共查询到17条相似文献,搜索用时 140 毫秒
1.
基于高光谱遥感反射比的太湖水体叶绿素a含量估算模型 总被引:19,自引:1,他引:19
旨在寻找叶绿素a的高光谱遥感敏感波段并建立其定量估算模型。通过对太湖水体的连续监测,获得了从2004年6月到8月3个月的太湖水体高光谱数据和水质化学分析数据。利用实测的高光谱数据分析计算太湖水体的离水辐亮度和遥感反射比;然后,通过相关分析寻找反演叶绿素a浓度的高光谱敏感波段,进而建立反演太湖水体叶绿素a浓度的高光谱遥感定量估算模型,并用相关数据对模型进行精度分析。研究发现,水体的遥感反射比光谱在719nm和725nm存在两个峰,其中719nm处的峰更明显且稳定。通过模型的对比分析,发现用这两个峰值处的遥感反射比参与建模可以提高叶绿素a的估算精度;并且认为由反射比比值变量R719/R670所建立的线性模型对叶绿素a浓度的估算精度最理想。 相似文献
2.
应用MODIS数据反演河北省海域叶绿素a浓度 总被引:6,自引:0,他引:6
为了建立更加合理、准确的叶绿素a遥感反演模型,利用地物光谱仪测定了河北省海域水面的光谱反射率,分析了光谱反射率与实测叶绿素a浓度之间的关系.在此基础上,通过MODIS数据各波段及波段组合的反射率与实测叶绿素a浓度的相关分析,确定第1波段(B1)为最佳反演波段,建立了应用B1反演叶绿素a浓度的遥感模型,并对模型精度进行验证.结果表明:该模型相关系数为0.66,反演结果均方根误差为0.48 mg/m3,模型精度优于SeaDAS的OC3标准经验算法;该模型反演河北省海域表层水体的叶绿素a浓度有较好的效果. 相似文献
3.
鄱阳湖叶绿素a浓度遥感定量模型研究 总被引:1,自引:0,他引:1
叶绿素a浓度是反映湖泊水体营养状况的重要指标,本研究通过分析水体叶绿素a浓度与高光谱反射特征的相互关系,采用一阶微分值和峰值比值法分别建立了叶绿素a的高光谱定量反演模型,在此基础上与同步MODIS数据敏感波段建立卫星定量反演模型。结果表明:叶绿素a荧光峰出现在波段690nm-700nm,波段696nm一阶微分值相关系数最大;波段700nm与波段680nm的比值与其对数相关性较好,MODIS数据波段2和波段1比值的指数模型为最佳的回归模型。 相似文献
4.
目前, 针对太湖水体的叶绿素波段敏感性的分析, 大多集中在实测的高光谱反射率数据或者图像提取反射率与叶绿素浓度的统计分析结果上, 缺乏基于水体光学特性的研究, 并且两者之间的一致性也一直缺少论证。研究中采用2004 年4 月和2007 年8 月的两期数据, 首先从水质参数的生物光学特性入手, 基于生物光学模型, 利用叶绿素和其他水质参数的吸收和后向散射系数, 模拟计算其他水质参数不变, 叶绿素浓度处于不同水平时的水面反射率, 分析实测反射率对叶绿素浓度变化的响应; 利用MODIS 的波段响应函数把实测光谱模拟成宽波段的MODIS 反射率, 以此作为桥梁进而对实测的高光谱反射率和MODIS 图像提取反射率与叶绿素浓度的相关程度的一致性进行分析, 为利用MODIS 图像进行水质参数反演时的反演因子选择提供了依据。 相似文献
5.
6.
应用TM数据估算沿岸海水表层时绿素浓度模型研究 总被引:3,自引:0,他引:3
本研究以大亚湾为实验区,以陆地卫星TM数据为信息源,结合表层海水叶绿素浓度实测资料建立模型。在对叶绿素光谱特征及遥感估算叶绿素浓度机理研究基础上,选取了TMI—TM4波段的75种波段组合为子因素,以叶绿素浓度为母因素,利用灰色系统理论,分析各波段组合与叶绿素浓度之间的关联度。将关联度最大的5种波段组合分别建模,得到5个估算表层海水叶绿素浓度的反演模型。误差分析表明,各模型的最大相对误差在19%以下,平均绝对相对误差在11.2%以下,相对标准误差在6.7%以下,模型精度较高。研究表明:(TM3×TM4)是估算沿岸海水表层叶绿素浓度的最佳波段组合,采用(TM3×TM4)与TM1、TM2或ln(TM十TM2)、In(TM1×TM2)之比值并不能改善估算精度。 相似文献
7.
应用实测光谱估测太湖梅梁湾附近水体叶绿素浓度 总被引:40,自引:1,他引:40
遥感方法测定水体中叶绿素含量的核心问题是建立遥感数据和叶绿素含量的定量关系;利用太湖梅梁湾附近水体的实测光谱和水质采样实验室分析,从数量上揭示了位于682nm附近和706nm附近对叶绿素含量估测最重要的两个光谱特征,分别通过比值(R706/R682、R706/R572)、微分、面积、峰高、峰谷距离以及反射峰位置等建立与叶绿素浓度的线性或非线性相关回归模型,通过R2、平均误差以及RMS误差等的分析对比,认为比值和反射峰位置对叶绿素浓度有很好的指示作用,是估测太湖梅梁湾附近水体叶绿素浓度的最好方法,其中与反射峰位置的指数关系R2达0.9199,与R706/R682的直接线性关系R2达0.9038。相对于706nm附近的峰谷距离、峰高以及反射峰面积而言,反射峰的位置是指示叶绿素浓度最敏感的变量;572nm附近、682nm附近以及706nm附近等三个波段对探测叶绿素a具有重要作用。 相似文献
8.
