| 应用粒子滤波同化条件植被温度指数的土壤水分估测 |
| |
| 引用本文： | 解毅,王鹏新,李俐,荀兰,张树誉.应用粒子滤波同化条件植被温度指数的土壤水分估测[J].干旱地区农业研究,2018,36(3):236-243. |
| |
| 作者姓名： | 解毅 王鹏新 李俐 荀兰 张树誉 |
| |
| 作者单位： | 中国农业大学信息与电气工程学院;农业部农业灾害遥感重点实验室;陕西气象局 |
| |
| 基金项目： | 国家自然科学基金项目（41371390） |
| |
| 摘 要： | 为了准确地获取2013—2015年关中平原冬小麦主要生育期土壤含水量(0~20 cm)的时空信息,基于Landsat-8遥感数据反演条件植被温度指数(CVTI),并结合CVTI和实测土壤水分间的线性相关性构建土壤水分反演模型。应用粒子滤波(PF)算法同化基于CVTI反演的和CERES-Wheat模型模拟的土壤水分,得到以天为步长的土壤水分同化值,利用土壤水分实测值分别检验土壤水分模拟值、反演值和同化值的精度。结果表明,CVTI和实测土壤水分间的线性相关性显著,尤其在小麦拔节期和抽穗~灌浆期,其相关性达到极显著水平(P0.01);土壤水分同化值和实测值间的线性相关性(r=0.96,P0.001)大于土壤水分模拟值和实测值间的相关性(r=0.71,P0.01)以及土壤水分反演值和实测值间的相关性(r=0.89,P0.001);土壤水分同化值的均方根误差(RMSE)和平均相对误差(MRE)比土壤水分模拟值的RMSE和MRE分别降低了0.025 cm~3·cm~(-3)和2.70%,比土壤水分反演值的RMSE和MRE分别降低了0.016 cm~3·cm~(-3)和4.15%,同化过程提高了时间序列土壤含水量的估测精度。因此,基于CVTI和PF算法能够较为准确估测关中平原小麦主要生育期的土壤含水量。
|
| 关 键 词： | 条件植被温度指数（CVTI） 土壤含水量 小麦生育期 估测 作物生长模型 粒子滤波 关中平原 |
Soil moisture estimation by using particle filter assimilated conditional vegetation temperature index |
| |
| XIE Yi,WANG Peng-xin,LI Li,XUN Lan,ZHANG Shu-yu.Soil moisture estimation by using particle filter assimilated conditional vegetation temperature index[J].Agricultural Research in the Arid Areas,2018,36(3):236-243. |
| |
| Authors: | XIE Yi WANG Peng-xin LI Li XUN Lan ZHANG Shu-yu |
| |
| Abstract: | In order to exactly obtain the temporal-spatial information of soil moisture (0~20 cm) in main growing period of winter wheat in Guanzhong Plain during 2013 to 2015, in this paper, based on the Landsat-8 remote sensing data, inverted the conditional vegetation temperature index (CVTI), constructed the soil moisture inverting model combined with the linear correlation between CVTI and measured soil moisture. Using the particle filter (PF) algorithm was assimilated the soil moisture based on CVTI inversion and simulation of the CERES-Wheat model, obtained the assimilation of the soil moisture value with daily step length. The precision of the simulated , inverted and assimilated soil moisture values were inspected by the field-measured soil moisture respectively. The results were indicated that: The linear correlation between CVTI and field-measured soil moisture was notable, especially at jointing stage and heading-filling stage of wheat, the relativity was achieved extremely significant (P<0.01). The linear correlation between the assimilated and measured soil moisture (r=0.96, P<0.001) was large estimated than the correlation between the simulated and measured soil moisture (r=0.71, P<0.01) and the correlation between inverted and measured soil moisture (r=0.89, P<0.001). The root mean square errors (RMSEs) and mean relative errors (MREs) of the assimilated soil moisture was reduced 0.025 mm3·mm-3 and 2.70% respectively compared with the simulated soil moisture, and was reduced 0.016 mm3·mm-3 and 4.15% respectively compared with the inverted soil moisture. These results indicated that the assimilation process was improved estimating accuracy of the soil moisture time series. Therefore, based on the CVTI and PF algorithm, the soil moisture content in main growing period of wheat in Guanzhong Plain can be accurately estimated. |
| |
| Keywords: | conditional vegetation temperature index soil moisture wheat growing period estimation crop growth model particle filter Guanzhong Plain |
| 本文献已被 CNKI 等数据库收录！ |
| 点击此处可从《干旱地区农业研究》浏览原始摘要信息 |
| 点击此处可从《干旱地区农业研究》下载免费的PDF全文 |
|
