基于Landsat 8生长时序遥感信息的玉米干旱监测研究

发布时间：2018-03-27 21:12

本文选题：Landsat　切入点：8　出处：《石河子大学》2017年硕士论文

【摘要】：【目的】干旱地区水分科学合理使用是影响作物生长及社会稳定的重要因素之一,作物干旱胁迫精准监测是实现精准灌溉的基础,是作物节本增效的途径。本研究利用Landsat8卫星数据分析玉米生育期干旱植被指数变化规律及其与农学干旱指标的关系,探索Landsat 8卫星数据应用于玉米生长期干旱监测的理论与方法,揭示研究区玉米干旱过程及旱情空间分布特征,为精准灌溉提供理论依据,提供实现农业节本增效的技术及方法。【方法】利用“3S”技术,通过大田调查与定点采样技术相结合,研究玉米生育期表征其干旱状况的农学参数以及基于Landsat 8数据的8种干旱遥感监测植被指数的定量关系,构建了新型的玉米干旱监测指数。在此基础上,将新型干旱监测植被指数应用于研究区2014-2016年的玉米生长期内干旱胁迫监测,提出在玉米整个生育期利用卫星数据进行干旱胁迫监测的方法。【结果】1.实现了基于Landsat 8卫星数据的大田玉米种植面积的精准监测,根据研究区作物的生长发育特点以及物候期的不同,利用决策树分类方法对研究区2014-2016年玉米种植面积进行了提取。2016年玉米种植面积提取精度高于2015年和2014年,2015年高于2014年,干旱程度不同对提取精度会造成一定的影响,水量充足年份,玉米长势一致有利于提高提取精度。2.分析比较了归一化植被指数(NDVI),归一化水分指数(NDWI),水胁迫指数(MSI1、MSI2),植被供水指数(VSWI),多波段干旱指数(MBDI),温度植被干旱指数(TVDI),并综合水胁迫指数(MSI2)和温度指数(LST)构建了新型的玉米干旱监测指数——冠层温度水分指数(CTWDI),并与地面调查的土壤相对含水量,玉米冠层含水量,叶绿素含量进行了相关性分析。结果表明:冠层温度水分指数CTWDI与玉米冠层含水量,叶绿素含量相关性最高,拟合方程决定系数也最高,用CTWDI监测玉米干旱状况具有一定的优势。3.玉米高产试验田是在充分水肥管理基础上实现的,其植被指数变化具有年际间的一致性,特别是在玉米生育期的中后期。利用监督分类方法,以高产田作为水分充足样本田,通过干旱植被指数的差异来判断相对于高产田的干旱程度。分析了2014-2016年研究区玉米生育期(5月-9月)内的NDVI、MSI2、CTWDI指数时间变化以及空间分布特点,得出3年内玉米生育期内的干旱风险评价指标。4.应用冠层温度水分指数(CTWDI)对研究区2014-2016年大田玉米干旱程度进行了分级,同时根据团场统计调查的产量数据对研究区玉米干旱遥感监测分级结果进行了精度检验。结果表明:2014年总体精度和Kappa系数分别为87.3%和0.84,2015年相对精度和Kappa系数分别为83%和0.71,2016年的分别为81.2%和0.57。越干旱年份,干旱监测结果越可靠。【结论】基于Landsat 8生长时序遥感信息对研究区玉米干旱监测结果较好,实现了利用遥感技术对研究区玉米的干旱监测。
[Abstract]:[Objective] the rational use of water in arid areas of science is one of the important factors affecting social stability and growth of crops, crop drought monitoring is the foundation to realize the precision of precision irrigation, is a way to crops efficiency. This study using the Landsat8 satellite data analysis arid vegetation index variation of maize and its relationship with agricultural drought index, exploration the application of Landsat 8 satellite data on maize growth theory and method of drought monitoring, revealing the distribution characteristics of drought process and spatial drought corn research area, to provide a theoretical basis for precision irrigation, provide implementation techniques and methods of agricultural high efficiency. [method] the use of "3S" technology, through investigation and field sampling point combination study on Maize agronomic parameters characterizing the drought conditions and 8 Drought Remote Sensing Monitoring Vegetation Index Based on Landsat 8 data The quantitative relationship between the number of constructed maize drought monitoring index model. On this basis, the model of Drought Monitoring Vegetation index used in the study area 2014-2016 years of maize growth period of drought stress monitoring method is proposed for monitoring drought stress in the whole growth period of maize using satellite data. [result] 1. to achieve a precise monitoring of Landsat 8 satellite data based on field corn planting area, according to the growth characteristics and different phenological periods of crop in study area, using the decision tree classification method to extract.2016 corn planting area extraction accuracy is higher than in 2015 and 2014 of 2014-2016 years of corn planting area in 2015 than in 2014, the degree of drought on the extraction accuracy will be different a certain impact, adequate water year, corn growing consistently helps to improve the extraction accuracy of.2. analysis and comparison of the normalized difference vegetation index The number (NDVI), normalized difference water index (NDWI), water stress index (MSI1, MSI2), vegetation water index (VSWI), multiband drought index (MBDI), the Temperature Vegetation Drought Index (TVDI), and the water stress index (MSI2) and temperature index (LST) and constructed a new type of maize drought the monitoring index of canopy temperature moisture index (CTWDI), and the ground survey of relative soil water content, canopy water content, chlorophyll content were analyzed. The results showed that the canopy temperature and moisture index CTWDI and maize canopy water content, the chlorophyll content of the highest correlation coefficient of fitting equation of decision is the highest, with CTWDI maize drought monitoring the status of.3. has the advantage of high yield of maize field test is based on a certain sufficient fertilizer and water management, consistent with the inter annual change of vegetation index, especially in maize growth in the late period. Using supervised classification method With sufficient water as high fields, like Honda, the differences in drought vegetation index to determine the degree of drought. Compared to the high yield field analysis of 2014-2016 years of the study area for maize growth period (May -9 months) in NDVI, MSI2, CTWDI index, time variation and spatial distribution of the 3 years of maize growth period drought risk assessment index.4. canopy temperature moisture index (CTWDI) of the study area 2014-2016 years of field corn drought degree were graded, and according to the accuracy test of maize drought monitoring by remote sensing classification results yield data farm survey. The results showed that in 2014 the overall accuracy and Kappa coefficient were 87.3% and 0.842015 years the relative accuracy and Kappa coefficient were 83% and 0.712016 years were 81.2% and 0.57. more drought, drought monitoring results more reliable. [Conclusion] Based on the Landsat 8 long The monitoring results of maize drought in the study area were better by the sequence remote sensing information, and the drought monitoring of Maize in the study area was realized by using remote sensing technology.

【学位授予单位】：石河子大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：S513;S127;S423

【参考文献】