自然资源遥感, 2024, 36(4): 9-22 doi: 10.6046/zrzyyg.2023245

综述

土壤盐渍化光学遥感监测方法研究进展

骆振海,, 张超,, 冯绍元, 唐敏, 刘锐, 孔纪迎

扬州大学水利科学与工程学院,扬州 225009

Advances in research on methods for optical remote sensing monitoring of soil salinization

LUO Zhenhai,, ZHANG Chao,, FENG Shaoyuan, TANG Min, LIU Rui, KONG Jiying

College of Hydraulic Science and Engineering, Yangzhou University, Yangzhou 225009, China

通讯作者: 张超(1990-),男,博士,副教授,主要从事多尺度农田水肥盐遥感监测与模拟研究。Email:zhangc1700@yzu.edu.cn。

责任编辑: 张仙

收稿日期: 2023-08-7   修回日期: 2024-03-15  

基金资助: 国家自然科学基金项目“多源遥感与作物-水盐耦合模型同化的盐渍化农田水盐反演方法研究”(52379049)
江苏省研究生科研与实践创新计划项目(SJCX23_1945)

Received: 2023-08-7   Revised: 2024-03-15  

作者简介 About authors

骆振海(2000-),男,硕士研究生,主要研究方向为土壤盐渍化遥感定量监测。Email: MZ120221105@stu.yzu.edu.cn。

摘要

土壤盐渍化是导致土壤肥力下降、生产力衰退、植被覆盖减少以及作物产量降低的重要原因之一。光学遥感监测技术具有宏观、及时、动态和低成本等优点,对于实现土壤盐渍化的动态监测具有重要意义。然而,目前少有对多尺度遥感数据、多类型遥感特征参量以及反演模型进行系统性梳理和总结的综述研究。为此,该文首先对光学遥感数据源进行了梳理,并总结了目前盐渍土监测相关研究中使用的遥感数据来源与尺度平台,将多源遥感数据分为卫星、航空和地面3种不同平台; 其次,整理了目前主流的建模特征参量以及统计回归和机器学习2类经典的反演方法,并分析了各方法的研究现状; 最后,对遥感数据源间的融合进行了探讨,比较了各建模方法的优缺点,并结合当前研究热点,展望了将数据同化与深度学习应用于土壤盐渍化监测研究的前景。

关键词： 土壤盐渍化; 土壤盐分; 光学遥感; 反演模型; 特征参量

Abstract

Soil salinization is identified as a major cause of decreased soil fertility, productivity, vegetation coverage, and crop yield. Optical remote sensing monitoring enjoys advantages such as macro-scale, timeliness, dynamics, and low costs, rendering this technology significant for the dynamic monitoring of soil salinization. However, there is a lack of reviews of the systematic organization of multi-scale remote sensing data, multi-type remote sensing feature parameters, and inversion models. This study first organized the optical remote sensing data sources and summarized the remote sensing data sources and scale platforms utilized in current studies on saline soil monitoring. Accordingly, this study categorized multi-source remote sensing data into three different platforms: satellite, aerial, and ground. Second, this study organized the mainstream characteristic parameters for modeling and two typical inversion methods, i.e., statistical regression and machine learning, and analyzed the current status of research on both methods. Finally, this study explored the fusion of remote sensing data sources and compared the pros and cons of various modeling methods. Furthermore, in combination with current hot research topics, this study discussed the prospects for the application of data assimilation and deep learning to soil salinization monitoring.

Keywords： soil salinization; soil salinity; optical remote sensing; inversion model; characteristic parameter

PDF (1421KB) 元数据 多维度评价 相关文章 导出 EndNote| Ris| Bibtex  收藏本文

本文引用格式

骆振海, 张超, 冯绍元, 唐敏, 刘锐, 孔纪迎. 土壤盐渍化光学遥感监测方法研究进展[J]. 自然资源遥感, 2024, 36(4): 9-22 doi:10.6046/zrzyyg.2023245

LUO Zhenhai, ZHANG Chao, FENG Shaoyuan, TANG Min, LIU Rui, KONG Jiying. Advances in research on methods for optical remote sensing monitoring of soil salinization[J]. Remote Sensing for Land & Resources, 2024, 36(4): 9-22 doi:10.6046/zrzyyg.2023245

0 引言

土壤盐渍化问题在全球范围内普遍存在。跟据联合国教科文组织(United Nations Educational, Scientific and Cultural Organization,UNESCO)和世界粮农组织(Food and Agriculture Organization of the United Nations,FAO)的数据统计,全球盐渍土面积约为9.54×109 hm2[1],其中中国盐渍土总面积约为3.6×107 hm2[2]。我国耕地中盐渍化面积达到920.9×104 hm2,占全国耕地面积的6.62%。土壤盐渍化是导致土地肥力下降、生产力衰退、植被覆盖度减少以及作物产量降低的重要原因之一[3]。盐渍土也是我国中低产土壤中的主要类型之一,而土壤盐渍化又是我国干旱、半干旱和滨海地区最常见的土地退化问题,严重影响着当地的作物产量、经济发展和环境可持续性。

盐渍土的形成是一个复杂的过程,因此对其进行探测和动态监测是一项相对困难的工作。传统监测方法通常涉及人工野外取样和分析,这种方法不仅耗时且工作量大,只适用于小范围测量,也可能存在样本的代表性问题,因此难以实现大面积动态监测[4]。遥感技术通过获取地表影像数据,能够快速获取大范围内不同时间下的地物信息。这种技术具有宏观、综合、及时、动态和低成本等特点,在大面积土壤盐度反演中被广泛应用[5]。对含盐量不同的土壤,可见光和近红外波段的光谱响应存在差异,高盐分土壤的光谱响应较强[6]。利用遥感技术了解土壤盐分可能发生变化的时间、方式和地点,实现土壤盐渍化的动态监测,对于土壤和水资源高效利用具有重要意义。

在过去的几十年中,研究学者们在盐渍化土壤的光谱特征、遥感数据源以及土壤盐分反演方法等方面取得了显著进展,经历了从遥感影像目视解译到基于影像光谱特征的数字遥感影像计算机自动分类处理的演变,并且利用了诸如可见光、多光谱、高光谱和热红外等多源遥感数据来进行相关研究[7]。目前,利用遥感数据特征参量构建反演模型已成为土壤盐渍化研究的热点[8-9] 。其中,定量评价土壤盐化程度的核心任务是建立土壤含盐量、土壤电导率、盐渍化面积和遥感特征参量之间的关系,建立可靠的遥感数据经验和物理反演模型,以实现对研究区内土壤盐渍化时空动态的监测[10-11]。

本研究在Web of Science 核心合集数据库中以 “soil salinization”“salinity”“salt”和“remote sensing” 作为主题词检索了2000—2023年内共2 163篇相关研究文献。通过可视化分析,筛选出现频次超过30次的关键词,其中,出现频率较高的关键词主要有remote sensing(遥感)、salinity(盐分)、soil salinity(土壤盐分)、soil moisture(土壤水分)、band(波段)、UAV(无人机)、satellite(卫星)、model(模型)等。这表明土壤盐分监测依赖于一定的模型和技术方法。利用光谱数据结合反演模型监测土壤盐渍化状况已成为常用的研究手段。在以往的研究中,研究者们主要集中于探索遥感技术在土壤盐渍化反演中的应用,特别是在单个建模因子的选择、模型建立及其精度验证等方面,如Sahbeni等[12]深入分析和讨论了利用遥感技术进行土壤盐分测绘和监测的相关工作; Ma等[13]对中国盐渍土管理的研究历程、土壤盐渍化监测的研究进展以及主要的建模方法进行了梳理。但目前对多尺度遥感数据、多类型遥感特征参量以及反演模型进行系统性梳理和总结的综述研究较少。因此本文的主要内容将围绕以下几个方面进行梳理和总结: ①总结光学遥感反演盐渍土信息的主要数据源; ②梳理常用的土壤盐渍化反演模型; ③分析目前土壤盐渍化遥感监测的局限性,并对未来的建模和反演方法进行展望。

1 光学遥感数据源

遥感是一种重要的对地观测技术,它主要利用卫星或飞机上的传感器,在无需物理接触的情况下获取遥感影像和地物信息[14]。光学遥感成像是遥感技术的一个重要分支,其传感器工作波段限于可见光至红外波段范围(350~2 500 nm),包括可见光、近红外和短波红外,被广泛应用于植被监测、水质监测、精准农业等领域[15]。光学遥感技术提供了高分辨率的多种类型影像数据,覆盖范围广且具有时间上的连续性,能够获取丰富的遥感信息。基于光谱信息表征土壤理化性质变化的特点,已验证了光学遥感数据在监测土壤盐渍化方面的有效性[16]。光学遥感主要获取地物的辐射亮度和反射率数据[17],依赖于卫星、航空和地面遥感平台等几种主要方式。

