中高分辨率遥感协同反演冬小麦覆盖度
发布时间：2018-08-22 18:08
【摘要】：为了开展高精度、高时空分辨率的植被覆盖度(fraction vegetation cover,FVC)监测,该文以华北地区冬小麦地为研究对象,采用4期高分一号卫星多光谱(GF1-PMS)、多光谱宽幅(GF1-WFV)与环境一号卫星多光谱(HJ1-CCD)3种传感器同期影像数据集,基于像元二分法模型,研究多源中高分辨率遥感影像协同估算FVC方法。以基于高空间分辨率GF1-PMS影像反演的FVC作为检验数据,对单源直接获取法、多源全生育期法、多源分期法3种反演模型进行了分析比较。研究结果表明:HJ1-CCD、GF1-WFV数据与GF1-PMS数据的FVC直接反演结果具有较高的一致性,但在冬小麦的初期生长阶段,受卫星观测角度效应的影响,GF1-WFV与HJ1-CCD的FVC结果偏高,偏差随冬小麦的成熟封垄而逐渐减弱;多源分期法的时空反演得到的FVC精度最高,GF1-WFV的决定系数为0.984,均方根误差为0.030;HJ1-CCD的决定系数为0.978,均方根误差为0.034;而在缺少GF1-PMS匹配数据时,可通过多源全生育期法提高GF1-WFV与HJ1-CCD数据的反演精度,GF1-WFV的决定系数为0.964,均方根误差为0.044;HJ1-CCD的决定系数为0.950,均方根误差为0.052。通过多传感器的联合反演获取时间序列的高精度的FVC数据,可为研究植被生长状况及生态环境动态变化提供数据基础。
[Abstract]:In order to monitor vegetation coverage with high precision and high spatial and temporal resolution, the winter wheat field in North China was studied in this paper. Based on pixel dichotomy model, the method of multi-source and high-resolution remote sensing image co-estimation of FVC was studied by using the simultaneous image data sets of four high-fractional-one satellite multispectral (GF1-PMS), multi-spectral wide amplitude (GF1-WFV) and environmental satellite multispectral (HJ1-CCD) sensors. Taking FVC inversion based on high spatial resolution GF1-PMS image as the test data, three kinds of inversion models, namely single source direct acquisition method, multi-source full growth period method and multi-source staging method, are analyzed and compared. The results show that there is a high consistency between the FVC direct inversion results of the GF1-PMS data and the GF1-PMS data. However, in the initial growth stage of winter wheat, the FVC results of GF1-WFV and HJ1-CCD are higher than those of HJ1-CCD due to the influence of satellite observation angle effect. The deviation decreases gradually with the mature ridging of winter wheat, the determination coefficient of FVC is 0.984, the root mean square error is 0.030 HJ1-CCD is 0.978, and the root mean square error is 0.034, but in the absence of GF1-PMS matching data, the determination coefficient of GF1-WFV is 0.978, and the root mean square error is 0.034 in the absence of GF1-PMS matching data. The inversion accuracy of GF1-WFV and HJ1-CCD data can be improved by multi-source whole growth period method. The determination coefficient of GF1-WFV is 0.964, the RMS error is 0.044 HJ1-CCD is 0.950, and the root mean square error is 0.052. The high precision FVC data of time series can be obtained by the joint inversion of multi-sensors, which can provide a data basis for the study of vegetation growth and the dynamic changes of ecological environment.
【作者单位】： 遥感科学国家重点实验室北京师范大学地理科学部;环境保护部卫星环境应用中心;北京林业大学精准林业北京市重点实验室;
【基金】：国家重点研发计划(2016YFD0800903)
【分类号】：S127;S512.11
本文编号：2197896
[Abstract]:In order to monitor vegetation coverage with high precision and high spatial and temporal resolution, the winter wheat field in North China was studied in this paper. Based on pixel dichotomy model, the method of multi-source and high-resolution remote sensing image co-estimation of FVC was studied by using the simultaneous image data sets of four high-fractional-one satellite multispectral (GF1-PMS), multi-spectral wide amplitude (GF1-WFV) and environmental satellite multispectral (HJ1-CCD) sensors. Taking FVC inversion based on high spatial resolution GF1-PMS image as the test data, three kinds of inversion models, namely single source direct acquisition method, multi-source full growth period method and multi-source staging method, are analyzed and compared. The results show that there is a high consistency between the FVC direct inversion results of the GF1-PMS data and the GF1-PMS data. However, in the initial growth stage of winter wheat, the FVC results of GF1-WFV and HJ1-CCD are higher than those of HJ1-CCD due to the influence of satellite observation angle effect. The deviation decreases gradually with the mature ridging of winter wheat, the determination coefficient of FVC is 0.984, the root mean square error is 0.030 HJ1-CCD is 0.978, and the root mean square error is 0.034, but in the absence of GF1-PMS matching data, the determination coefficient of GF1-WFV is 0.978, and the root mean square error is 0.034 in the absence of GF1-PMS matching data. The inversion accuracy of GF1-WFV and HJ1-CCD data can be improved by multi-source whole growth period method. The determination coefficient of GF1-WFV is 0.964, the RMS error is 0.044 HJ1-CCD is 0.950, and the root mean square error is 0.052. The high precision FVC data of time series can be obtained by the joint inversion of multi-sensors, which can provide a data basis for the study of vegetation growth and the dynamic changes of ecological environment.
【作者单位】： 遥感科学国家重点实验室北京师范大学地理科学部;环境保护部卫星环境应用中心;北京林业大学精准林业北京市重点实验室;
【基金】：国家重点研发计划(2016YFD0800903)
【分类号】：S127;S512.11
【相似文献】
相关期刊论文 前2条
1 刘建华,胡光道;高分辨率遥感影像的土地利用分类[J];资源开发与市场;2004年04期
2 陈杰;陈铁桥;梅小明;邵权斌;邓敏;;基于最优尺度选择的高分辨率遥感影像丘陵农田提取[J];农业工程学报;2014年05期
，本文编号：2197896
本文链接：https://www.wllwen.com/kejilunwen/nykj/2197896.html
