基于光谱转换的土壤盐分反演与动态分析
发布时间：2019-01-19 14:55
【摘要】：为了对黄河三角洲地区进行大范围、长期的盐渍土监测,以研究区2000—2016年间4景Landsat-5 TM,EO-1 ALI,Landsat-8 OLI时间序列影像及Hyperion高光谱数据为基础开展土壤盐分定量反演分析.将Hyperion数据按照光谱响应函数分别重采样为TM,ALI,OLI模拟数据,采用数值回归方法计算TM,ALI与OLI对应波段间的光谱转换系数,从而将TM,ALI影像转换为OLI时序影像.分别采用偏最小二乘回归模型与多元线性回归模型建立土壤光谱与盐分参量间的预测关系,并将最优预测模型应用于OLI时序影像进行盐分反演制图,通过叠置方法进行盐渍土演化分析.结果表明,光谱转换方法提高了多传感器间数据一致性.偏最小二乘回归-电导率(PLSR-EC)模型的相关系数为0.700,采用2012年电导率实测值检验该模型反演结果,相关系数为0.690.研究区内高盐分土壤减少并向低盐分土壤转化.
[Abstract]:In order to monitor the salinized soil in the Yellow River Delta region for a long time, the study area will have four Landsat-5 TM,EO-1 ALI, from 2000 to 2016. Based on Landsat-8 OLI time series images and Hyperion hyperspectral data, quantitative inversion analysis of soil salinity was carried out. The Hyperion data are resampled to TM,ALI,OLI simulation data according to the spectral response function, and the spectral conversion coefficients between the corresponding bands of TM,ALI and OLI are calculated by numerical regression method, and the TM,ALI images are converted into OLI time series images. Partial least square regression model and multivariate linear regression model were used to establish the prediction relationship between soil spectral and salt parameters, and the optimal prediction model was applied to OLI time series image for salt inversion mapping. The evolution of saline soil was analyzed by superposition method. The results show that the spectral conversion method improves the data consistency between sensors. The correlation coefficient of partial least square regression conductivity (PLSR-EC) model is 0. 700. The inverse result of the model is tested by the 2012 conductivity data, and the correlation coefficient is 0. 690. In the study area, the high salinity soil decreased and transformed to the low salt soil.
【作者单位】： 东南大学交通学院;中国科学院南京地理与湖泊研究所;
【基金】：国家自然科学基金资助项目(41471352)
【分类号】：O212.1;S156.4
[Abstract]:In order to monitor the salinized soil in the Yellow River Delta region for a long time, the study area will have four Landsat-5 TM,EO-1 ALI, from 2000 to 2016. Based on Landsat-8 OLI time series images and Hyperion hyperspectral data, quantitative inversion analysis of soil salinity was carried out. The Hyperion data are resampled to TM,ALI,OLI simulation data according to the spectral response function, and the spectral conversion coefficients between the corresponding bands of TM,ALI and OLI are calculated by numerical regression method, and the TM,ALI images are converted into OLI time series images. Partial least square regression model and multivariate linear regression model were used to establish the prediction relationship between soil spectral and salt parameters, and the optimal prediction model was applied to OLI time series image for salt inversion mapping. The evolution of saline soil was analyzed by superposition method. The results show that the spectral conversion method improves the data consistency between sensors. The correlation coefficient of partial least square regression conductivity (PLSR-EC) model is 0. 700. The inverse result of the model is tested by the 2012 conductivity data, and the correlation coefficient is 0. 690. In the study area, the high salinity soil decreased and transformed to the low salt soil.
【作者单位】： 东南大学交通学院;中国科学院南京地理与湖泊研究所;
【基金】：国家自然科学基金资助项目(41471352)
【分类号】：O212.1;S156.4
【相似文献】
相关期刊论文 前10条
1 杨巧梅;杨宇辉;;石油市场理性情况下的动态分析[J];科技情报开发与经济;2006年10期
2 罗龙;;大国开放度变动的动态分析[J];数量经济技术经济研究;1990年01期
3 程耀东 ,牛加毅 ,齐津;一种结构动态分析中有限元网格细分方法[J];浙江大学学报(自然科学版);1992年S1期
4 孟宪宝,程庆水;关于折旧的动态分析[J];石家庄经济学院学报;1998年02期
5 杨振军,赵迎红,黄涛;对装备费用—效能动态分析的几点看法[J];价值工程;1999年03期
6 郑万明,张银生,马军海;在动态分析中不同分布的相位随机临界值对获取数据的影响(续)[J];天津商学院学报;1999年03期
7 潘建伟;;中国区域收入不均等动态分析与路径选择——基于广义熵视角的分析[J];经济与管理研究;2010年07期
8 罗持久,李桂青,张汝清;结构动态分析的函数变换算法[J];计算结构力学及其应用;1990年03期
9 王文亮;环向圈连带冠叶片的动态分析[J];复旦学报(自然科学版);1991年02期
10 邱m，
本文编号：2411476
本文链接：https://www.wllwen.com/kejilunwen/yysx/2411476.html