1.1 卫星遥感监测

自20世纪70年代开始,学者们开始使用卫星遥感影像数据进行土壤盐渍化监测的相关研究,主要基于土壤反射率随含盐量变化规律进行定量反演[18]。到了90年代后期,随着遥感数据源的逐渐丰富,研究方法变得更加复杂[19]。此后,随着航天技术的进步,卫星传感器的空间分辨率不断提高,从Landsat,MODIS,Sentinel系列等中分辨率卫星扩展到包括QuickBird,GeoEye,WorldView等高分辨率卫星(表1),这实现了更精细的地表特征观测。另一方面,传感器的光谱分辨率也显著提升,高光谱遥感数据为盐渍土的精准定量监测提供了更多的光谱波段信息,发挥着重要作用。数据的分析与处理也经历了丰富的发展历程,从早期的人工可视化解读逐渐演变为自动化特征提取和分类算法,并进一步拓展到基于统计学、机器学习和深度学习等算法的遥感影像分类、识别和反演工作[20]。在土壤盐渍化监测中,常用的光学数据源诸如高分辨率影像、高光谱影像和多光谱影像等能够提供详细的土壤信息,为研究人员提供了广泛的选择和应用空间。高分辨率遥感影像可用于准确描述地物纹理和形状,而高光谱影像则因其具有较高的光谱波段信息,在遥感建模和特征识别方面得到广泛的应用[21]。另外,多光谱影像具备宽阔的视场、实时信息采集和周期性覆盖能力,使得监测土壤盐分变化变得更加容易[22]。

表1   常用卫星遥感主要参数

Tab.1  Main parameters and features of commonly used satellite remote sensing

卫星名称波段范围/μm空间分辨率/m时间分辨率
Landsat70.45~12.5全色: 15
多光谱: 30		重访周期16 d
Landsat80.43~12.51多光谱: 15/30
热红外: 100		重访周期16 d
Sentinel-2B0.4~2.410/20/60单星重访周期10 d
高分一号
(GF-1)		全色:0.45~
0.90
多光谱:0.45~0.89		PMS全色: 2
PMS多光谱: 8
WFV多光谱: 16		重访周期4 d
高分六号
(GF-6)		高分相机:
0.45~0.90
宽幅相机:
0.45~0.89		PMS全色: 2
PMS多光谱: 8
WFV多光谱:16		与GF-1组网运行后,将使遥感数据获取的时间分辨率从4 d缩小到2 d
Terra,Aqua0.4~14.4MODIS: 250~1 000
ASTER: 15~90		重访周期1~2 d
资源三号
(ZY-3)		全色: 0.50~0.80
多光谱: 0.45~0.89		正视全色: 2.1
前、后视全色: 3.5
正视多光谱: 6		重访周期5 d
HJ-1B0.43~12.5CCD相机: 30
红外多光谱相机: 150/300		重访周期4 d
Pleiades-1A0.43~0.95全色: 0.5
多光谱: 2.0		重访周期1 d
HJ-1A0.43~0.95CCD相机: 30
高光谱成像仪: 100		重访周期4 d
RapidEye0.45~0.9全色0.61~0.72
多光谱2.44~2.88		重访周期1 d
QuickBird0.45~0.9全色: 0.61
多光谱: 2.44		重访周期1~
3.5 d
IKONOS全色0.45~0.9
多光谱:0.45~0.88		全色: 1
多光谱: 4		重访周期1~
3 d
WorldView-20.4~1.04全色: 0.5
多光谱: 1.8		重访周期1.1 d
WorldView-30.4~1.04全色: 0.31
多光谱: 1.24		重访周期1 d
SPOT-60.45~0.89全色: 1.5
多光谱: 6		重访周期2~3 d

新窗口打开| 下载CSV


Gorji等[23]分析了基于Landsat8 OLI和Sentinel-2A数据构建的光谱指数和实测土壤电导率的关系,对乌鲁米耶湖西部的土壤盐分进行了估算; Allbed等[24]通过构建土壤盐度指数和IKONOS影像波段反射率之间的关系,对土壤盐分的空间变化进行了研究; 陈实等[25]基于MODIS数据监测分析了北疆农区土壤盐渍化状况及其空间动态变化; 陈俊英等[26]在无人机和GF-1卫星遥感数据的基础上,进行了河套灌区沙壕灌域地区土壤盐渍化监测研究; Farahmand等[27]评估了基于光学Sentinel-2A影像数据的各种非线性回归模型估算土壤盐分含量的能力。自我国的GF-1卫星发射以来,空间对地观测能力得到了显著提升,并在各领域得到了广泛的应用。在基于卫星遥感影像的土壤盐渍化监测研究中,可见光到红外光谱中的特征参量能够更准确地估算土壤盐分含量,从而提高了土壤盐渍化监测的准确性[3]。

1.2 航空遥感监测

近年来,航空遥感平台如无人机等技术迅速发展,并逐渐融入民用领域,成为农业研究和应用中热点工具。无人机具有便携、高灵活性和飞行时间自由等优点,能够在较低高度和不同类型的区域飞行,并捕获具有高时空分辨率的影像[28]。在农业遥感监测中,固定翼、多旋翼和无人驾驶直升机是常用的无人机制式(表2)。根据不同的监测任务,可以在无人机平台上安装多种传感器,如数码相机、多光谱相机、高光谱相机、热成像相机和激光雷达等光学设备(表3)。相对于卫星传感器,机载传感器具有提供更高分辨率影像、受云层和大气气溶胶等噪声干扰较小等优点,使其更适用于农业生产实践。特别是在当前农业生产趋于高度区域化、一体化、精准化以及智能化的趋势下,低空遥感平台展现出更广阔的发展前景[29]。Ivushkin等[30]研究指出,在土壤盐渍化监测中,无人机搭载高光谱相机、多光谱相机、热红外相机和激光雷达等设备显示出巨大的潜力。

表2   农业遥感监测的无人机特性

Tab.2  UAV characteristics for agricultural remote sensing monitoring

类别常见机型优点缺点
固定翼DJI Phantom 4,Parrot Bebop 2,DJI Mavic Air 2续航时间长,负载大,飞行速度快,可操作范围大起飞需要助跑,着陆需要滑翔,不能悬停
多旋翼大疆精灵,大疆M600,大疆S1000可水平和垂直飞行、起降,可悬停在特定位置,自主导航,结构简单续航时间短,负载小,对恶劣环境的抵抗力较差
直升机K-MAX,JT8D,VSR700垂直起降,悬停在给定位置,飞行稳定性高机翼结构复杂,维护成本高

新窗口打开| 下载CSV


表3   不同类型的机载传感器及其特性

Tab.3  Different types of airborne sensors and characteristics

类型传感器光谱波段①波长范围/μm特性参考文献
数码相机Sony DSC-QX 100R,G,B—分辨率: 2 020万像素
感光度: 160~12 800
质量: 179 g		[31]
Nikon D90R,G,B—分辨率: 1 230万像素
感光度: 200~3 200
质量: 620 g		[32]
多光谱成像仪XCam Multi-SpectrumG,R,R-edge,NIR0.55~0.79高度自动化,凝视成像
质量: 470 g		[33]
Micro-MCAG,R,R-edge,NIR0.45~1分辨率: 130万像素
质量: 497~1 000 g
镜头焦距: 9.6 nm		[34]
Parrot SequoiaG,R,R-edge,NIR0.55~0.79分辨率: 120万像素
质量: 72 g
帧频: 1帧/s		[35]
高光谱成像仪Nano-Hyperspec340个波段0.4~1空间像素: 1 020
光圈: F/2.5
质量: 1 000 g		[36]
Rikola最大380个波段0.5~0.9波段反射率: 30帧/s
质量: 720 g
地面采样距离: 100 m时为6.5 cm		[30,36]
Gaia Sky-mini128/256/520/
1 040		0.4~1质量: 1 000 g
像素间距: 6.45 μm
横向视场: 168 m		[37]
热红外成像仪Tau© 2—8~14红外分辨率: 640×512
像素尺寸: 17 μm
温度范围: -20~100 ℃		[38]
Fluke TiX620—7.5~14图像分辨率: 640×480
质量1.5 kg
温度范围: -40~600 ℃		[39]
激光雷达VUX©-1UAV——质量: 3 600 g,波长: 1 550 nm、光
斑直径: 25 nm		[30]

