A Spectral Index for Estimating Soil Salinity in the Yellow River Delta Region of China Using EO-1 Hyperion Data

Soil salinization is one of the most common land degradation processes. In this study, spectral measurements of saline soil samples collected from the Yellow River Delta region of China were conducted in laboratory and hyperspectral data were acquired from an EO-1 Hyperion sensor to quantitatively map soil salinity in the region. A soil salinity spectral index (SSI) was constructed from continuum-removed reflectance (CR-reflectance) at 2052 and 2203 nm, to analyze the spectral absorption features of the salt-affected soils. There existed a strong correlation (r = 0.91) between the SSI and soil salt content (SSC). Then, a model for estimation of SSC with SSI was established using univariate regression and validation of the model yielded a root mean square error (RMSE) of 0.986 and an R2 of 0.873. The model was applied to a Hyperion reflectance image on a pixel-by-pixel basis and the resulting quantitative salinity map was validated successfully with RMSE = 1.921 and R2 = 0.627. These suggested that the satellite hyperspectral data had the potential for predicting SSC in a large area.

Derivation of salt content in salinized soil from hyperspectral reflectance data: A case study at Minqin Oasis, Northwest China

Soil salinization is a serious ecological and environmental problem because it adversely affects sustainable development worldwide, especially in arid and semi-arid regions. It is crucial and urgent that advanced technologies are used to efficiently and accurately assess the status of salinization processes. Case studies to determine the relations between particular types of salinization and their spectral reflectances are essential because of the distinctive characteristics of the reflectance spectra of particular salts. During April 2015 we collected surface soil samples(0–10 cm depth) at 64 field sites in the downstream area of Minqin Oasis in Northwest China, an area that is undergoing serious salinization. We developed a linear model for determination of salt content in soil from hyperspectral data as follows. First, we undertook chemical analysis of the soil samples to determine their soluble salt contents. We then measured the reflectance spectra of the soil samples, which we post-processed using a continuum-removed reflectance algorithm to enhance the absorption features and better discriminate subtle differences in spectral features. We applied a normalized difference salinity index to the continuum-removed hyperspectral data to obtain all possible waveband pairs. Correlation of the indices obtained for all of the waveband pairs with the wavebands corresponding to measured soil salinities showed that two wavebands centred at wavelengths of 1358 and 2382 nm had the highest sensitivity to salinity. We then applied the linear regression modelling to the data from half of the soil samples to develop a soil salinity index for the relationships between wavebands and laboratory measured soluble salt content. We used the hyperspectral data from the remaining samples to validate the model. The salt content in soil from Minqin Oasis were well produced by the model. Our results indicate that wavelengths at 1358 and 2382 nm are the optimal wavebands for monitoring the concentrations of chlorine and sulphate compounds, the predominant salts at Minqin Oasis. Our modelling provides a reference for future case studies on the use of hyperspectral data for predictive quantitative estimation of salt content in soils in arid regions. Further research is warranted on the application of this method to remotely sensed hyperspectral data to investigate its potential use for large-scale mapping of the extent and severity of soil salinity.

Black Soils Classification by Ground Spectral Process and Analysis

1 Introduction Black soils are a soil type with good properties and high fertility,which is very suitable for plant growth(Liu et al.,2015).Black soil resources are widely distributed in North America,Eurasia,and South America,and cover about 916million ha around the world,35 million ha of this in northeast China(Liu et al.,2012).

Spectral Response of Different Eroded Soils in Subtropical China: A Case Study in Changting County, China

Hyper-spectral data is widely used to determine soil properties. However, few studies have explored the soil spectral characteristics as response to soil erosion. This study analysed the spectral response of different eroded soils in subtropical China and then identify the spectral characteristics and soil properties that better discriminate soils with different erosion degrees. Two methods were compared: direct identification by inherent spectral characteristics and indirect identification by predictions of critical soil properties. Results showed that the spectral curves for different degrees of erosion were similar in morphology, while overall reflectance and characteristics of specific absorption peaks were different. When the first method is applied, some differences among different eroded groups were found by integration of associated indicators.However, the index of such indicators showed apparent mixing and crossover among different groups, which reduced the accuracy of identification.For the second method, the correlation between critical soil properties, such as soil organic matter(SOM), iron and aluminium oxides and reflectance spectra, was analysed. The correlation coefficients for the moderate eroded group were primarily between-0.3 to-0.5, which were worse than the other twogroups. However, the maximum value of R2 was obtained as 0.86 and 0.94 for the non-apparent eroded and the severe group. Furthermore, these two groups also showed some differences in the spectral response of iron complex state(Fep), Aluminium amorphous state(Alo) and the modelling results for soil organic matter(SOM). The study proved that it is feasible to identify different degrees of soil erosion by hyper-spectral data, and that indirect identification by modelling critical soil properties and reflectance spectra is much better than direct identification. These results indicate that hyper-spectral data may represent a promising tool in monitoring and modelling soil erosion.