9.
10.
11.
12.
不同季相针叶树种高光谱数据识别分析 总被引:24,自引:2,他引:22
利用高分辨率光谱仪在实地测得的光谱数据来识别美国加州的6种主要针叶树种。树冠阴面和阳面的高光谱数据分别在1996年夏、秋测得。首先对原始光谱数据作简单处理,然后进行6种数据变换:对数变换、一阶微分变换、对数变换后一阶微分变换、归一化变换、归一化变换后一阶微分变换及归一化后对数变换。采用相邻窄波段逐步加宽的办法,测试不同波段宽度对树种识别精度的影响。所有的变换方法及波段宽度试验最后均由神经元网络算法产生的树种分类精度来评价。试验结果表明对数变换后一阶微分和归一化变换后一阶微分能够获得高于94%的平均精度;归一化变换和微分处理能够限制阴影的影响;20nm的波段宽度用于识别此6种针叶树种是较为理想的。我们发现太阳高度角变化对树种识别影响不大。 相似文献
13.
Changliang Yi 《Journal of the Indian Society of Remote Sensing》2013,41(4):957-967
Accurately estimating phytoplankton Chlorophyll-a(Chla) concentration from remotely sensed data is particularly challenging in turbid, productive waters. In this study, a weighted Chla concentration algorithm (WCA) are constructed to smooth the performance of three-bands semi-analytical algorithm(TSA) and four-bands semi-analytical algorithm(FSA). The performance of WCA, TSA and FSA algorithms are calibrated and validated by three independently datasets collected from Chesapeake Bay, USA, Yellow River Estuary, China, and Taihu Lake, China. Results of this study indicated that: (1) The accuracy and stability of TSA, FSA and WCA in Chesapeake Bay have a superior performance than it in Taihu Lake and Yellow River Estuary; (2) In Taihu Lake and Yellow River Estuary, the TSA and FSA algorithms are not stable for estimating Chla concentration, especially in Taihu Lake, the accuracy and stability of TSA, FSA and WCA algorithms are quite bad; (3) The WCA can greatly improve the accuracy and stability of TSA and FSA algorithms, but it is greatly depended on the performance of TSA and FSA algorithms; and (4) Although the TSA, FSA and WCA algorithms are semi-analytical algorithm, however, the optimal bands, accuracy and stability of these algorithms are very timely and located dependence. 相似文献
14.
通过研究,发现在大样本(N=335)的情况下MODIS 500 mB1-B3与太湖悬浮物浓度具有较好的线性关系(R2=0.659)。同时,运用实测光谱深入分析悬浮物的敏感波段,进一步揭示了采用B1-B3估测的理论依据,并建立了适用于各个季节的太湖悬浮物通用遥感估测模型。 相似文献
15.
Particulate organic carbon (POC) plays an important role in the carbon cycle in water due to its biological pump process. In the open ocean, algorithms can accurately estimate the surface POC concentration. However, no suitable POC-estimation algorithm based on MERIS bands is available for inland turbid eutrophic water. A total of 228 field samples were collected from Lake Taihu in different seasons between 2013 and 2015. At each site, the optical parameters and water quality were analyzed. Using in situ data, it was found that POC-estimation algorithms developed for the open ocean and coastal waters using remote sensing reflectance were not suitable for inland turbid eutrophic water. The organic suspended matter (OSM) concentration was found to be the best indicator of the POC concentration, and POC has an exponential relationship with the OSM concentration. Through an analysis of the POC concentration and optical parameters, it was found that the absorption peak of total suspended matter (TSM) at 665 nm was the optimum parameter to estimate POC. As a result, MERIS band 7, MERIS band 10 and MERIS band 12 were used to derive the absorption coefficient of TSM at 665 nm, and then, a semi-analytical algorithm was used to estimate the POC concentration for inland turbid eutrophic water. An accuracy assessment showed that the developed semi-analytical algorithm could be successfully applied with a MAPE of 31.82% and RMSE of 2.68 mg/L. The developed algorithm was successfully applied to a MERIS image, and two full-resolution MERIS images, acquired on August 13, 2010, and December 7, 2010, were used to map the POC spatial distribution in Lake Taihu in summer and winter. 相似文献
16.
17.
Based on in situ water sampling and field spectral measurements in Dianshan Lake, a semi-analytical three-band algorithm was used to estimate Chlorophylla (Chla) content in case II waters. The three bands selected to estimate Chla for high concentrations included 653, 691 and 748 nm. An equation, based on the difference in reciprocal reflectance between 653 and 691 nm, multiplied by reflectance at 748 nm as [Rrs−1(653) − Rrs−1 (691)] Rrs(748), explained 85.57% of variance in Chla concentration with a root mean square error (RMSE) of <6.56 mg/m3. In order to test the utility of this model with satellite data, HJ-1A Hyperspectral Imager (HSI) data were analyzed using comparable wavelengths selected from the in situ data [B67−1(656) − B80−1(716)] B87(753). This model accounted for 84.3% of Chla variation, estimating Chla concentrations with an RMSE of <4.23 mg/m3. The results illustrate that, based on the determined wavelengths, the spectrum-based model can achieve a high estimation accuracy and can be applied to hyperspectral satellite imagery especially for higher Chla concentration waters. 相似文献