① B为蓝色波段; R为红色波段; G为绿色波段; NIR为近红外波段; R-edge为红边波段。

新窗口打开| 下载CSV


无人机具备携带不同传感器能力,可以在可见光、近红外和中红外波段上进行土壤盐分含量的反演,从而提高估算精度和可靠性[40-41]。例如,Xie等[33]使用配备XCam多光谱相机的多旋翼无人机采集了中国浙江省杭州湾南岸的多光谱影像,并结合卫星数据采用支持向量机(support vector machines, SVM)分类方法对研究区裸地和植被进行分类,根据分类结果构建模型,充分发挥了无人机在提供准确光谱信息和大尺度卫星光谱数据互补方面的优势; Hu等[42]则利用搭载在无人机上的高光谱相机,定量表征了田间尺度的土壤盐分含量和状态。近年来,热红外传感器由于技术的改进和成本的降低越来越受到关注。目前,有2种主要类型的热红外成像设备,一种是能够捕捉点或线的扫描设备,另一种是具有二维红外焦平面阵列的扫描设备。热红外成像具有非入侵、非接触和无损的优势,能够快速确定植被冠层和土壤表面的温度分布。Tian等[39]利用热红外遥感数据研究了土壤盐分胁迫对植被蒸散和生长的影响,研究发现土壤盐分越高,植被冠层覆盖率越低。另外,机载激光雷达是一种新型的主动遥感技术,其最显著的优势是能够直接获取高精度三维点云数据。乔纪纲等[43]结合激光雷达和多光谱数据,对莺歌海滨岸带进行地表信息提取,深入研究了该区域土地退化、植被分布以及湿地特征。尽管无人机在农业监测中得到广泛应用,但仍然面临一些挑战,包括续航能力、负载能力以及监测的有效性等方面的问题。此外,由于无人机搭载不同类型的相机,获取的影像通常呈现多元化和不规则的形态,需进一步借助专业软件进行特定的数据预处理和分析。未来的发展需要尽快实现轨迹飞行、数据采集和诊断填图的智能化和自动化,以更好地应对当前的挑战,并进一步促进低空遥感在农业监测中的应用[44]。

1.3 地面遥感监测

随着近端传感技术的不断发展,各种地面遥感平台逐渐增多。近端传感技术,如时域反射仪、大地电导率仪、地物光谱仪、荧光光谱仪(表4)等,以及光学、热红外、微波影像等技术也日益成熟,为多要素、多尺度一体化盐渍化土壤水盐信息的获取提供了新的手段[45]。例如,通过采集不同深度的土壤盐分数据,可利用电磁感应仪以水平和垂直2种模式来测量不同深度土壤电导率,进而根据相关关系计算出土壤盐分含量[46]。Deng等[47]以渭干河—库车河三角洲绿洲为研究区,运用电磁感应技术对典型地块的土壤电导率进行测量,以评估该地区土壤剖面中盐分的空间分布; Kahaer等[48]对野外采集的土壤样品进行了室内高光谱测量和电导率测定,通过参数筛选,建立了土壤电导率的高光谱估算模型,成功实现了土壤盐分的有效监测。电磁感应仪在评估多时空尺度下土壤盐渍化的性质、起源和演变方面得到广泛应用。Wu等[22]利用EM38-MK2电导率仪测量了伊拉克中部地区的土壤电导率,并结合卫星遥感影像建立了遥感盐分反演模型,盐分含量预测准确率达到82.6%; Wu等[11]使用SR-3500光谱仪对平罗地区的土壤样品进行反射率测量,并基于Landsat影像、理化性质以及敏感波段,建立了一种新的模型来模拟和预测研究区土壤盐渍化状况,结果表明,绿光、蓝光和近红外光与土壤盐分含量存在显著相关性。电磁能、土壤表面和盐分的物理化学特征之间的相互作用使得在干旱地区遥感监测土壤盐碱化成为可能。相对于非盐渍土,受盐渍影响的土壤在光谱的可见光和近红外区域显示出更高的光谱反射率。Xu等[49]使用ASD光谱仪测定了中国内蒙古自治区河套灌区表层的不同土壤水分和盐浓度下的反射率,利用540 nm,1 740 nm,2 010 nm和2 350 nm波段数据得到的盐分估算模型决定系数R2达到0.951。

表4   盐渍土遥感监测的近地仪器

Tab.4  Near-earth instruments for remote sensing monitoring of saline soils

类 型传感器特性优点缺点
土壤
含水量
测量		TDR利用电磁波的传播时间来测量土壤含水量非侵入性,实时测量,高精度成本较高,结果需要根据特定的土壤类型进行校准
传导式
电导率
仪		Thermo
Scientific
Orion		适用于浅层土壤电导率测量快速测量,无需采样,使用范围广易受环境影响,空间分辨率有限,数据解释复杂
感应式
电导率
仪		EM38-
MK2		适用于较深层土壤和地下水电导率测量
手持式
便携式
光谱仪		SR-3500
地面
光谱仪		操作界面简
单、可快速
获取光谱
数据		高效、快速,可提供多个参数进行分析覆盖的波长范围有限,需要校准,需要大量的参考样本
高分辨率
光谱仪		ASD
光谱仪		可提供更精细的光谱信息

新窗口打开| 下载CSV


2 基于光学遥感数据的土壤盐渍化反演方法

土壤盐渍化的反演方法主要包括直接方法和间接方法2种。直接方法是指从遥感影像中直接对盐渍化土壤进行解译。由于盐分积累,土壤表层形成盐壳或盐皮,在地表呈现浅色或灰白色。遥感影像中的部分可见光和近红外波段数据对土壤盐分变化较为敏感,通过将其与土壤盐分含量关联构建土壤盐度指数,可以有效监测裸地土壤的盐渍化状况[50]。间接方法主要通过植被特征、土壤温度、土壤水分和土壤理化性质等间接反演土壤盐渍化特征。受盐胁迫的植物会在形态上发生变化(如叶绿素、干物质和叶面积指数等),在光谱上呈现出不同的响应特征,可以提取此类特征作为特征参量建立其与土壤盐分含量之间的关系。

2.1 遥感特征参量

利用遥感数据定量评估土壤盐分的核心方法是明确相关盐分指标与遥感数据之间的关系。遥感影像通过预处理后具有丰富的地物光谱信息,大量研究表明,植被指数、盐分指数、水分指数以及温度指数等常用于土壤盐分的反演研究,本文总结了现有研究中有代表性的遥感特征参量,如表5所示。

表5   常用光谱指数计算公式

Tab.5  Commonly used spectral indices calculation formulas

特征参量变量名称公式①参考文献
盐分
指数		盐分指数(salinity index, SI)B×R　[9]
盐分指数(SI1)G×R　[17,51
52]
盐分指数(SI2)G2+R2+NIR2[17,51
52]
盐分指数(SI3)G2+R2　[17,51
52]
盐分指数(SI6)B×RG[52]
盐分指数(SI7)NIR×RG[53]
亮度指数(brightness index, BI)R2+NIR2　[51,54]
归一化盐分指数(normalized salinity index, NDSI)R-NIRR+NIR[55]
盐分指数(salinity index,SI-T)RNIR×100[56]
强度指数(intensity index, INT1)G+R2[51,57]
植被
指数		土壤调节植被指数(soil-adjusted vegetation index, SAVI)(NIR-R)×1.5NIR+R+0.5[17,58]
归一化植被指数(normalized vegetation index, NDVI)NIR-RNIR+R[16,17
59]
重整化差异植被指数(renormalize differential vegetation index, RDVI)NIR-RNIR+R　[60]
绿色归一化差分植被指数(green normalized differential vegetation index, GNDVI)NIR-GNIR+G[60]
三角植被指数(triangular vegetation index, TVI)NIR-RNIR+R+0.5　[60]
差值植被指数(differential vegetation index, DVI)NIR-R[16,17]
归一化差值绿色指数(normalized difference green index, NDGI)G-RG+R[61]
增强化归一植被指数(enhanced normalized differential vegetation index, ENDVI)NIR+SWIR2-RNIR+SWIR2+R[11]
水分
指数		水分指数(water index, WI)NIRSWIR[10]
归一化水分指数(normalized differential water index, NDWI)NIR-SWIRNIR+SWIR[15,62]
温度温度植被干旱指数(temperature vegetation drought index, TVDI)TS-TSminTSmax-TSmin[63]
植被温度条件指数(vegetation temperature condition index, VTCI)LSTNDVImax-LSTNDVILSTNDVImax-LSTNDVImin[64]

① B,R,G,NIR,SWIR和SWIR2分别为蓝光、红光、绿光、近红外和短波红外波段反射率; L为常数(通常取值为1); TS为地表温度; TSmax和TSmin分别为TS的最大值和最小值; LSTNDVI为NDVI像素值的地表温度; LSTNDVImax和LSTNDVImin分别为LSTNDVI的最大值和最小值。