Hyper-spectral characteristics and classification of farmland soil in northeast of China

The physical and chemical heterogeneities of soils make the soil spectral different and complicated, and it is valuable to increase the accuracy of prediction models for soil organic matter（SOM） based on pre-classification. This experiment was conducted under a controllable environment, and different soil samples from northeast of China were measured using ASD2500 hyperspectral instrument. The results showed that there are different reflectances in different soil types. There are statistically significant correlation between SOM and reflectence at 0.05 and 0.01 levels in 550–850 nm, and all soil types get significant at 0.01 level in 650–750 nm. The results indicated that soil types of the northeast can be divided into three categories： The first category shows relatively flat and low reflectance in the entire band; the second shows that the spectral reflectance curve raises fastest in 460–610 nm band, the sharp increase in the slope, but uneven slope changes; the third category slowly uplifts in the visible band, and its slope in the visible band is obviously higher than the first category. Except for the classification by curve shapes of reflectance, principal component analysis is one more effective method to classify soil types. The first principal component includes 62.13–97.19% of spectral information and it mainly relates to the information in 560–600, 630–690 and 690–760 nm. The second mainly represents spectral information in 1 640–1 740, 2 050–2 120 and 2 200–2 300 nm. The samples with high OM are often in the left, and the others with low OM are in the right of the scatter plot（the first principal component is the horizontal axis and the second is the longitudinal axis）. Soil types in northeast of China can be classified effectively by those two principles; it is also a valuable reference to other soil in other areas.

Predicting Surface Roughness and Moisture of Bare Soils Using Multi- band Spectral Reflectance Under Field Conditions

Soil surface roughness, denoted by the root mean square height(RMSH), and soil moisture(SM) are critical factors that affect the accuracy of quantitative remote sensing research due to their combined influence on spectral reflectance(SR). In regards to this issue, three SM levels and four RMSH levels were artificially designed in this study; a total of 12 plots was used, each plot had a size of 3 m × 3 m. Eight spectral observations were conducted from 14 to 30 October 2017 to investigate the correlation between RMSH, SM, and SR. On this basis, 6 commonly used bands of optical satellite sensors were selected in this study, which are red(675 nm), green(555 nm), blue(485 nm), near infrared(845 nm), shortwave infrared 1(1600 nm), and shortwave infrared 2(2200 nm). A negative correlation was found between SR and RMSH, and between SR and SM. The bands with higher coefficient of determination R^2 values were selected for stepwise multiple nonlinear regression analysis. Four characterized bands(i.e., blue, green, near infrared, and shortwave infrared 2) were chosen as the independent variables to estimate SM with R^2 and root mean square error(RMSE) values equal to 0.62 and 2.6%, respectively. Similarly, the four bands(green, red, near infrared, and shortwave infrared 1) were used to estimate RMSH with R^2 and RMSE values equal to 0.48 and 0.69 cm, respectively. These results indicate that the method used is not only suitable for estimating SM but can also be extended to the prediction of RMSH. Finally, the evaluation approach presented in this paper highly restores the real situation of the natural farmland surface on the one hand, and obtains high precision values of SM and RMSH on the other. The method can be further applied to the prediction of farmland SM and RMSH based on satellite and unmanned aerial vehicle(UAV) optical imagery.

