Knowledge Resource Center for Ecological Environment in Arid Area
| 基于中高分辨率遥感的地表植被参数反演算法研究 | |
| 其他题名 | The algorithm research of land surface vegetation parameter retrieval based on mid-high resolution remote sensing |
| 赵岩 | |
| 出版年 | 2012 |
| 学位类型 | 硕士 |
| 导师 | 王思远 |
| 学位授予单位 | 中国科学院大学 |
| 中文摘要 | 本文针对中高分辨率遥感数据进行地表植被参数反演模型的研究，主要包括植被指数、叶面积指数以及地表温度，这些参数广泛应用于地表植被分析，具有重要的研究价值。\n本文选择ASTER数据作为试验数据，以黑河流域四个重点实验区之一的中游旱区水文试验区的两个加密观测区-花寨子和盈科绿洲为主要研究区域。\n首先，本文对地表植被遥感反演的数据预处理关键环节-辐射定标、大气校正进行了深入研究，详细讨论了其处理流程和关键技术，明确提出了大气层顶平均太阳光谱辐照度的量化公式及其推演过程解释，并验证了公式的正确性，同时试验了FLAASH和6S两种大气校正方法，结果表明，这两种大气校正方法都能较好地去除大气以及环境对地物反射率的影响，特别是去除以2.66微米和2.74微米为中心的大气水汽吸收带的影响，效果尤为明显。\n针对植被指数反演模型，本文采用归一化植被指数(NDVI)，并利用其估算植被覆盖度，以研究试验区域植被覆盖度变化以及植被空间分布格局并分析其成因。在估算植被覆盖度时，提出一种结合土地覆盖调查数据资料确定反演模型相关参数的方法，并对比了ASTER与TM数据反演估算植被覆盖度的结果。\n针对叶面积指数反演模型，本文提出通过估算GVI，结合MODIS LAI产品，选择均质区域作为回归分析和算法验证样点进行回归分析，并得到统计方程，再将该统计方程应用于算法验证样点，最后进行精度评价，给出各评价指标的具体数值，并根据GVI数值分组，结合相关误差给出误差修正公式进行数据修正。\n针对地表温度反演模型，本文利用ASTER 13、14波段和分裂窗算法反演地表温度，并提出使用NDVI间接估算地表比辐射率，采用改进后的大气水汽含量反演模型估算大气水汽含量并根据统计方程估算大气透过率。同时结合MODIS LST/LSE产品，选择均质区域作为样点进行精度评价，给出各评价指标的具体数值，并根据样点类型分类，结合相关误差给出误差修正公式进行数据修正。\n最后，针对基于中高分辨率遥感数据的产品反演生成系统项目需求以及遥感数据特点，本文研究了用于生成陆地、大气、海洋参数产品的系统构建，提出了一系列关键技术解决系统构建过程中出现的问题，建立了基于中高分辨率遥感数据的陆地、大气、海洋参数产品反演原型系统。 |
| 英文摘要 | Based on the information obtained by the airborne remote sensing system sensors, research is done on vegetation parameters retrieval model for mid to high resolution remote sensing data. The research mainly focuses on vegetation index, leaf area index and surface temperature. These parameters, which are of great importance, are widely used in vegetation analysis.\nASTER data is used as test data. Huzhaizi and YingkeluIu, two of the high frequent observed areas, are chosen to do research on. Located in the middle reaches’, they are dry land in the hydrologic test area, which is one of the four key experiment regions in the Heihe river basin.\n\n\nFirst, in-depth study is done on the radiometric calibration and atmospheric correction, which is the key link in vegetation remote sensing data retrieval pre-processing. The processing and key techniques are discussed in detail. The average solar spectral irradiance quantitative formula of the top of the atmosphere is put forward, explained and verified. And also, FLAASH and 6S atmospheric correction method is tested. The results showed that these two methods can remove the impact caused by atmosphere and environment on the surface reflectance, particularly the effect caused by 2.66 micron and 2.74 micron atmospheric water vapor absorption band.\nIn the vegetation index retrieval model, normalized difference vegetation index (NDVI) is applied. NDVI is used to estimate vegetation coverage in order to study the vegetation coverage changes in the test area and the vegetation spatial distribution structure, and analyzed the causes. When estimating vegetation coverage, a method which is used to determine corresponding parameters in the inverse model, is proposed by integrating the land coverage survey data. The retrieval results received from ASTER and TM data is compared.\n\nFor the leaf area index retrieval model, this thesis proposes to estimate greenness vegetation index (GVI) by combining MODIS LAI products. First, regression analysis is done by selecting homogeneous regions as the regression analysis and algorithm validation sample points; second, the statistical equations received from the first step is used to validate the sample points in the algorithm; then, accuracy evaluation is done to give the specific values for each evaluation index; finally, the evaluation results are categorized according to the GVI value, and data correction is done using the error correction formula retrieved by combining with corresponding errors.\n\nThe land surface temperature retrieval model uses the ASTER 13, 14 band and the split window algorithm to retrieve land surface temperature. The use of NDVI to indirectly estimate the surface emissivity is put forward. The improved atmospheric water vapor content retrieval model is used to estimate atmospheric water vapor content, and the atmospheric transmittance is estimated according to the statistical equation. With the use of MODIS LST/LSE product, homogeneous regions are selected as the sample points for the accuracy evaluation, and then the specific values is calculated for each evaluation index. Sample points are classified according to their type. Last, data correction is done using the error correction formula retrieved by combining with corresponding errors.\nLast, according to the mid-high remote sensing data retrieval product generation systems’ requirements, as well as the features of remote sensing data, this paper analyzed the system construction techiniques applied in terrestrial, atmospheric and oceanographic parameters products, and a series of methods are proposed to solve the problems occurred during the system construction process. A mid-high resolution remote sensing data retrieval prototype system for land, atmosphere and ocean parameter product is built. |
| 中文关键词 | 中高分辨率 ; 地表植被 ; 数据预处理 ; 参数反演模型 ; 系统构建 |
| 英文关键词 | mid-high resolution remote sensing land surface vegetation data preprocessing parameter retrieval model system construction |
| 语种 | 中文 |
| 国家 | 中国 |
| 来源学科分类 | 信号与信息处理 |
| 来源机构 | 中国科学院对地观测与数字地球科学中心 |
| 资源类型 | 学位论文 |
| 条目标识符 | http://119.78.100.177/qdio/handle/2XILL650/287069 |
| 推荐引用方式 GB/T 7714 | 赵岩. 基于中高分辨率遥感的地表植被参数反演算法研究[D]. 中国科学院大学,2012. |
| 条目包含的文件 | 条目无相关文件。 | |||||
| 个性服务 |
| 推荐该条目 |
| 保存到收藏夹 |
| 导出为Endnote文件 |
| 谷歌学术 |
| 谷歌学术中相似的文章 |
| [赵岩]的文章 |
| 百度学术 |
| 百度学术中相似的文章 |
| [赵岩]的文章 |
| 必应学术 |
| 必应学术中相似的文章 |
| [赵岩]的文章 |
| 相关权益政策 |
| 暂无数据 |
| 收藏/分享 |
除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。