新窗口打开| 下载CSV


1)光谱特征参数。光谱特征参数是指在一定波长范围内的光谱数据中,通过对光谱曲线进行分析和计算获得的描述性指标。这些参数可以用来表征不同物质或对象的光谱特性,包括反射率、吸收率、辐射亮度等。土壤盐分会对土壤表面的反射率产生影响,盐渍化土壤在光谱的可见光和近红外区域呈现出比非盐渍化土壤更高的光谱响应[65]。Sidike等[66]采用偏最小二乘回归方法筛选了土壤盐分敏感波段,研究结果表明近红外波段对土壤盐分的光谱响应最为显著; Fan等[67]通过统计分析方法证实土壤盐分与NIR和SWIR波段的相关性更高; Zhang等[68]通过原始反射率相关图、一阶导数光谱相关图和偏最小二乘回归分析,发现395~410 nm,483~507 nm,632~697 nm,731~762 nm,812~868 nm,884~909 nm和918~930 nm波长是土壤盐分的敏感光谱波段。此外,有学者提出采用NDSI结合遥感影像可实现盐渍土区域监测[55]。Abdullah等[69]通过主成分分析(principal component analysis, PCA)构建了基于5个盐度指数和11个环境变量的土壤盐分估算模型,并利用土壤调查和土地覆盖图的地面土壤盐分预测值评估模型的预测精度。同时,Khan等[70]研究表明,BI,NDSI和SI等指数在估算巴基斯坦旁遮普省中部土壤盐分方面表现出较好的有效性。

2)植被指数和水分指数。植被指数是从植被的光谱反射特性中提取出来的一类指标。在高盐分胁迫下,地表的植被形态会因根系吸水不足和离子毒害作用而发生变化,这种变化可以通过光学遥感技术进行诊断和分析。分散的植被或土壤表面的盐生植物可以作为土壤盐分状态的标志,通过利用植被的反射率可以间接监测土壤盐渍化情况。NDVI和SAVI等几种植被指数常被用于评估和绘制土壤盐分的间接指标[16-17]。Allbed等[71]通过分析NDVI值和盐度指数的特性,实现了对多光谱图像中土壤盐度和植被覆盖变化的监测; Wu等[11]结合了多光谱冠层响应盐分指数(canopy response salinity index,CRSI)、垂直干旱指数(perpendicular dryness index,PDI)以及ENDVI,建立了大尺度、高精度的土壤盐分遥感定量反演模型; 张思源等[72]选用引入短波红外波段的ENDVI和半干旱区反演效果最优的SI3构建ENDVI-SI3特征空间,建立改进型盐渍化监测指数模型,为半干旱区盐渍化反演特征空间中特征参量的选取提供了新思路。水分指数是一类用于表征土地或植被水分含量和干旱程度的指标。遥感技术获取的光谱数据能够反映出土地或植被在不同波段上的吸收与反射特性,进而推测水盐分布状况。由于土壤盐分通常与土壤水分含量密切相关,水分指数如NDWI和WI等可以用于估算土壤盐分含量。丁建丽等[73]结合Landsat TM数据和野外实测数据,分析土壤盐分与修改型土壤调整植被指数(modified soil-adjusted vegetation index,MSAVI)以及湿度指数(wetness index,WI)的关系,在此基础上提出了MSAVI-WI特征空间的概念,并构建了土壤盐渍化遥感监测指数模型(MWI); Chi等[17]建立了基于改进的陆面综合因子体系的盐分预测模型,该体系预测因子由光谱值、盐度指数、植被指数、热湿指数等组成,研究利用此系统绘制了黄河三角洲湿地的土壤盐分含量分布图。

3)热红外辐射。土壤盐分会改变土壤表面的热红外辐射特征,进而影响热红外辐射亮温和地表温度。Tian等[39]基于热红外遥感对中国甘肃石羊河地区的作物水分状况及其盐胁迫耐受性进行分析,研究指出随着土壤盐分含量的增加,作物冠层覆盖率降低,生长速度减慢; Ivushkin等[59]结合MODIS卫星热图像、NDVI和EVI植被指数以及土壤盐分数据图,证明了卫星反演的冠层温度与土壤盐分水平之间存在显著的相关性; Tajgardan等[74]通过PCA和回归分析方法,利用高级星载热发射和反射辐射数据,绘制了伊朗北部地区的土壤盐分分布图。

4)土壤理化性质。土壤质地也是影响土壤盐分含量的重要因素之一。地表粗糙度变化导致光影区域分布发生变化,进而改变土壤的光学反射特性。土壤团聚体的大小和形状也影响反射特性,直径膨胀的土壤骨料会减少反射,而光滑平坦的表面则反射率更高。此外,土壤有机质含量和矿物质成分同样影响光谱反射。Howari等[75]研究了不同类型的盐分结皮的光谱反射和吸收特征,揭示了其光谱响应差异,并明确了土壤颗粒大小对光谱反射曲线的影响; Song等[51]基于光谱和地形指数与广义可加模型的综合方法对中国黄河三角洲地区的土壤盐分进行了估算。

2.2 遥感反演模型

监测区域土壤盐渍化状况时,受时空变化的影响,构建高鲁棒性的遥感反演模型成为盐渍土监测研究的重中之重。通过遥感技术获取数据和图像信息,并选择合适的特征参量来建立模型,进而评估研究区域内的土壤盐分含量的空间分布。在构建土壤盐渍化遥感监测模型时,通常采用2种方法: 一是统计回归模型,利用线性回归、偏最小二乘回归(partial least squares regression,PLSR)等统计方法,根据遥感特征参量和地面实测数据的经验关系进行建模; 二是机器学习模型,例如BP神经网络(back propagation neural network,BPNN)、SVM、随机森林(random forest,RF)等,通过训练大量样本数据,学习土壤盐分与遥感特征参量之间的关系,构建更复杂和准确的模型。对文中引用的参考文献涉及的数据来源、建模特征参量、建模方法及反演精度等进行了梳理(表6)。

表6   多平台、多源遥感数据土壤盐分指标反演研究

Tab.6  Summary of research study on soil salinity inversion based on multi-platform and multi-source remote sensing data

建模类别建模特征参量反演目标变量建模方法结 果参考文献
统计
回归
模型		光谱参数、植被指数、垂直干旱指数、反射率土壤盐分含量相关性分析、多元线性回归(multiple linear regression, MLR)植被指数中,土壤盐分响应最高决定系数R2=0.577[11]
盐度指数、植被指数、热湿指数土壤盐分含量PLSR研究区四季土壤盐分含量平均值分别为8.00 g/kg,7.53 g/kg,7.83 g/kg和6.90 g/kg[17]
植被指数、盐度指数土壤电导率MLRR2=0.77
R2=0.75		[23]
盐度指数、电导率土壤盐度空间变化回归分析R2=0.65[24]
盐度指数土壤盐分含量MLR相关系数R>0.3[26]
植被指数、冠层温度、植被
株高		作物株高、气孔导度、盐分含量MLRR2=0.64[30]
土壤有机质含量、反射率土壤盐分含量相关性分析、MLR校准: R2=0.684
验证: R2=0.663		[37]
土壤水分、电导率、热红外数据、作物生长数据土壤盐分含量回归分析R2=0.86[39]
反射率、土壤湿度和质地土壤盐分含量、土壤含水量多元逐步回归R2=0.47[49]
盐分指数、植被指数土壤盐分含量线性和非线性回归R2=0.59[53]
盐度指数、植被指数土壤电导率PLSRR2=0.52[57]
盐度指数、亮度指数、植被指数土壤盐分含量MLR、多元逐步回归R2=0.992
均方根误差(root mean square error,RMSE)为0.195 g/kg		[66]
反射率、土壤盐分含量PLSRR2=0.749[67]
植被指数土壤盐分含量PLSRR2=0.50~0.58[68]
植被指数土壤盐分含量线性回归平均误差(mean error,ME)和RMSE分别为-0.61 ds/m和52.2 ds/m[74]
盐度指数、亮度指数、
植被指数		土壤盐分含量MLR标准误差约为12.1 μs/cm[76]
电导率土壤盐分含量PLSR相关系数R= 0.700[77]
pH值、电导率土壤盐分含量PLSR,MLR土壤pH值和电导率模型的R2平均值分别为0.77和0.48[78]
电导率、反射率土壤盐分含量PCA,PLSRPLSR和PCA模型的校准精度分别为R2= 0.862和R2= 0.537[79]
植被指数、盐度指数土壤盐分含量相关性分析R2=0.739(无盐渍化)、0.469(轻度)、0.677(重度)[25]
电导率、盐度指数土壤盐分含量逻辑回归模型R2=0.88,RMSE=20.85 ds/m[27]
植被指数、水分指数土壤盐分含量特征空间MWI与土壤表层盐分含量相关性较高(R=0.844)[73]
植被指数、冠层温度土壤盐分含量统计分析、方差分析F=0.245,P=0.865[59]
盐分指数土壤盐分含量监督最大似然分类法分类精度约为90%[52]
机器
学习
模型		反射率、电导率土壤盐分含量PLSR,SVM相对差异百分比(relative percent difference,RPD)为3.35%[3]
植被指数、盐分指数、反射率土壤盐分含量MLR,PLSR,SVM,RFRF拟合精度最高,训练: R2=0.870,验证: R2=0.766[8]
电导率、反射率土壤盐分含量PLSR,人工神经网络(artificial neural network,anN)PLSR: R>0.81,RPD>2.1%
ANN: R>0.92,RPD>2.3%		[16]
反射率、光谱指数土壤盐分含量MLR,RF,SVM,BPNNR2=0.770(裸土)
R2=0.676(植被覆盖)		[33]
盐分指数土壤盐分含量BPNN,SVR,RF校准: R2=0.835,验证: R2=0.812,RPD=2.299%[34]
反射率、植被指数、盐分指数土壤盐分含量PLSR,BPNN,SVR,RF建模: R2=0.724,RMSE=1.764 g/kg
验证: R2=0.745,RMSE=1.879 g/kg,RPD=2.211%		[35]
反射率土壤盐分含量RFR2=0.95[42]
植被指数土壤盐分含量BPNN,SVM,RFR2=0.885[60]
盐分指数土壤水盐信息BPNNR2=0.624[80]
植被指数、盐分指数、亮度指数土壤盐分含量BPNN建模: R2=0.769
验证: R2=0.774		[81]
地下水深度、灌溉水量、蒸发量土壤电导率SVM建模: MRE=2.14%,验证: MRE=3.48%,预测: MRE=6.37%[82]

