本文選題：漬害 + 溫度-植被指數(shù)特征空間��； 參考：《華中師范大學(xué)》2016年碩士論文

【摘要】：漬害低產(chǎn)農(nóng)田在中國南方湖區(qū)、沿江平原廣泛分布。由于漬害對(duì)糧食生產(chǎn)影響嚴(yán)重,發(fā)生存在歷史悠久,一直是漬害低產(chǎn)田治理中的重要內(nèi)容。傳統(tǒng)對(duì)漬害的鑒定、識(shí)別是在實(shí)地調(diào)研、實(shí)驗(yàn)和測(cè)量的基礎(chǔ)上進(jìn)行的,費(fèi)時(shí)費(fèi)力。因此,迫切需要尋求一種快速有效的監(jiān)測(cè)識(shí)別的方法,而遙感技術(shù)就為此提供了技術(shù)支持。本研究選取江漢平原腹地四個(gè)縣市作為研究區(qū),利用遙感數(shù)據(jù)和DEM數(shù)據(jù)對(duì)研究區(qū)的漬害農(nóng)田進(jìn)行嘗試判定提取并進(jìn)行空間分析。研究結(jié)果可以為漬害農(nóng)田的快速識(shí)別、監(jiān)測(cè)提供一種方法,豐富了濕潤湖區(qū)平原的土壤水分含量的研究,具有重要的理論和現(xiàn)實(shí)意義。本研究選取2005年4月20日的Land sat 5 TM遙感影像數(shù)據(jù),在融合、裁剪等預(yù)處理的基礎(chǔ)上,進(jìn)行輻射定標(biāo)和大氣校正,并對(duì)校正前后地物波譜曲線變化進(jìn)行了分析；對(duì)研究區(qū)遙感影像進(jìn)行地物分類并進(jìn)行分類精度評(píng)價(jià)；在對(duì)地表真實(shí)溫度反演中,地表比輻射率的估算采用像元分解法。其次,通過LST-NDVI和LST-EVI構(gòu)建的特征空間對(duì)“干邊”和“濕邊”進(jìn)行擬合,可以發(fā)現(xiàn)構(gòu)建的LST-EVI特征空間具有明顯的三角形特征。因此,采用LST-EVI特征空間的擬合系數(shù)計(jì)算出研究區(qū)的TVDI。最后,在ArcGIS 10.0中對(duì)地物分類中的農(nóng)田像元、DEM數(shù)據(jù)和TVDI進(jìn)行空間分析,獲取漬害農(nóng)田像元；分析了漬害的三個(gè)主要影響因子：水系、地勢(shì)起伏度和高程,并對(duì)漬害農(nóng)田進(jìn)行了空間分布統(tǒng)計(jì)。本研究驗(yàn)證了漬害遙感識(shí)別原理的科學(xué)性、合理性、可行性和實(shí)踐性,為漬害農(nóng)田的遙感識(shí)別判定提供了一種快速有效的方法。本文的主要成果有：(1)利用LST-Ⅵ特征空間對(duì)漬害農(nóng)田進(jìn)行判定識(shí)別。LST-Ⅵ特征空間表征了土壤含水量、溫度和植被之間的關(guān)系,因此可以用于漬害研究。本研究提出和驗(yàn)證了該方法的可行性。(2)基于GIS空間分析,提出遙感識(shí)別判定漬害農(nóng)田的閾值標(biāo)準(zhǔn)：同時(shí)滿足0.12EVI0.31和23.1 mDEM34 m的農(nóng)田耕地像元；農(nóng)田耕地像元的TVDI滿足0TVDI0.45。(3)研究結(jié)果顯示研究區(qū)2005年漬害農(nóng)田的面積約為339.69km2,集中分布的高程范圍為24 m～29m。
[Abstract]:Waterlogging and low yield farmland are widely distributed in the lake region of southern China and the plain along the Yangtze River. Because of the serious impact of waterlogging on grain production and the existence of a long history, it has been an important content in the treatment of waterlogging and low yield fields. Traditional identification of stains is based on field investigation, experiment and measurement, which is time-consuming and laborious. Therefore, there is an urgent need to find a rapid and effective method of monitoring and identification, and remote sensing technology provides technical support for this. In this study, four counties and cities in the hinterland of Jianghan Plain were selected as the study areas. Remote sensing data and Dem data were used to determine and extract the waterlogged farmland in the study area and spatial analysis was carried out. The results can provide a method for rapid identification and monitoring of waterlogged farmland, and enrich the study of soil moisture content in humid lake plain, which has important theoretical and practical significance. In this study, the data of Land sat 5 TM remote sensing image from April 20, 2005 were selected to carry out radiometric calibration and atmospheric correction on the basis of pre-processing such as fusion and clipping, and the changes of spectral curve of ground object before and after correction were analyzed. The classification accuracy of the remote sensing image is evaluated and the pixel decomposition method is used to estimate the surface specific emissivity in the inversion of the real surface temperature. Secondly, by fitting "dry edge" and "wet edge" with the feature space constructed by LST-NDVI and LST-EVI, we can find that the constructed LST-EVI feature space has obvious triangle characteristics. Therefore, the fitting coefficients of LST-EVI feature space are used to calculate TVDI in the study area. Finally, in ArcGIS 10.0, the field pixel Dem data and TVDI were analyzed in ArcGIS 10.0, and the three main influencing factors of waterlogging were analyzed: water system, topographic fluctuation and elevation. The spatial distribution of waterlogged farmland was analyzed. This study verifies the scientific rationality feasibility and practicality of the principle of remote sensing identification of waterlogging damage and provides a rapid and effective method for the identification of waterlogging damage farmland. The main results of this paper are as follows: (1) the relationship among soil moisture content, temperature and vegetation is characterized by LST- 鈪，

本文編號(hào)：2054081