Quantitative retrieval of soil salt content based on measured spectral data

Choosing the Minqin Oasis, located downstream of the Shiyang River in Northwest China, as the study area, we used field-measured hyperspectral data and laboratory-measured soil salt content data to analyze the characteristics of saline soil spectral reflectance and its transformation in the area, and elucidated the relations between the soil spectral re-flectance, reflectance transformation, and soil salt content. In addition, we screened sensitive wavebands. Then, a multiple linear regression model was established to predict the soil salt content based on the measured spectral data, and the accuracy of the model was verified using field-measured salinity data. The results showed that the overall shapes of the spectral curves of soils with different degrees of salinity were consistent, and the reflectance in visible and near-infrared bands for salinized soil was higher than that for non-salinized soil. After differential transformation, the correlation coefficient between the spectral reflectance and soil salt content was obviously improved. The first-order differential transformation model based on the logarithm of the reciprocal of saline soil spectral reflectance produced the highest accuracy and stability in the bands at 462 and 636 nm; the determination coefficient was 0.603, and the root mean square error was 5.407. Thus, the proposed model provides a good reference for the quantitative extraction and monitoring of regional soil salinization.

Soil Salinity Detection in Semi-Arid Region Using Spectral Unmixing, Remote Sensing and Ground Truth Measurements

Soil salinity is one of the serious environmental problems ravaging the soils of arid and semi-arid region, thereby affecting crop productivity, livestock, increase level of poverty and land degradation. Hyperspectral remote sensing is one of the important techniques to monitor, analyze and estimate the extent and severity of soil salt at regional to local scale. In this study we develop a model for the detection of salt-affected soils in arid and semi-arid regions and in our case it’s Ghannouch, Gabes. We used fourteen spectral indices and six spectral bands extracted from the Hyperion data. Linear Spectral Unmixing technique (LSU) was used in this study to improve the correlation between electrical conductivity and spectral indices and then improve the prediction of soil salinity as well as the reliability of the model. To build the model a multiple linear regression analysis was applied using the best correlated indices. The standard error of the estimate is about 1.57 mS/cm. The results of this study show that hyperion data is accurate and suitable for differentiating between categories of salt affected soils. The generated model can be used for management strategies in the future.

基于GF-1数据的耕地土壤镉(Cd)含量遥感估算方法

本文采用多种光谱变换和回归分析方法探索了使用GF-1卫星影像监测耕地土壤镉(Cd)含量的可行性。首先针对获取的GF-1原始影像数据,在完成预处理及剔除植被信息后进行倒对数、平方根和反正弦平方根变换,生成4套光谱影像;然后分别用采样点5 m缓冲区内各套影像光谱统计值与Cd含量进行相关性分析和多种回归分析。选择模型决定系数最高(>95%)的反正弦平方根变换后的自适应重加权回归方法构建的线性回归模型作为遥感估算模型。遥感估算结果在稻田积水、边缘地带等出现了异常估算值;笔者分析原因后应用线性插值的方法得到最终估算结果。相关性分析和建模精度表明该方法是可行的,有望应用于实际土壤质量监测和土地管理中。

基于波谱响应特征的雄安新区农田土壤重金属含量反演

分析并监测雄安新区农田土壤污染状况,对于保障粮食安全、建设绿色雄安具有重要意义。该研究以雄安新区为研究区,基于多源遥感数据珠海一号(Zhuhai-1)OHS数据、哨兵二号(Sentinel-2)L2A级数据的波谱响应特征及实地测得的农田土壤重金属含量数据,在对土壤重金属含量单因子与多因子污染评价的基础上,筛选出3种超标的农田重金属元素铅(Pb)、铜(Cu)、锌(Zn),采用偏最小二乘回归方法(partial least squares regression,PLSR)建立农田土壤重金属含量反演模型。利用Zhuhai-1提取土壤样本点的原始光谱反射率以及4种变换后的光谱反射率,Sentinel-2提取7种对重金属胁迫敏感的植被指数,将其与3种土壤重金属含量作相关性分析,筛选出敏感波段与植被指数,即波谱响应特征,构建土壤重金属含量反演模型。结果表明,3种模型整体反演精度较为优良,Pb含量反演模型决定系数(determination coefficient,R^(2))为0.490,均方根误差(root mean squared error,RMSE)为4.66 mg/kg,平均绝对值误差(mean absolute error,MAE)为1.92 mg/kg;Cu含量反演模型R^(2)为0.491,RMSE为16.85 mg/kg,MAE为3.69 mg/kg;Zn含量反演模型R^(2)为0.664,RMSE为20.63 mg/kg,MAE为9.36 mg/kg。将该模型应用于雄安新区农田区域,得到大部分农田土壤中Pb含量均未超过风险筛选标准,在研究区西南部、西部部分区域Cu含量超过土壤污染风险筛选值,同时在研究区西部、西南部Zn污染较严重,雄安新区东南部部分农田有Zn零星分布,其他区域Cu和Zn含量未超过国家土壤污染风险管控值。因此,利用多源遥感数据波谱响应特征反演土壤重金属Pb、Cu和Zn含量,能够快速准确地实现对雄安新区土壤重金属污染情况的调查,同时为大面积土壤重金属含量监测提供解决方案。