新窗口打开| 下载CSV


2.2.1 统计回归

回归分析是一种统计过程,用于估计变量之间的关系,特别是解释自变量与因变量之间的关系。当涉及多个自变量时,称为多元回归。研究表明,在特定条件下,各类遥感特征参量与土壤盐分呈现出较强的相关关系,在土壤盐分建模中具有巨大的潜力[14]。Bouaziz等[76]利用MODIS数据和多元线性回归,在土壤盐分监测研究中发现,将盐分指数(SI2)与波段3(648 nm)反射率结合到统计模型中,能够更好地反映出土壤盐分在空间的扩散情况。Judkins等[83]研究发现,Landsat5 TM传感器的波段7、变换型归一化植被指数(transformed normalized difference vegetation index,TNDVI)以及通过缨帽变换法(tasseled cap transformation,TCT)得到的缨帽指数3和5,与土壤盐分变化呈现较高的相关性,随后将这些光谱参量混合到多元回归模型中,成功进行了土壤表面盐分制图。Allbed等[24]利用遥感数据对沙特阿拉伯哈萨绿洲进行了研究,采用基于遥感指标的统计回归模型,成功预测和绘制了该地区的土壤盐分含量空间变化图; 吴霞等[84]利用Landsat8数据构建了盐渍化评价指数,基于相关分析与曲线回归分析法对宁夏银北灌区土壤盐度进行了定量分析和预测。

然而,基于线性回归的反演模型无法完全解释变量之间的复杂相互作用[85]。混合模型通过组合多个模型来克服各模型中的局限性,可提供更为准确、可靠的模型反演精度,逐渐应用于数据采集较差地区的土壤盐分含量预测[86],如统计模型中的协同克里格和回归克里格法可以利用多个变量之间的回归分析来预测目标变量在低采样区的分布[87]。

2.2.2 PLSR

PLSR方法综合了PCA、典型相关性分析和多元线性回归的优点,在同时多个遥感特征参量建模条件下,能够有效地处理变量之间强共线性和噪声影响等问题[77]。PLSR首先对各变量成分进行有效性验证,并根据相关性自动筛选和重组变量,最终提取具有最佳系统解释力的新成分进行回归建模[88]。近年来,PLSR被广泛用于高光谱数据建模,能够建立光谱数据与土壤理化性质之间的可靠关系。例如,Bai等[78]将PLSR模型应用于HJ-1A高光谱数据,成功绘制了中国松嫩平原北部土壤盐度和碱度的空间分布图,Zhang等[79]建立了土壤盐分与土壤光谱反射率的PLSR模型和PCA模型,并证实了PLSR模型准确性明显优于PCA模型; Sidike等[66]采用PLSR对中国平罗地区的土壤盐分进行估算,结果显示其估算精度明显优于逐步回归方法。

2.2.3 特征空间

光谱特征空间是指由一组光谱特征参数构成的多维空间。在该空间中,每个点代表了经光谱数据计算得到的某个物体或区域的独特光谱特征。通过分析土壤盐渍化参数(如电导率、盐分含量)在二维或三维特征空间中到达某个特征点的距离,可以反映不同盐分含量程度,并清晰揭示不同参数之间的变化趋势。

研究表明,通过各类特征参量构建不同维度的特征空间反演模型,能够实现土壤盐渍化的定量监测[89]。丁建丽等[73]从Landsat TM影像中提取了改良土壤调整后的植被指数(modified soil-adjusted vegetation index,MSAVI)和WI,构建了二维特征空间,并证实其与干旱区绿洲表层的土壤盐分密切相关; 王飞等[90]综合植被和土壤信息,提出了NDVI-SI特征空间概念,对新疆玉田绿洲土壤盐碱化进行定量分析和监测; Guo等[91]利用土壤与植被的协同关系构建了二维特征空间模型,证实该模型在监测黄河三角洲地区土壤盐渍化方面具有较大的潜力; 冯娟等[92]以表面反照率为基础构建特征空间,发现基于地表反照率和土壤调节植被指数的监测模型能够准确、高效地分析研究区域的土壤盐渍化程度。除此之外,三维技术通过引入更多的特征空间,以弥补二维特征空间对盐渍化土壤多因素分析不足。如Yao等[93]基于地表温度、SI和NDVI构建了三维特征空间模型,对中国渭干河-库车绿洲的土壤盐分含量进行了评估,研究结果表明,该模型能够快速、相对准确的监测研究区土壤盐渍化状态。

2.2.4 BPNN

BPNN是一种前馈式的神经网络,由多个神经元组成,每个神经元与前一层的所有神经元相连。与传统的数学方程构建方法不同,BPNN能够学习与识别复杂系统输入和输出数据集之间的非线性关系。其展示出强大的自学习能力、适应性和抗干扰能力,已被证实在土壤理化参数反演方面具有较大的潜力。Wang等[80]利用Landsat8卫星数据成功建立了基于BPNN的土壤水分和盐分的预测模型; Zhang等[81]基于无人机多光谱数据、Sentinel-2A卫星数据以及地面实测盐分数据的多源数据融合监测方法,发现综合估算模型中的BPNN是土壤盐分含量的最佳预测模型,其能够快速且准确地监测区域土壤盐渍化分布。

2.2.5 RF

RF是一种集成学习方法,该方法通过使用随机不相关的决策树来建立分类和回归问题的预测模型[94]。近年来,RF在植被生长指标反演和土壤理化参数估计中得到广泛的应用。如黄晓宇等[8]在干旱地区绿洲土壤盐分研究中,基于Landsat8 OLI影像建立了多个土壤盐分反演模型,并指出相比于经典统计模型,RF建模方法的反演精度更高; Sui等[95]基于水文连通性度量和RF算法,在海岸带土壤盐分研究中开发了基于原始观测和卫星数据的土壤盐分反演模型; 胡婕[96]以高光谱一阶微分、宽带谱指数和窄带谱指数作为输入变量,比较了PLSR和RF方法对土壤盐分的估算效果,结果表明,RF模型能够更好地利用光谱数据预测土壤盐分含量,并在裸土区模型的预测精度表现最佳。

2.2.6 SVM

SVM是一种能够实现结构风险最小化思想的方法,可有效解决小样本、非线性和高维数据等问题。相比传统的统计方法,SVM具有较强的表达能力、泛化能力和学习效率等优势,便于与多源信息相结合,从而实现更高的反演精度[97]。Cai等[98]基于多光谱和纹理特征组合,利用SVM分类器对受盐分影响的土壤进行识别,证实了SVM分类器能够有效提取银川平原土壤盐渍化分布信息; Guan等[82]在土壤电导率值的动态预测中引入SVM理论,构建了土壤盐分动态预测模型,用于盐渍化灌区的灌溉水管理,结果表明,相较于神经网络模型,SVM在土壤电导率值预测方面具有更大的优势。

