岷江上游生态水信息指标参数遥感反演与变化监测
发布时间：2019-01-24 19:38
【摘要】：森林植被作为地球生物圈的重要组成部分,与土壤、大气在多种维度上进行着多种形式的水分和热量的交换。在植被与水文的相互作用中,森林植被通过林地蒸散发、大气降水截留、枯落物截留、地表径流、土壤入渗、水分储存等水文过程与水文循环构成复杂的相互制约与反馈机制。生态水层以地球表面的植被层为中心,形成水循环过程中一个特殊的过渡带或缓存带,直接影响着水循环中各水资源量的滞留时间与动态分配过程。生态水层作为水循环圈层中一个相对独立的实体,在森林水文循环中扮演重要角色,但其资源量具有宏观性与空间分异性,尤其在生态与水文过程参数随空间分布剧烈变化的内陆河流域,通过常规方法难以有效获取其资源量。因此,如何有效利用遥感等新技术的多尺度分布式数据获取能力,实现对生态水的量化观测,为查明岷江上游流域地区的生态水资源量及其时空分布与变化特征提供理论与技术支持,对区域水循环研究具有重要的理论价值与实践意义。生态水前期研究表明,利用植被生物物理参数、土壤含水等因子在多波段、多时相遥感数据中的波谱特征宏观反映,能够对部分生态水参数建立遥感反演模型。但生态水前期研究侧重于光学遥感领域,在气候条件恶劣的地区,难以有效获取高质量光学遥感影像,对开展生态水研究尤其是其时空动态研究具有一定局限性。微波遥感具有全天候、全天时对地观测能力,近年来在水文研究中应用日益广泛,已在大气降水、土壤水分、积雪等水文变量以及叶面积指数、植被分类信息等植被生态参数的估算中取得一定成效。因此,本文基于前期生态水研究成果,以岷江上游毛尔盖地区为试验区,从光学遥感与微波遥感两个角度探索高原山区生态水参数的定量遥感反演方法,并通过多时相光学与微波遥感数据,实现对生态水部分参数的变化监测。本论文主要研究内容与创新成果如下:(1)利用时序雷达数据集建立了研究区土壤含水量估算模型。应用先进积分方程模型(AIEM)模拟了Wagner等提出的相对土壤水分估算模型的干湿端区间长度,建立了研究区的相对土壤水分估算模型,并将估算结果转换为土壤体积含水量。同时,基于较大范围地表粗糙度与入射角条件下的模拟结果,确定了C波段、L波段10dB近似区间长度,该结论对于长期干旱得不到湿端或长期湿涝得不到干端的地区具有现实意义。(2)发展了一种基于多时相雷达遥感影像的土地覆盖分类方法。基于7景合成孔径干涉雷达数据,采用多时相图像融合处理技术有效抑制了雷达影像的相干斑噪声,针对山区地形设计了一种基于升-降轨的双视向阴影补偿算法,改善了影像质量,利用后向散射系数、纹理特征、相干系数、数字高程4个特征参量实现了支持向量机监督分类,总分类精度为81.77%。(3)基于Van Genuchten土壤持水曲线模型,明确了土壤水饱和系数(SMS)的数学形式定义,并基于土壤体积含水量与雷达后向散射系数成强正相关且与尺度无关的研究结论,在假定模型各特定后向散射系数获取时的地表状况不变的前提下,推导了SMS的微波遥感反演模型。在此基础上,基于时序雷达数据集建模思路,建立了适合研究区地表状况的SMS微波遥感反演模型。(4)基于野外实测光谱与叶片等效水层厚度,建立了研究区基于复比植被指数(MSI/SR)的植被含水量光学遥感反演模型。采用温度植被干旱指数(TVDI),引入了地表温度高程校正模型,使用EVI替代NDVI,在此基础上构建了Ts/EVI空间,通过野外实测数据与TVDI的回归分析,建立了土壤含水量光学遥感反演模型。(5)采用多时相雷达遥感数据,对研究区主要地类土壤含水量(SMC)与土壤水饱和系数(SMS)进行了变化监测,结果显示总体上9月SMC与SMS比7月略有下降,不同植被类型的平均SMC与SMS呈现常绿林地-灌木林地-草地逐次降低;采用多时相光学遥感数据,实现了主要地类生态水涵养模数(MEC)的变化监测,结果显示总体上9月MEC比6月略有下降,不同植被类型的平均MEC呈现常绿林-灌木林-草地逐次降低。
[Abstract]:Forest vegetation, as an important part of the earth's biosphere, is in various dimensions of water and heat exchange with the soil and the atmosphere. In the interaction of the vegetation and the hydrology, the forest vegetation forms a complex mutual restriction and feedback mechanism through the hydrological processes and the hydrological cycle such as the evapotranspiration of the forest land, the interception of the atmospheric precipitation, the interception of the litter, the surface runoff, the infiltration of the soil, the water storage and the like. The ecological water layer is the center of the vegetation layer on the surface of the earth, and a special transitional zone or buffer zone is formed in the process of water circulation, and the retention time and the dynamic allocation process of the water resources in the water circulation are directly affected. As a relatively independent entity in the water circulation loop layer, the ecological water layer plays an important role in the hydrological cycle of the forest, but the amount of the ecological water layer has the characteristics of macroscopic and spatial heterogeneity, especially in the inland river basin with great changes of the ecological and hydrological process parameters with the space distribution, it is difficult to obtain the amount of resources effectively by the conventional method. Therefore, how to effectively utilize the multi-scale distributed data acquisition capability of new technologies such as remote sensing to realize the quantitative observation of the ecological water is to provide the theoretical and technical support for the identification of the ecological water resources, the spatial and spatial distribution and the change characteristics of the upper reaches of the Minjiang River, It is of great theoretical value and practical significance to study the regional water circulation. The preliminary study of the ecological water shows that the spectral characteristics of the soil water and other factors in the multi-band and multi-phase remote sensing data are reflected in the multi-band and multi-phase remote sensing data, and the remote sensing inversion model can be established for some of the ecological water parameters. However, the early research of ecological water has focused on the field of optical remote sensing, and it is difficult to obtain high-quality optical remote sensing images in areas with bad weather conditions, and has certain limitations on the research of ecological water, especially its space-time dynamic research. The microwave remote sensing has an all-weather and all-day observation capability. In