农田土壤理化参数快速获取技术研究进展与展望

[目的/意义]土壤是农业基本的生产资料,其质量与农业高效生产和可持续发展密切相关。由于以往对农田的高强度利用以及土壤侵蚀等原因,导致部分农田出现土壤有机质明显下降、地力减弱和生态功能退化等现象。土壤理化参数作为揭示土壤空间特征、评估土壤肥力的关键指标,对农田可持续利用起着至关重要的作用。因此,土壤理化参数信息的快速获取极为必要。[进展]探讨了农田土壤理化参数获取技术的研究意义,总结了当前用于农田土壤理化参数信息获取的主要技术,包括以电化学分析和光谱分析为主的实验室快速检验技术,以电磁感应、探地雷达、多光谱、高光谱和热红外为主的近地快速感知技术,以直接反演法、间接反演法和结合分析法为主的卫星遥感技术,以及近年的新型快速获取技术,如生物传感、环境磁学、太赫兹光谱和伽马能谱等,梳理了各方法的优缺点及适用情况。[结论/展望]结合农田环境的作业需求,依据未来研究的侧重方向提出发展建议,包括开发便携化、智能化和经济型的近地土壤信息获取系统及设备,实现土壤信息的原位快速检测。优化低空土壤信息获取平台的性能,完善数据的解译方法;联合多因素构建卫星遥感反演模型,利用多种共享开放的云计算平台实现数据的深度挖掘。深入探索多源数据融合在土壤理化参数信息获取中的研究与应用,构建泛化能力强、可靠性高的土壤信息感知算法和模型等,从而实现土壤理化参数信息快速、精准和智能化获取。

基于sCARS的淮北平原土壤有机质含量高光谱建模

为确定淮北平原砂姜黑土土壤有机质(SOM)最佳反演模型,探寻最佳特征波长筛选方法,提高模型预测精度。利用原始光谱进行倒数对数(Log(1/R))、标准正态变量变换(SNV)、去包络线(CR)、一阶微分(FDR)处理,采用稳定竞争性自适应重加权算法(sCARS)筛选特征变量,对比分析竞争性自适应重加权算法(CARS)、相关系数法(|r|≥0.47)和显著性水平法(p≤0.01)所得结果,建立SOM含量的偏最小二乘(PLSR)模型,并对比精度差异。结果表明:(1)全波段范围内,SOM含量与原始光谱呈极显著负相关,与Log(1/R)光谱呈极显著正相关,与SNV光谱相关性明显增强。CR和FDR光谱与SOM含量呈不同程度的正负相关性。(2)对比全波段,CARS和sCARS算法能够有效去除光谱冗余信息,筛选得到特征波段数目仅占全波段的1%~5%。筛选后模型精度更高,相对分析误差(RPD)均大于1.8。(3)相比于CARS算法,sCARS算法具备更好的稳定性和精确性。筛选到的特征波段主要分布在800~850、1850~1900、2050~2500 nm区域。(4)Log(1/R)-sCARS模型精度最佳,建模集和预测集的决定系数(R2)分别提升了0.08和0.28,RPD值为3.05,对SOM含量预测极好。

