of visible, infrared data, have a strong dependence on land cover type, they are poor in temporal-spatial adaptability. In addition, for drought monitoring based on microwave remote sensing, no mature algorithm is available and no drought index has been built yet. (2) Another problem is the lack of forward-based drought monitoring model using remote sensing technology. The current models are basically inversion-type ones, which have the effect of latency, because they are built by the various drought-induced land surface phenomena and natural features objects responses. (3) In addition, it is the lack of time sequence data-based drought monitoring methods in the time and frequency domain. As the methods that currently exist are basically spatial monitoring methods and mainly for drought monitoring that is based on single remote sensing images. They do not give full play to the advantages of high temporal resolution remote sensing data. (4) Finally, there is a problem of the lack of large-scale space-time adaptive method for drought monitoring based on remote sensing, as the current drought monitoring studies are conducted for small areas, various methods are deficient in regional adaptability, coefficients of regional influence need to be established. Accordingly, their value of practical application and development are limited. For the above-mentioned problems, the proposition of a large-scale space-time adaptive forward-based drought monitoring method based on time-sequence remote sensing data has an important practical significance for drought disaster prevention and mitigation, sustainable development of national economy and society. This dissertation focus on the following content and work:(1) The status of domestic and foreign remote sensing drought monitoring were comprehensively summarized, followed by the analysis for the existing problems and deficiencies in the field. (2) A detailed and exact definition of drought was presented, following on which is a summary of formation and characteristics of drought and a analysis of the impact of drought, taken the drought condition in China into consideration. (3) Existing methods of remote sensing drought monitoring were classified in detail. Then following on a systematical summary of the drought monitoring methods is a comparative analysis of the advantages and disadvantages of each method in sub-categories. (4) Based on the analysis of parameters which affect solar radiation model, after the factors of reflectivity and brightness temperature in clear sky been extracted, with the combination of remote sensing spectral information, the two functions affected by these factors were specifically analyzed and the clear sky brightness temperature and reflectivity calculation formula with normal dynamic range was constructed in consideration of the impact of each factor. Consequently adaptive cloud detection method was built with correlate experimental analysis and accuracy evaluation. (5) Directing to the deficiency of MODIS remote sensing drought monitoring model, the research of remote sensing drought monitoring model of cloud parameter method was done intergrated with site data. In addition, the quantitative accuracy assessment was conducted among and within the methods, following on which was a qualitative analysis of monitoring results of nearly two-year comparative large domestic drought events. (6) Relying on the analysis of the monitoring results of cloud parameters method in the long time series and the selection of the indicators for background values of cloud parameter, the research on the construction of the ambient field of cloud parameters in monitoring blind area was done and related expereiments were conducted combined with the site data.This dissertation made innovations in the following there aspects:(1)Accorded to the radiation spectrum curve of surface features under clear sky, a clear sky brightness temperature and reflectivity formulas were established and the dynamic range of the clear sky brightness temperature and reflectivity was proposed with the consideration of land cover types. Accordingly, an adaptive method of cloud detection was presented.(2) Aiming at the deficiency of MODIS remote sensing drought monitoring model, a successive model was constructed, which was modified in time and space scales to build cloud parameters drought monitoring model that was effectively varified by the experiments and practical operation in remote sensing drought monitoring.(3) Based on the analysis of monitoring results of cloud parameters model in longtime series, the method of establishment of ambient field of cloud parameters, using aselected indictor, was promoted in the interesting area which extracted from thismodel in China area. The experimental results demonstrate that the construction ofbackground field for cloud parameters contributed to improve monitoring accuracy inmonitoring blind area.According to the research of this dissertation, conclusions could be derived asfollows:(1) There is a direct relationship between the damage of drought and the conditionssuch as national produce and life activity, social economic development, national foodsecurity and even national political stability, so drought disaster prevention andmitigation using high technology is an indispensable component of a harmonioussociety. Compared to traditional ground-based observation of the meteorologicaldrought monitoring method, drought monitoring based on remote sensing technologyhas unparalleled advantages.(2) The results of cloud detection directly affects the accuracy of fruits from droughtmonitoring of cloud parameter model whose core is three cloud parameters. As for thelarge study area and long duration, the traditional cloud detection methods with fixedthreshold could hardly meet the accuracy requirements. Howerver, an adaptivemethod of cloud detection included the analysis of the variation of brightnesstemperature and reflectance under clear sky and the consideration of clouds'characteristics of high reflectivity and low brightness temperature. And relatedexperiment results demonstrated that this method was better, so that it could serve as asource of information for the cloud parameters drought monitoring model.(3) Drought, as a continuous process, whose occurrence is free from time and spaceconstraints and whose general scope is flaky, is diffucult to be monitored using remotesensing technology in a large-space and long-running scale. In this paper, aiming atthese problems, remote sensing drought monitoring model based on cloud parameterwas built by analyzing the correlation between clouds and drought, structuring thethree cloud parameters which are independent of each other and modifying the modelfrom time and space scales to enhance its adaptability in time and spatial scales. Theresults of experiment which based on FY-2 series of data, high time resolution of thedomestic geostationary meteorological satellite, showed that this model could getcomparative high precision in drought monitoring. (4) For the interseting area that abtained from the analysis of years of droughtmonitoring of cloud parameter model in some region, the studies on the constructionof ambient field of cloud parameters were done combined with real data and the fruitsof years of drought monitoring. The experimental results demonstrated that theconstruction of background field for cloud parameters contributed to improvemonitoring accuracy in monitoring blind area. However, in this paper, relatedexperiments were only carried out in subregion, so that the promotion of backgroundcloud parameters remains to be further studied.Keywords:drought monitoring, FY-2C/D/E, dynamic threshold, cloud detecting,cloud parameter, temporal and spatial modification, background field...
|
PDF Full Text Request