recent years, the application of microwave remote sensing has been widely used in the study of hydrology. It has achieved some effect in the estimation of the hydrological variables such as atmospheric precipitation, soil moisture, snow and other vegetation ecological parameters such as leaf area index and vegetation classification information. Therefore, based on the research results of the early-stage ecological water, the quantitative remote sensing and inversion method of the ecological water parameters in the high-altitude mountain area is explored from two angles of optical remote sensing and microwave remote sensing based on the research results of the early-stage ecological water, and the multi-time phase optical and microwave remote sensing data are adopted, and the change monitoring of the partial parameters of the ecological water is realized. The main contents and innovations of this thesis are as follows: (1) The model of soil moisture content in the study area is established by using the time series radar data set. By using the advanced integral equation model (ADEM), the relative soil moisture estimation model proposed by Wagner et al. is simulated, and the relative soil moisture estimation model of the study area is established, and the estimated result is converted into the soil volume water content. At the same time, the approximate interval length of C-band and L-band is determined based on the simulation results under the condition of large-range surface roughness and incident angle. (2) The method of land cover classification based on multi-time-phase radar remote sensing image is developed. Based on the 7-view synthetic aperture interference radar data, the coherent speckle noise of the radar image is effectively suppressed by using the multi-time phase image fusion processing technology, a dual-view shadow compensation algorithm based on the lift-down rail is designed for mountainous terrain, the image quality is improved, and the backward scattering coefficient is utilized, The classification of the support vector machine is realized by the texture feature, the coherence coefficient, the digital elevation and the four characteristic parameters. The total classification accuracy is 81.77%. (3) Based on the water-holding curve model of the Van Genuchten soil, the mathematical form definition of the soil water saturation coefficient (SMS) is defined, and the study conclusion is based on the study conclusion that the soil volume water content is strongly positive correlation with the backward scattering coefficient of the radar and is not related to the scale, The microwave remote sensing inversion model of SMS is derived on the premise of the assumption that the surface condition of the model after each specific backward scattering coefficient is unchanged. On this basis, based on the time-series radar data set modeling method, an SMS microwave remote sensing inversion model suitable for the surface condition of the study area is established. (4) Based on the field measured spectrum and the thickness of the equivalent water layer of the blade, the optical remote sensing inversion model of the vegetation water content based on the complex ratio vegetation index (MSI/ SR) was established. Using the temperature vegetation drought index (TVDI), the surface temperature elevation correction model was introduced. Using the EVI instead of the NDVI, the Ts/ EVI space was constructed, and the soil moisture content optical remote sensing inversion model was established by the regression analysis of the field measured data and the TVDI. (5) The main soil moisture content (SMC) and soil water saturation coefficient (SMS) in the study area were monitored by multi-time-phase radar remote sensing data. The results showed that the SMC and SMS decreased slightly in September. The average SMC and SMS of different vegetation types show that the green forest land-forest land-grassland is gradually reduced, and the multi-time phase optical remote sensing data is adopted to realize the change monitoring of the main ecological water conservation modulus (MEC). The results show that the MEC has a slight decrease in September MEC than in June. The average MEC of different vegetation types showed a gradual decrease of the evergreen forest-deciduous forest-grassland.