土壤质地对灌区土壤盐分高光谱反演精度影响研究

针对土壤质地对高光谱反演土壤盐分精度的影响不明确问题,于2023年4月1-10日在内蒙古河套灌区沈乌灌域共采集了132个不同盐渍化程度的土壤样品,并同步采集了对应的光谱信息,研究了不同盐渍化程度下土壤光谱反射率的变化特征以及不同土壤质地光谱特征与土壤盐分的相关性,探讨了土壤样本适宜的数学变换方法,并筛选敏感波段,建立了基于全部样本以及不同土壤质地下的土壤盐分含量的高光谱反演模型。结果表明:随着土壤盐分含量的增加,高光谱反射率逐渐增大;随着土壤粒度的减小,不同土壤质地下土壤盐分与不同波段的反射率及其相关系数呈先增加后下降的变化趋势。通过对光谱数据进行数学变换后,发现以倒数对数微分、对数微分、平方根微分3种变换效果最佳。通过建立多元逐步线性回归(BPNN)、偏最小二乘回归(PLSR)、支持向量机回归(SVM)以及BP神经网络(BPNN)4种模型,对光谱变换下的盐分含量进行了估算,4种模型的估算精度由高到低表现为:BPNN>SVM>PLSR>MLSR。相较于全部样本的土壤盐分估算结果,考虑不同土壤质地的盐分估算精度均有所提升,其中砂粒质地估算精度R2由0.918提升到0.962,RPD由3.493提升到4.313;粉粒质地估算精度R2由0.866提升到0.902,RPD由2.613提升到3.310;黏粒质地估算精度R2由0.876提升到0.926,RPD由2.651提升到3.953,且在3种土壤质地背景下建立的模型均达到了出色模型的标准。说明在考虑土壤质地的前提下进行含盐量的高光谱反演,有利于提升反演精度。

黄棕壤性水稻土有机质含量高光谱反演研究

以黄棕壤性水稻土为研究对象,利用ASD FieldSpec®4地物波谱仪获取土壤高光谱反射率曲线,分析土壤有机质(SOM)含量的分布形态和高光谱特征,基于原始光谱(R)、一阶微分(FD)、二阶微分(SD)、倒数的对数(LR)、倒数一阶微分(FDR)和对数一阶微分(FDL)这6种光谱数据,分别建立了黄棕壤性水稻土SOM含量偏最小二乘回归模型(PLSR)、支持向量机模型(SVM)和BP神经网络模型(BPNN),并比较分析了这3种模型预测精度的差异。结果表明:(1)SOM含量与原始光谱反射率呈弱相关关系,经FD处理后,光谱曲线特征突出明显,光谱FD、SD、FDR和FDL变换能有效提升光谱反射率与SOM含量的相关性。(2)PLSR、SVM和BPNN模型对SOM含量低值(1.98%)的预测效果均较差;数理统计有助于模型精度的评价。(3)SVM模型的预测效果整体优于PLSR和BPNN模型;光谱FD变换的SVM模型对SOM含量的预测效果最好,其验证集的R2、RMSE和RPD分别为0.902、0.257和2.287,可为实现快速、准确地测定黄棕壤性水稻土SOM含量提供新的模型参考和技术思路。

微分光谱变换方法对土壤重金属含量反演精度的影响研究

随着我国工农业的日益发展,土壤中以镍(Ni)、铁(Fe)、铜(Cu)、铬(Cr)、铅(Pb)等为代表的重金属污染对人类生活产生了严重影响。高光谱遥感技术具有实时、无损、快速等优点,为高效准确地获取土壤重金属含量提供了科学手段。而在利用高光谱数据反演土壤重金属含量时,微分光谱变换方法的选择对遥感反演土壤重金属含量的精度有显著影响。为明确二者关系,基于研究区采集的60个土壤样品,测定其Ni、Fe、Cr、Cu、Pb等含量以及350~2500 nm波段范围的光谱反射率。在相关系数(CC)分析法的基础上通过改进离散粒子群算法(MDBPSO)优选遥感探测土壤重金属含量的特征波段。最终以优选出的特征波段作为自变量利用随机森林(RF)算法构建了Ni、Fe、Cr、Cu、Pb等重金属含量的估测模型。在对原始反射率数据进行高斯平滑的基础上,对比分析了一阶微分(R′)、对数倒数的一阶微分(1/lgR)′、倒数的一阶微分(1/R)′、指数的一阶微分(e^(R))′四种微分光谱变换方法对土壤重金属反演精度的影响。结果表明,在CC分析法的基础上,MDBPSO算法可以有效地降低光谱数据的冗余度,提高模型的运行效率。其中R′、(1/lgR)′、(1/R)′、(e^(R))′中对Ni、Fe、Cr、Cu、Pb敏感的特征波段个数分别至少减少了154、363、135、744和889个。(1/lgR)′、R′、R′、(1/R)′、R′光谱变换方法分别应用到Ni、Fe、Cr、Cu、Pb特征波段的组合运算中,得到的估测模型的精度优于其他微分变换方法;模型检验集的决定系数分别为0.913、0.906、0.872、0.912、0.876,均方根误差分别为0.743、0.095、2.588、1.541、1.453。本研究为利用遥感数据反演土壤重金属含量微分光谱变换方法的选择提供了科学的参考,为进一步实现土壤重金属含量的大面积高精度遥感监测提供新的思路。