各类机器学习算法均能使预测模型从光谱和空间模式中进行学习,并根据输入的数据特征进行评估。研究表明,在土壤盐分遥感反演中,BPNN收敛速度较慢,存在局部极小值,且没有结构判定的理论支持[99]; SVM基于结构风险最小化原则,可以较好地解决局部极小值、非线性和高维数等实际问题[13],但其对参数和核函数的选择过于敏感,在求解多分类问题时存在不足[100]; RF对异常值不太敏感[101],在预测结果方面具有高准确性、计算变量重要性的优点,能够对大量预测变量之间复杂的相互作用进行解释[102]。目前,将多机器学习方法相结合构建土壤盐渍化反演模型的手段也在逐渐应用,以尽可能提高模型反演精度。

3 研究和应用展望

伴随着遥感分辨率和相关技术的不断发展,卫星、航空和近地遥感平台将会构成星空地一体化的动态监测网,为土壤盐渍化监测提供高时间、空间、光谱和辐射分辨率的多元数据产品,并表现出以下的研究和应用趋势:

1)多源遥感数据融合。在区域土壤盐渍化监测中,单一传感器数据难以满足大面积、高精度和高效率的需求。因此,融合不同光学遥感数据源,实现高精度、大尺度的土壤盐渍化监测,已逐渐成为当前研究的热点。在区域土壤盐渍化定量反演中,根据应用的目的和目标不同,多源光学遥感数据的融合方式也不尽相同。例如融合具有不同时空分辨率的两个或多个影像,可以改进影像的质量,进而建立基于融合影像的土壤盐分反演模型。或将卫星影像与无人机影像相融合,使用高空间或光谱分辨率的无人机近地影像对卫星影像进行校正,以此克服无人机影像覆盖范围有限的问题,进而构建土壤盐分估算模型以进行大面积反演。另外,还可以将卫星遥感影像和地物光谱融合,地谱分析是定量遥感反演的基础,可以实现土壤盐度的精确分析。地面光谱与卫星影像数据相互补不仅可以提高遥感反演的精度和实用性,而且通过这2种数据的融合还可以提高土壤盐分预测的精度。

2)反演方法对比与协同。遥感技术被认为是监测土壤表面盐分含量的有效手段之一。然而,数据挖掘的不足严重制约了其高效、高精度监测的发展。线性回归模型在建模速度、小样本数据和简单关系方面具有优势,因此成为许多非线性模型的基础,但其预测精度通常低于非线性模型。BPNN算法具有强大的非线性拟合能力和自主学习能力,SVM算法能够避免传统的归纳到演绎的过程,而RF算法更适用于处理非线性数据,具有实现简单、训练速度快和抗过拟合能力强的特点。然而,BPNN的学习速度慢且容易陷入局部最小值,可能需要更多的训练时间才能收敛到最佳解,而且可能停留在非理想解上,无法达到全局最优解; SVM更适用于小样本、非线性的样本集合,这也意味着该模型容易受到数据质量的影响,对超参数调整很敏感。目前,机器学习算法已经在土壤盐渍化监测领域得到广泛应用。面向未来,研究的焦点可以转向开发更先进的模型融合策略,特别是通过创新性地结合不同机器学习算法,来克服土壤盐渍化监测中面临的特殊挑战。此外,为了增强模型的泛化能力和在实际应用中的价值,研究可能将深入探索自适应算法,以自动调整和优化模型参数。同时,将更广泛地利用多元化遥感数据源,如卫星、无人机和地面观测数据的综合,进一步提升监测精度。这些发展不仅能够提升土壤盐渍化反演的准确性,也将为土壤管理与农业生产决策提供更加可靠的支持,推动农业科技的进步。

3)数据同化与深度学习。基于遥感信息的数据同化,主要思路是将遥感观测数据融入陆面模型,并通过不断校正数据模型以达到精确评估与预测的目的。如使用遥感观测数据作为驱动变量,以土壤溶质运移模型和植被生长模型作为模型算子,通过同化算法(如EnKF,4DVar)将遥感观测变量融入模型中,实现目标参量动态模拟与反演。土壤盐分的同化误差不仅受到遥感观测变量和植被生长特征诊断效能的影响,还与物理模型中土壤水盐运移模块和盐分胁迫模块的机理和模型参数鲁棒性相关。目前,土壤参量同化反演的研究主要集中在土壤含水量方面,对于土壤盐分含量的数据同化技术的研究仍有待进一步开展。此外,随着更强大的图形处理器逐渐开发,基于卷积神经网络的深度学习技术的发展彻底改变了图像分析的基本规则,其在机器学习算法组成的基础上,增加了神经网络中的层级和非线性变换以及训练过程的效率,从而建立更为准确和真实的输出。目前,深度学习方法多以RGB影像单一数据源为主,植被特征提取、杂草识别和病虫害诊断为主要目标,针对土壤理化参数的量化研究鲜有报道。未来研究可以进一步探讨利用不同深度学习模型和多模态遥感数据进行综合评估和量化土壤盐渍化信息的有效性和准确性。

参考文献

李洪义. 滨海盐土三维土体电导率空间变异及可视化研究[D]. 杭州: 浙江大学, 2008.

[本文引用: 1]

Li H Y. Three dimensional variability and visualization of soil electrical conductivity in coastal saline land[D]. Hangzhou: Zhejiang University, 2008.

[本文引用: 1]

杨劲松.

中国盐渍土研究的发展历程与展望

[J]. 土壤学报, 2008, 45(5):837-845.

[本文引用: 1]

Yang J S.

Development and prospect of the research on salt-affected soils in China

[J]. Acta Pedologica Sinica, 2008, 45(5):837-845.

[本文引用: 1]

Ali A A A, Weindorf D C, Chakraborty S, et al.

Combination of proximal and remote sensing methods for rapid soil salinity quantification

[J]. Geoderma, 2015, 239:34-46.

[本文引用: 3]

Sreenivas K, Venkataratnam L, Narasimha R P V.

Dielectric properties of salt-affected soils

[J]. International Journal of Remote Sensing, 1995, 16(4):641-649.

[本文引用: 1]

Wang L, Wei Y.

Estimating the total nitrogen and total phosphorus content of wetland soils using hyperspectral models

[J]. Acta Ecologica Sinica, 2016, 36(16):5116-5125.

[本文引用: 1]

Rao B R M, Sharma R C, Ravi Sankar T, et al.

Spectral behaviour of salt-affected soils

[J]. International Journal of Remote Sensing, 1995, 16(12):2125-2136.

[本文引用: 1]

刘全明.

含盐土壤盐渍化雷达反演模拟研究

[J]. 测绘通报, 2014(9):43-46.

DOI:10.13474/j.cnki.11-2246.2014.0290      [本文引用: 1]

准确、快速提取大范围地表土壤盐渍化空间分布是一个迫切急需解决的科学难题。本文以河套灌区解放闸灌域土壤盐分雷达监测为例，研究基于RADARSAT-2数据的盐渍化信息提取。利用成熟的BP神经网络技术，建立了四极化雷达影像灰度值反演土壤盐分的人工智能模型，经实测数据检验能够在一定程度上满足盐渍化监测的需要，优于传统盐渍土分类方法，可促进微波遥感在土壤盐渍化监测中的开拓应用。

Liu Q M.

On radar inversion and simulation of salty soil salinization

[J]. Bulletin of Surveying and Mapping, 2014(9):43-46.

DOI:10.13474/j.cnki.11-2246.2014.0290      [本文引用: 1]

准确、快速提取大范围地表土壤盐渍化空间分布是一个迫切急需解决的科学难题。本文以河套灌区解放闸灌域土壤盐分雷达监测为例，研究基于RADARSAT-2数据的盐渍化信息提取。利用成熟的BP神经网络技术，建立了四极化雷达影像灰度值反演土壤盐分的人工智能模型，经实测数据检验能够在一定程度上满足盐渍化监测的需要，优于传统盐渍土分类方法，可促进微波遥感在土壤盐渍化监测中的开拓应用。

黄晓宇, 王雪梅, 卡吾恰提·白山.

基于Landsat8 OLI影像干旱区绿洲土壤含盐量反演

[J]. 自然资源遥感, 2023, 35(1):189-197.doi: 10.6046/zrzyyg.2022047.

[本文引用: 3]

Huang X Y, Wang X M, Kawuqiati B.

Inversion of soil salinity of an oasis in an arid area based on Landsat8 OLI images

[J]. Remote Sensing for Natural Resources, 2023, 35(1):189-197.doi: 10.6046/zrzyyg.2022047.