【学位授予单位】：成都理工大学
【学位级别】：博士
【学位授予年份】：2016
【分类号】：P237;P332
，
本文编号：2414774
[Abstract]:Forest vegetation, as an important part of the earth's biosphere, is in various dimensions of water and heat exchange with the soil and the atmosphere. In the interaction of the vegetation and the hydrology, the forest vegetation forms a complex mutual restriction and feedback mechanism through the hydrological processes and the hydrological cycle such as the evapotranspiration of the forest land, the interception of the atmospheric precipitation, the interception of the litter, the surface runoff, the infiltration of the soil, the water storage and the like. The ecological water layer is the center of the vegetation layer on the surface of the earth, and a special transitional zone or buffer zone is formed in the process of water circulation, and the retention time and the dynamic allocation process of the water resources in the water circulation are directly affected. As a relatively independent entity in the water circulation loop layer, the ecological water layer plays an important role in the hydrological cycle of the forest, but the amount of the ecological water layer has the characteristics of macroscopic and spatial heterogeneity, especially in the inland river basin with great changes of the ecological and hydrological process parameters with the space distribution, it is difficult to obtain the amount of resources effectively by the conventional method. Therefore, how to effectively utilize the multi-scale distributed data acquisition capability of new technologies such as remote sensing to realize the quantitative observation of the ecological water is to provide the theoretical and technical support for the identification of the ecological water resources, the spatial and spatial distribution and the change characteristics of the upper reaches of the Minjiang River, It is of great theoretical value and practical significance to study the regional water circulation. The preliminary study of the ecological water shows that the spectral characteristics of the soil water and other factors in the multi-band and multi-phase remote sensing data are reflected in the multi-band and multi-phase remote sensing data, and the remote sensing inversion model can be established for some of the ecological water parameters. However, the early research of ecological water has focused on the field of optical remote sensing, and it is difficult to obtain high-quality optical remote sensing images in areas with bad weather conditions, and has certain limitations on the research of ecological water, especially its space-time dynamic research. The microwave remote sensing has an all-weather and all-day observation capability. In recent years, the application of microwave remote sensing has been widely used in the study of hydrology. It has achieved some effect in the estimation of the hydrological variables such as atmospheric precipitation, soil moisture, snow and other vegetation ecological parameters such as leaf area index and vegetation classification information. Therefore, based on the research results of the early-stage ecological water, the quantitative remote sensing and inversion method of the ecological water parameters in the high-altitude mountain area is explored from two angles of optical remote sensing and microwave remote sensing based on the research results of the early-stage ecological water, and the multi-time phase optical and microwave remote sensing data are adopted, and the change monitoring of the partial parameters of the ecological water is realized. The main contents and innovations of this thesis are as follows: (1) The model of soil moisture content in the study area is established by using the time series radar data set. By using the advanced integral equation model (ADEM), the relative soil moisture estimation model proposed by Wagner et al. is simulated, and the relative soil moisture estimation model of the study area is established, and the estimated result is converted into the soil volume water content. At the same time, the approximate interval length of C-band and L-band is determined based on the simulation results under the condition of large-range surface roughness and incident angle. (2) The method of land cover classification based on multi-time-phase radar remote sensing image is developed. Based on the 7-view synthetic aperture interference radar data, the coherent speckle noise of the radar image is effectively suppressed by using the multi-time phase image fusion processing technology, a dual-view shadow compensation algorithm based on the lift-down rail is designed for mountainous terrain, the image quality is improved, and the backward scattering coefficient is utilized, The classification of the support vector machine is realized by the texture feature, the coherence coefficient, the digital elevation and the four characteristic parameters. The total classification accuracy is 81.77%. (3) Based on the water-holding curve model of the Van Genuchten soil, the mathematical form definition of the soil water saturation coefficient (SMS) is defined, and the study conclusion is based on the study conclusion that the soil volume water content is strongly positive correlation with the backward scattering coefficient of the radar and is not related to the scale, The microwave remote sensing inversion model of SMS is derived on the premise of the assumption that the surface condition of the model after each specific backward scattering coefficient is unchanged. On this basis, based on the time-series radar data set modeling method, an SMS microwave remote sensing inversion model suitable for the surface condition of the study area is established. (4) Based on the field measured spectrum and the thickness of the equivalent water layer of the blade, the optical remote sensing inversion model of the vegetation water content based on the complex ratio vegetation index (MSI/ SR) was established. Using the temperature vegetation drought index (TVDI), the surface temperature elevation correction model was introduced. Using the EVI instead of the NDVI, the Ts/ EVI space was constructed, and the soil moisture content optical remote sensing inversion model was established by the regression analysis of the field measured data and the TVDI. (5) The main soil moisture content (SMC) and soil water saturation coefficient (SMS) in the study area were monitored by multi-time-phase radar remote sensing data. The results showed that the SMC and SMS decreased slightly in September. The average SMC and SMS of different vegetation types show that the green forest land-forest land-grassland is gradually reduced, and the multi-time phase optical remote sensing data is adopted to realize the change monitoring of the main ecological water conservation modulus (MEC). The results show that the MEC has a slight decrease in September MEC than in June. The average MEC of different vegetation types showed a gradual decrease of the evergreen forest-deciduous forest-grassland.
【学位授予单位】：成都理工大学
【学位级别】：博士
【学位授予年份】：2016
【分类号】：P237;P332
，
本文编号：2414774
本文链接：https://www.wllwen.com/shoufeilunwen/jckxbs/2414774.html
