【作者】 卫炜；

【导师】 周清波； 毛克彪；

【作者基本信息】 中国农业科学院 ， 农业遥感， 2012， 硕士

【摘要】 土壤水分作为农田生态系统的重要组成部分，是农作物生长发育的基本条件，也是土壤内部化学、生物和物理过程中不可缺少的介质，影响着土壤的肥力。在农业科学研究中，土壤水分是研究农作物水分胁迫、进行旱情监测和农作物估产等的一个重要指标。因此，土壤水分监测在水循环规律研究、水资源合理利用、农田灌溉以及旱情灾害预报等领域具有重要的意义。使用遥感方法监测土壤水分具有宏观、高时效和经济性等特点，与传统的可见光和红外遥感相比较，微波遥感在估算土壤水分方面拥有非常明显的优势。主动和被动微波遥感存在优势互补的特点，二者的联合应用，特别是土壤表面有农作物覆盖的条件下，将有利于充分发挥各自的优势，简化土壤水分监测的过程和提高土壤水分估算的质量。多数情况下被动微波遥感被用来进行土壤水分绝对值的反演，而主动微波遥感常被用来进行土壤水分变化相对值的估算，这也是由它们各自的特点所决定的。无论是主动还是被动微波遥感，在估算土壤水分时都需要考虑地表粗糙度和植被覆盖的影响，并设法将这些影响加以消除。本研究在SMEX02土壤水分试验中的PALS系统观测及地面数据支持下，研究在农作物覆盖地区的主被动微波遥感联合监测土壤水分的方法。分别对基于PALS辐射计数据的被动微波遥感土壤水分反演方法和基于PALS雷达数据的主动微波遥感土壤水分变化探测法作了研究分析，在此基础上提出了基于PALS系统的主被动微波遥感联合监测土壤水分方法，并对所提出的联合监测方法在研究区进行了应用验证。本研究的主要成果包括以下几个方面：（1）利用AIEM模型针对SMEX02试验中的PALS辐射计的参数设置，通过模拟的地表辐射状况数据库，分析了在PALS辐射计的传感器参数设置及其试验区的土壤质地参数条件下，参数化的Qp半经验模型及运用该模型的土壤水分反演方法的适用性。通过结合“ωτ”模型来描述植被覆盖的影响，将在AMSR-E的C波段数据的基础上建立的参数化的Qp半经验模型及相应的土壤水分反演方法成功的应用到PALS辐射计的L波段数据上。（2）用AIEM模型按照PALS雷达的参数设置模拟了L波段（1.26GHz）时不同粗糙度条件下后向散射系数与土壤表面菲涅耳反射率的关系，发现不同粗糙度下后向散射系数与菲涅耳反射率之间都可以用一条过原点的直线来拟合，并且土壤水分变化（比值）与相应的菲涅耳反射率变化之间存在良好的线性关系。由此通过菲涅耳反射率在后向散射系数变化和土壤水分变化之间建立联系，将在全极化机载雷达（AirSAR）的L波段（1.2GHz）数据上发展而来的土壤水分变化探测方法也成功的应用于PALS雷达的L波段数据。（3）在研究和分析被动微波遥感土壤水分反演方法和主动微波遥感土壤水分变化探测法的基础上，提出了基于主被动微波遥感联合的土壤水分监测方法。通过对主被动微波遥感的联合应用，充分发挥了它们各自的优势，不需要大量获取植被含水量和地表温度等辅助数据就可以实现对农作物覆盖地区土壤水分的估算以及动态变化监测。基于主被动微波遥感联合的土壤水分监测方法不但简化了土壤水分监测的过程，而且在土壤水分估算精度上也有了一定的提高，为主被动微波遥感的联合应用提供了新的思路。更多还原

【Abstract】 As an important component of the farmland ecosystem, soil moisture is one of the essential requirements of crop growth. It also plays a unique role in the chemical, biological and physical processes within soil that influences the soil fertility. In agricultural science, soil moisture is a vital index of crop water stress, drought supervision and yield estimation researches. Therefore, soil moisture monitoring has significant value for the water cycle law research, water resource utilization, field irrigation and drought forecast, etc. Monitoring soil moisture by remote sensing can be considered as macroscopically, efficiently and economically. Compared with optical and infrared remote sensing, microwave method has obvious superiorities in soil moisture estimation. Active and passive microwave remote sensing are advantage complementary. It implies that the combined application of them would contribute to the maximization of their strengths and therefore simplify the process of soil moisture monitoring and improve the accuracy of soil moisture estimation, especially in the vegetation covered area.In most cases passive microwave remote sensing is used to retrieve soil moisture content while active microwave remote sensing is used to estimate soil moisture variation, which is decided by their characteristics. The effect of surface roughness and vegetation cover must be considered and managed to eliminate in soil moisture estimation no matter to the active or passive remote sensing. This dissertation devotes to the combined application of active and passive remote sensing in soil moisture monitoring for the vegetation covered area with the support of SMEX02PALS data. On the research and analysis of the soil moisture content retrieve method using PALS radiometer data and the soil moisture variation estimate method using PALS radar data, soil moisture monitoring method using the combined application of active and passive remote sensing is proposed. At last, the new method is applied and verified by using the SMEX02PALS and in situ experimental data. The results and innovations are mainly in the following aspects:(1) On the settings of PALS radiometer and conditions of SMEX02test region, a simulated radiation database is built by using the AIEM model. The applicability of the parameterized surface model Qp model and the soil moisture retrieve method developed on in the soil texture conditions of SMEX02test region and parameter settings of PALS radiometer it is tested by analyzing the database. With the " ω-τ " model to describe the effect of vegetation, the Qp model and the soil moisture retrieve method developed from the C-band data of AMSR-E are successfully applied to the L-band data of PALS radiometer.(2) The relationship between L-band (1.26GHz) backscattering coefficients and Fresnel reflectivities on the soil surface in different roughness conditions is simulated by using the AIEM model on the settings of PALS radar. The results show that the relationship can be fitted by a zero axial straight line, and a linear relationship also exists between the soil moisture variation (specific value) and the corresponding Fresnel reflectivity variation. The contact between backscattering coefficients variation and soil moisture variation is established through the Fresnel reflectivity. Therefore, the soil moisture variation estimate method developed from the L-band (1.2GHz) data of AirSAR is also successfully applied to the L-band data of PALS radar.(3) On the foundation of researching and analyzing of the soil moisture content retrieve and variation estimate method using passive and active microwave remote sensing respectively, the soil moisture monitoring method using the combined application of active and passive remote sensing is proposed. The strengths are maximized by the combination of them. Estimation of soil moisture and variation monitoring in the vegetation covered area can be done without massive ancillary data. The combined method propose in this dissertation not only simplifies the process of soil moisture monitoring but also improves it’s accuracy, which provides a new idea for the combination of active and passive remote sensing.更多还原