[本文引用: 3]

周磊, 贺聪聪, 吕爱锋, 等.

柴达木盆地土壤盐渍化程度快速动态监测

[J]. 测绘科学, 2021, 46(7):99-106,114.

[本文引用: 2]

Zhou L, He C C, Lyu A F, et al.

Research on quick dynamic monitoring of soil salinization in Qaidam Basin

[J]. Science of Surveying and Mapping, 2021, 46(7):99-106,114.

[本文引用: 2]

Zhang X, Huang B, Liu F.

Information extraction and dynamic evaluation of soil salinization with a remote sensing method in a typical county on the Huang-Huai-Hai Plain of China

[J]. Pedosphere, 2020, 30(4):496-507.

[本文引用: 2]

Wu D, Jia K, Zhang X, et al.

Remote sensing inversion for simulation of soil salinization based on hyperspectral data and ground analysis in Yinchuan,China

[J]. Natural Resources Research, 2021, 30(6):4641-4656.

[本文引用: 5]

Sahbeni G, Ngabire M, Musyimi P K, et al.

Challenges and opportunities in remote sensing for soil salinization mapping and monitoring:A review

[J]. Remote Sensing, 2023, 15(10):2540.

[本文引用: 1]

Ma Y, Tashpolat N.

Current status and development trend of soil salinity monitoring research in China

[J]. Sustainability, 2023, 15(7):5874.

[本文引用: 2]

Benediktsson J A, Chanussot J, Moon W M.

Very high-resolution remote sensing:Challenges and opportunities

[J]. Proceedings of the IEEE, 2012, 100(6):1907-1910.

[本文引用: 2]

Mulla D J.

Twenty five years of remote sensing in precision agriculture:Key advances and remaining knowledge gaps

[J]. Biosystems Engineering, 2013, 114(4):358-371.

[本文引用: 2]

Farifteh J, Van der Meer F, Atzberger C, et al.

Quantitative analysis of salt-affected soil reflectance spectra:A comparison of two adaptive methods (PLSR and ANN)

[J]. Remote Sensing of Environment, 2007, 110(1):59-78.

[本文引用: 4]

Chi Y, Sun J, Liu W, et al.

Mapping coastal wetland soil salinity in different seasons using an improved comprehensive land surface factor system

[J]. Ecological Indicators, 2019, 107:105517.

[本文引用: 7]

Hick P T, Russell W.

Some spectral considerations for remote sensing of soil salinity

[J]. Soil Research, 1990, 28(3):417.

[本文引用: 1]

Daliakopoulos I N, Tsanis I K, Koutroulis A, et al.

The threat of soil salinity:A European scale review

[J]. Science of the Total Environment, 2016, 573:727-739.

[本文引用: 1]

Vaheddoost B, Guan Y, Mohammadi B.

Application of hybrid ANN-whale optimization model in evaluation of the field capacity and the permanent wilting point of the soils

[J]. Environmental Science and Pollution Research, 2020, 27(12):13131-13141.

[本文引用: 1]

Ghassemian H.

A review of remote sensing image fusion methods

[J]. Information Fusion, 2016, 32:75-89.

[本文引用: 1]

Wu W, Mhaimeed A S, Al-Shafie W M, et al.

Mapping soil salinity changes using remote sensing in Central Iraq

[J]. Geoderma Regional, 2014, 2:21-31.

[本文引用: 2]

Gorji T, Yildirim A, Hamzehpour N, et al.

Soil salinity analysis of Urmia Lake Basin using Landsat-8 OLI and Sentinel-2A based spectral indices and electrical conductivity measurements

[J]. Ecological Indicators, 2020, 112:106173.

[本文引用: 2]

Allbed A, Kumar L, Sinha P.

Mapping and modelling spatial variation in soil salinity in the Al Hassa Oasis based on remote sensing indicators and regression techniques

[J]. Remote Sensing, 2014, 6(2):1137-1157.

[本文引用: 3]

陈实, 徐斌, 金云翔, 等.

北疆农区土壤盐渍化遥感监测及其时空特征分析

[J]. 地理科学, 2015, 35(12):1607-1615.

DOI:10.13249/j.cnki.sgs.2015.012.1607      [本文引用: 2]

基于MODIS数据,利用归一化植被指数和盐分指数的二维特征空间关系建立土壤盐渍化遥感监测模型,对北疆农区2000年以来的土壤盐渍化状况及其空间动态变化进行了监测分析,并探讨了典型区土壤盐渍化的主要驱动因素。结果表明：① 土壤盐渍化遥感监测指数可以从宏观上定量刻画北疆农区的土壤含盐量;② 北疆农区土壤盐渍化空间特征呈现出总体上逆转、局部严重发展的态势;③ 土壤盐渍化等级在不同时间段的发展或逆转的方向主要由中度向重度及重度向盐土间的相互转化,其中重度盐渍化农用地的转化幅度最大;④ 不同土壤盐渍化等级中盐土的形成与农区降水量和干燥程度具有较好的相关性,未盐渍化（正常）和中度盐渍化与农区有效灌溉面积和农作物播种面积分别呈相关系数较高的正相关和负相关。

Chen S, Xu B, Jin Y X, et al.

Remote sensing monitoring and spatial-temporal characteristics analysis of soil salinization in agricultural area of northern Xinjiang

[J]. Scientia Geographica Sinica, 2015, 35(12):1607-1615.

DOI:10.13249/j.cnki.sgs.2015.012.1607      [本文引用: 2]

The soil secondary salinization caused by brackish water irrigation and so on in agricultural area of northern Xinjiang aggravated the degree of soil salinization, severely reduced agricultural productivity level.To explore the impact of soil salinization in agricultural area of northern Xinjiang on sustainable development of agriculture, the article used the two-dimensional spatial relationship of normalized difference vegetation index and salinity index to build remote sensing monitoring model of soil salinization based on MODIS Data, and analyzed soil salinization status and the spatial dynamic changes of this area since 2000, then discussed the main driving factors of soil salinization in the typical area. The results showed: 1) Remote sensing monitoring indicators of soil salinization can quantitatively characterize soil salinity in agricultural area of northern Xinjiang from the macro. 2) The spatial characteristics of soil salinization in agricultural area of northern Xinjiang showed a reversal of the overall and local serious development situation. 3) The development or reverse direction of soil salinization levels in different periods was mainly mutual transformation of moderate to severe and severe to the saline soil, and the conversion amplitude of severe salinization agricultural land was maximum. 4)There was a good correlation between formation of saline soil with precipitation and degree of dryness and it turned on high positive and negative correlations between not salinization (normal) and moderate salinization with rural effective irrigation area and crop planting area. The research results can provide scientific basis for the prevention of soil salinization.Research conclusion provided scientific guidance basis for the prevention and treatment of soil salinization in the agricultural area of northern Xinjiang and agricultural sustainable development, and offered certain research foundation in quantitative analysis and monitoring work of large scale salinization in arid areas.

陈俊英, 王新涛, 张智韬, 等.

基于无人机-卫星遥感升尺度的土壤盐渍化监测方法

[J]. 农业机械学报, 2019, 50(12):161-169.

[本文引用: 2]

Chen J Y, Wang X T, Zhang Z T, et al.

Soil salinization monitoring method based on UAV-satellite remote sensing scale-up

[J]. Transactions of the Chinese Society for Agricultural Machinery, 2019, 50(12):161-169.

[本文引用: 2]

Farahmand N, Sadeghi V.

Estimating soil salinity in the dried lake bed of urmia lake using optical sentinel-2 images and nonlinear regression models

[J]. Journal of the Indian Society of Remote Sensing, 2020, 48(4):675-687.

