| 引用本文: | 张晓春,刘海若,严忆辉,等.结合物候特征的农田土壤表层含水率遥感反演[J].灌溉排水学报,0,():-. |
| zhang xiaochun,liu hairuo,yan yihui,et al.结合物候特征的农田土壤表层含水率遥感反演[J].灌溉排水学报,0,():-. |
|
| |
|
|
| 本文已被：浏览 1435次 下载 0次 |
|
|
|
| 结合物候特征的农田土壤表层含水率遥感反演 |
|
张晓春1, 刘海若1, 严忆辉1, 唐蓉1, 张煜2
|
|
1.武汉大学水资源与水电工程科学国家重点实验室;2.长江水利委员会长江科学院 空间信息技术应用研究所
|
|
| 摘要: |
| 【目的】土壤水分是作物生长的必要条件，反演土壤含水率对监测土壤墒情、减少作物受渍害程度、提高作物产量具有重要意义。【方法】建立了一种结合作物物候特征的农田土壤表层含水率遥感反演方法。首先基于预处理后的Sentinel-1 SAR影像获得入射角数据和VV极化条件下的总后向散射系数，并从环境卫星影像提取相应日期的作物植被覆盖指数NDVI（Normalized Difference Vegetation Index）。然后基于入射角、后向散射系数和NDVI，通过水云模型得到去除植被影响后的后向散射系数。结合研究区内作物物候特征，分生育阶段建立后向散射系数与土壤相对体积含水率之间的关系，并建立适用于研究区域作物物候特征的反演模型。【结果】根据2种轮作作物冬小麦和夏玉米物候特征分为4个时间段函数，分别为冬小麦的播种—分蘖期、越冬期、返青期—成熟期，和夏玉米的全生育期。分生育期阶段选取部分实测点土壤含水率数据进行计算，得到各时间段的土壤含水率与后向散射系数关系的经验方程，其线性相关系数分别达到0.40、0.80、0.91和0.79。通过提取固镇县与利辛县的作物种植结构，结合经验方程反演得到研究区农田土壤表层含水率分布图。基于剩余部分实测土壤含水率数据分别对反演模型进行验证，结果显示冬小麦全生育期总的复相关系数为0.73，夏玉米全生育期总的复相关系数为0.82。【结论】通过结合物候特征分生育阶段建立函数，为农作物种植区域的土壤表层含水率反演提供一种快速准确的方法。 |
| 关键词: 土壤含水率；遥感反演；后向散射系数；物候特征；农作物区 |
| DOI： |
| 分类号:TV93 |
| 基金项目:国家重点研发计划项目(2018YFC1508302,2018YFC1508301)；国家自然科学基金（31871516）；湖北省自然科学基金(编号:2019CFB507)。 |
|
| Remote Sensing Retrieval of Soil Surface Moisture Content in Farmland Combined with Phenological Characteristics |
|
zhang xiaochun1, liu hairuo1, yan yihui1, tang rong1, zhang yu2
|
|
1.State Key Laboratory of Water Resources and Hydropower Engineering Science, Wuhan University;2.Spatial Information Technology Application Research Department, Changjiang River Scientific Research Institute
|
| Abstract: |
| 【Objective】Soil moisture is a necessary condition for crop growth. Soil moisture can not only affect various nutrients of soil dissolution, transfer and microbial activities, but also reflect local climate, vegetation, topography, soil texture and other natural conditions. The research of soil moisture content is widely used in many fields, such as ecological environment assessment, irrigation system compilation, waterlogging disaster monitoring and precision agriculture, etc. In meteorological standards, soil moisture content is the key parameter for monitoring soil moisture and analyzing the degree of crop waterlogging damage. At the same time, monitoring soil water content based on remote sensing images has the advantages of large monitoring time and space, fast analysis speed, economy and high efficiency. Therefore, the inversion of soil moisture content is of great significance for monitoring soil moisture content, reducing crop waterlogging damage and increasing crop yield. 【Method】A remote sensing retrieval method of farmland soil surface water content combined with crop phenological characteristics was adopted. Firstly, the incident angle data and the total backscattering coefficient under VV polarization are obtained based on the preprocessed Sentinel-1 SAR image, and the normalized difference vegetation index (NDVI) of crops at the corresponding date was extracted from the environmental satellite image. Then, based on the incident angle, backscattering coefficient and NDVI, the backscattering coefficient after removing the influence of vegetation was obtained through the water cloud model. According to the phenological characteristics of crops in the study area, the relationship between backscattering coefficient and soil relative volume moisture content was established in different growth stages, and an inversion model suitable for studying the phenological characteristics of crops in the study area was established. 【Result】According to the phenological characteristics of winter wheat and summer maize, the two rotation crops can be divided into four time period functions, which are sowing-tillering stage, overwintering stage, turning green stage-mature stage of winter wheat, and the whole growth period of summer maize. Then, the empirical equation of the relationship between soil moisture content and backscattering coefficient in each time period was obtained by calculating the data of soil moisture at some measured points in different growth stages, and the linear correlation coefficients were 0.40, 0.80, 0.91 and 0.79 respectively. By extracting the crop planting structure of Guzhen and Lixin, and combining with empirical equation inversion, the distribution map of surface water content of farmland soil in the study area was obtained. The inversion model was verified based on the remaining measured soil water content data, and the results showed that the total complex correlation coefficient of winter wheat during the whole growth period was 0.73, and that of summer maize during the whole growth period was 0.82. 【Conclusion】By establishing the function according to the phenological characteristics in different growth stages, it provides a fast and accurate method for inversion of soil surface water content in crop planting areas, and the results have good simulation accuracy in general rainfall years. From the results, this method is effective. |
| Key words: Soil moisture; Remote sensing inversion; Backscattering coefficient; Phenological characteristics; Crop region |
|
|
|
|