土壤As含量光谱指数反演方法评估

为探究基于光谱指数反演土壤砷(As)含量的有效性和适用性,选取河北省保定市某农田为研究区,利用PSR-3500便携式地物光谱仪和电感耦合等离子发射光谱法测定42个土壤样品实验室及野外原位光谱信号和As含量;基于实验室光谱、野外原位光谱及野外原位-直接标准化校正(DS)光谱,计算叶绿素指数(CI)、差值指数(DI)、和值指数(SI)、比值指数(RI)和简化归一化指数(NDI和NPDI),采用相关性分析提取强相关光谱指数参与土壤As含量随机森林回归建模。在此基础上,通过与相同光谱环境全波段建模方法的精度对比评估光谱指数的有效性;对比不同光谱环境中各类光谱指数建模精度的稳定性以评估其适用性;结合典型土壤组分的吸收特征波段尝试解析光谱指数提升土壤As含量反演性能的内在机制。结果表明:(1)相较于全波段建模,光谱指数法在“实验室光谱、野外原位光谱、野外原位-DS光谱”三种光谱环境中均能有效提升土壤As含量反演建模精度,R_(p)^(2)和RPD分别从0.243和1.2提升至0.730和2.009、0.264和1.213提升至0.669和1.809、0.334和1.279提升至0.678和1.841;(2)三种光谱环境中,DI、RI、NDI在野外原位光谱、SI和NPDI在野外原位-DS光谱环境中的适用性较差,CI综合适用性最强,R_(p)^(2)>0.66,RPD>1.8;(3)指数特征波段表现出与铁氧化物、粘土矿物和有机物吸收特征的相关性,但部分指数特征波段缺乏可解释性,无法揭示指数计算通过组合波段放大有效信息和消除噪声的统一规律。该研究可为后续发展基于光谱指数的土壤重金属遥感反演应用甚至卫星有效载荷研制中的波段设计提供科学依据。

重庆市南川区土壤锰元素遥感反演

土壤中的锰元素对植物生长起着重要作用,土壤锰含量过高或者缺少都将对植物产生不良影响,因此快速监测土壤中的锰含量尤为重要。目前利用遥感技术监测土壤锰含量的相关研究主要集中在利用土壤光谱估算土壤锰含量,而对于植被常年覆盖的南方地区,难以从卫星影像中获取土壤光谱。因此,引入植被光谱,探索植被覆盖区域土壤锰元素的快速监测方法。首先从Landsat 8影像中提取11种植被光谱指标,并运用皮尔逊相关系数(Pearson correlation coefficient)结合方差膨胀因子(VIF)筛选出最佳植被光谱指标;在此基础上,利用偏最小二乘回归(PLSR)、多元逐步回归(MSR)和BP神经网络(BPNN)算法构建最佳植被光谱指标与土壤锰元素之间的光谱响应模型,分析比较三个模型的估算效果从而确定最佳反演模型;最后,基于最佳反演模型进行土壤锰含量空间制图。以重庆市南川区为例,研究结果表明:3个植被光谱指标(比值植被指数,归一化植被指数和可见光大气阻抗植被指数)被确定为土壤锰元素最佳的光谱响应指标;BPNN光谱响应模型(R^(2)为0.78,RMSE为334.24,RPD为2.13)为土壤锰含量最佳反演模型,其土壤锰含量的制图精度(R^(2)为0.69,RMSE为567.64,RPD为1.30)。表明通过植被光谱指标反演土壤锰含量可行,该研究为区域尺度的土壤锰含量监测开拓了新思路。

滨海湿地土壤质地高光谱估测模型对比分析

土壤质地影响着植被分布、水土保持能力、微生物活动等多种物理、化学、生物和水文特性和过程。准确地获取土壤质地对湿地生态修复和保护具有重要意义。基于天津市滨海湿地57个实测表层土壤质地和可见光-近红外高光谱数据,对土壤样品进行S-G平滑以及一阶微分(FD)、倒数(RT)、倒数一阶微分(RTFD)、平方根(SR)、平方根一阶微分(SRFD)、倒数之对数(LR)和倒数之对数一阶微分(LRFD)八种变换,分析不同土壤质地类别的光谱曲线特征及土壤粒径含量与八种变换之间相关性。通过竞争性自适应重加权算法(CARS)优选特征波段,结合偏最小二乘(PLSR)、随机森林(RFR)和支持向量机(SVR)三种回归算法,对比不同光谱变换后的土壤粒径含量建模效果。结果表明:(1)湿地土壤质地类别主要为粉壤土和粉土,粉土在400~2400nm波段光谱反射率最高,砂土在400~2000nm波段光谱反射率最低,FD、RTFD和SRFD变换后波段反射率与土壤粒径含量的相关性明显提高,最大相关系数绝对值均达到0.58以上,最高达到0.70。(2)CARS算法筛选八种光谱变换的特征波段数为全波段数的1.05%~6.15%,有效降低光谱数据的信息冗余。(3)对比三种粒径含量估测模型,SRFD和RTFD光谱变换的SVR模型精度最好,优于其他两种模型,黏粒(SRFD)测试集(R^(2)=0.72,RMSE=1.86%,nRMSE=11.33%)、粉粒(SRFD)测试集(R^(2)=0.72,RMSE=2.82%,nRMSE=7.30%)和砂粒(RTFD)测试集(R^(2)=0.71,RMSE=5.75%,nRMSE=5.91%)。研究结果可为高光谱数据准确监测滨海湿地土壤质地提供依据与技术支撑。

基于光谱指数建模的沙井子灌区土壤盐分反演

为了快速准确地获取干旱地区表层土壤盐分信息,以沙井子灌区为研究区,利用地面采集的0~10 cm和10~20 cm深度的土壤盐分数据,以及同步获取的Landsat 9 OLI遥感影像上相应点位的波段反射率值,组合两波段和三波段光谱指数,建立低植被度覆盖下盐渍化监测SDI1、SDI2、SDI3、SDI4模型,并检验4类模型对不同土层深度土壤盐分的反演精度。结果表明:(1)当土层深度为0~10 cm时,4类盐渍化监测模型对土壤盐渍化等级分类精度分别为73.56%、66.35%、43.75%和74.52%;而当土层深度为10~20 cm时,相应的分类精度分别为61.06%、62.50%、66.35%和64.42%,说明灌区内土层最佳反演深度为0~10 cm。(2)三波段光谱指数构建的SDI4模型优于双波段光谱指数构建的其余3种模型,能够有效反演沙井子灌区土壤盐渍化程度。研究结果可为灌区土壤盐渍化治理和防治提供有效的技术参考。

融合作物类型的土壤盐分遥感反演方法研究

在沿海平原地区,土壤盐度是制约作物生长的非生物胁迫之一,也是作物种植的重要依据,作物类型能够间接反映土壤盐渍化程度,因此本文提出了一种融合作物类型信息的土壤盐分反演方法。以黄河三角洲典型滨海盐渍土地区为例,基于Sentinel-2 MSI影像,首先采用随机森林分类提取作物类型信息,并基于OneHot方式将作物类型信息编码;然后融合作物类型信息,结合环境协变量数据、地面实测盐分数据,采用自适应增强决策树模型(AB-DT)进行盐分反演;最后与其他机器学习方法,如支持向量机、随机森林、K最邻近和决策树进行盐分反演精度的对比。结果表明:①加入作物类型信息能够提高土壤盐分反演模型精度,所有模型中,融合作物类型变量的AB-DT反演模型精度最高,建模集R 2为0.86,测试集R 2为0.61;②加入作物类型信息能够修正误判的盐渍土级别,并使土壤盐分反演结果的地块边缘更加清晰。综上所述,加入作物类型信息,能够提高土壤盐分反演的准确性,为农田管理和农业决策提供更可靠的依据。