DOI:10.1007/s12524-019-01100-8      [本文引用: 2]

Soil salinization is one of the serious environmental problems in arid and semiarid regions. As an effective technique for monitoring soil salinity, remote sensing (RS) technology has been widely used to estimate soil salinity in recent years. Previous studies on soil salinity mapping based on RS images adopted linear regression (LR) between the field measured of electrical conductivity (EC) and the RS data. It is expected that nonlinear regression (NLR) models improve the accuracy of soil salinity mapping over LR. The main objectives of this study are: (1) evaluation the capability of various NLR models for estimating soil salinity based on optical Sentinel-2 RS images, (2) feature selection for soil salinity estimation, and (3) updated and accurate soil salinity map production in the dried lake bed of Urmia Lake. The investigated NLR models include: polynomials, rational functions, powers, exponential, gaussian, logarithmic, and sum of sinusoidal functions with different degrees. All these regression models were calibrated and evaluated separately based on 8 visible and infrared bands of the Sentinel-2 image and 17 salinity indices to estimate soil salinity in the dried lake bed of Urmia Lake (Iran). The evaluation results confirmed the superiority of the NLR models over the LR model for soil salinity estimation. The polynomial degree 3 (Poly-3) based on S3 index (S3=GxRBcould predict EC value with better accuracy than the best LR model (based on narrow NIR band). The R-2 and RMSE of the Poly-3 model were 0.98 and 8.16 dS/m while corresponding values of the best LR model were 0.88 and 20.85 dS/m in test samples, respectively. In general, the results show that the NLR models, along with RS data, have enough accuracy to estimate soil salinity. To compare these methods visually and estimate salt's distribution and concentration in this area, soil salinity maps were predicted by the best NLR model (EC=1.63x10-10xS33-9.95x10-6xS32+0.11xS3-151.7$${\text{EC}} = 1.63 \times 10<^>{ - 10} \times {\text{S}}3<^>{3} - 9.95 \times 10<^>{ - 6} \times {\text{S}}3<^>{2} + 0.11 \times {\text{S}}3 - 151.7$$\end{document}) and the other linear and NLR models in the dried lake bed of Urmia Lake.

Huang H, Deng J, Lan Y, et al.

Accurate weed mapping and prescription map generation based on fully convolutional networks using UAV imagery

[J]. Sensors, 2018, 18(10):3299.

[本文引用: 1]

李冰, 刘镕源, 刘素红, 等.

基于低空无人机遥感的冬小麦覆盖度变化监测

[J]. 农业工程学报, 2012, 28(13):160-165.

[本文引用: 1]

Li B, Liu R Y, Liu S H, et al.

Monitoring vegetation coverage variation of winter wheat by low-altitude UAV remote sensing system

[J]. Transactions of the Chinese Society of Agricultural Engineering, 2012, 28(13):160-165.

[本文引用: 1]

Ivushkin K, Bartholomeus H, Bregt A K, et al.

UAV based soil salinity assessment of cropland

[J]. Geoderma, 2019, 338:502-512.

DOI:10.1016/j.geoderma.2018.09.046      [本文引用: 4]

Increased soil salinity is a significant agricultural problem that decreases yields for common agricultural crops. Its dynamics require cost and labour effective measurement techniques and widely acknowledged methods are not present yet. We investigated the potential of Unmanned Aerial Vehicle (UAV) remote sensing to measure salt stress in quinoa plants. Three different UAV sensors were used: a WIRIS thermal camera, a Rikola hyperspectral camera and a Riegl VUX-SYS Light Detection and Ranging (LiDAR) scanner. Several vegetation indices, canopy temperature and LiDAR measured plant height were derived from the remote sensing data and their relation with ground measured parameters like salt treatment, stomatal conductance and actual plant height is analysed. The results show that widely used multispectral vegetation indices are not efficient in discriminating between salt affected and control quinoa plants. The hyperspectral Physiological Reflectance Index (PRI) performed best and showed a clear distinction between salt affected and treated plants. This distinction is also visible for LiDAR measured plant height, where salt treated plants were on average 10 cm shorter than control plants. Canopy temperature was significantly affected, though detection of this required an additional step in analysis - Normalised Difference Vegetation Index (NDVI) clustering. This step assured temperature comparison for equally vegetated pixels. Data combination of all three sensors in a Multiple Linear Regression model increased the prediction power and for the whole dataset R-2 reached 0.46, with some subgroups reaching an R-2 of 0.64. We conclude that UAV borne remote sensing is useful for measuring salt stress in plants and a combination of multiple measurement techniques is advised to increase the accuracy.

孙刚, 黄文江, 陈鹏飞, 等.

轻小型无人机多光谱遥感技术应用进展

[J]. 农业机械学报, 2018, 49(3):1-17.

[本文引用: 1]

Sun G, Huang W J, Chen P F, et al.

Advances in UAV-based multispectral remote sensing applications

[J]. Transactions of the Chinese Society for Agricultural Machinery, 2018, 49(3):1-17.

[本文引用: 1]

Xu L, Zheng C L, Wang Z C, et al.

A digital camera as an alternative tool for estimating soil salinity and soil surface roughness

[J]. Geoderma, 2019, 341:68-75.

[本文引用: 1]

Xie L, Feng X, Zhang C, et al.

A framework for soil salinity monitoring in coastal wetland reclamation areas based on combined unmanned aerial vehicle (UAV) data and satellite data

[J]. Drones, 2022, 6(9):257.

[本文引用: 3]

Wei G, Li Y, Zhang Z, et al.

Estimation of soil salt content by combining UAV-borne multispectral sensor and machine learning algorithms

[J]. PeerJ, 2020, 8:e9087.

[本文引用: 2]

Yu X, Chang C, Song J, et al.

Precise monitoring of soil salinity in China’s Yellow River Delta using UAV-borne multispectral imagery and a soil salinity retrieval index

[J]. Sensors, 2022, 22(2):546.

[本文引用: 2]

Zhu C, Ding J, Zhang Z, et al.

SPAD monitoring of saline vegetation based on Gaussian mixture model and UAV hyperspectral image feature classification

[J]. Computers and Electronics in Agriculture, 2022, 200:107236.

[本文引用: 2]

Sun M, Li Q, Jiang X, et al.

Estimation of soil salt content and organic matter on arable land in the Yellow River Delta by combining UAV hyperspectral and landsat-8 multispectral imagery

[J]. Sensors, 2022, 22(11):3990.

[本文引用: 2]

Das S, Christopher J, Apan A, et al.

UAV-Thermal imaging and agglomerative hierarchical clustering techniques to evaluate and rank physiological performance of wheat genotypes on sodic soil

[J]. ISPRS Journal of Photogrammetry and Remote Sensing, 2021, 173:221-237.

[本文引用: 1]

Tian F, Hou M, Qiu Y, et al.

Salinity stress effects on transpiration and plant growth under different salinity soil levels based on thermal infrared remote (TIR) technique

[J]. Geoderma, 2020, 357:113961.

[本文引用: 4]

Bhardwaj A, Sam L, Akanksha, et al.

UAVs as remote sensing platform in glaciology:Present applications and future prospects

[J]. Remote Sensing of Environment, 2016, 175:196-204.

[本文引用: 1]

Nevalainen O, Honkavaara E, Tuominen S, et al.

Individual tree detection and classification with UAV-based photogrammetric point clouds and hyperspectral imaging

[J]. Remote Sensing, 2017, 9(3):185.

[本文引用: 1]

Hu J, Peng J, Zhou Y, et al.

Quantitative estimation of soil salinity using UAV-borne hyperspectral and satellite multispectral images

[J]. Remote Sensing, 2019, 11(7):736.

[本文引用: 2]

乔纪纲, 艾彬, 邹春洋.

基于多源遥感的莺歌海滨岸环境特征分析

[J]. 热带地理, 2011, 31(5):456-462.

[本文引用: 1]

采用综合多源遥感方法提取地表信息,结合文献资料和实地调查数据探讨莺歌海滨岸带地面要素的分布特征及其相互关系.主要讨论了综合激光雷达(LiDAR)、极化合成孔径雷达(SAR)、光学和光谱遥感数据进行地表信息提取的适用方法,包括建立精细潮位线模型、近岸微地貌模型,并进行沙化土地、植被、湿地的提取与分类；分析了莺歌海滨岸区的海岸形态、土地退化、湿地与植被分布以及自然景观结构的特征.研究表明,综合遥感方法能有效地支持地表环境特征提取；研究区内滩涂和湿地资源利用不合理；存在较高的土地退化风险；湿地开发与盐场生产活动

Qiao J G, Ai B, Zou C Y.

Analysis of coastal environment in yinggehai with multi-source remote sensing images

[J]. Tropical Geography, 2011, 31(5):456-462.

[本文引用: 1]

采用综合多源遥感方法提取地表信息,结合文献资料和实地调查数据探讨莺歌海滨岸带地面要素的分布特征及其相互关系.主要讨论了综合激光雷达(LiDAR)、极化合成孔径雷达(SAR)、光学和光谱遥感数据进行地表信息提取的适用方法,包括建立精细潮位线模型、近岸微地貌模型,并进行沙化土地、植被、湿地的提取与分类；分析了莺歌海滨岸区的海岸形态、土地退化、湿地与植被分布以及自然景观结构的特征.研究表明,综合遥感方法能有效地支持地表环境特征提取；研究区内滩涂和湿地资源利用不合理；存在较高的土地退化风险；湿地开发与盐场生产活动

Zhang H, Wang L, Tian T, et al.

A review of unmanned aerial vehicle low-altitude remote sensing (UAV-LARS) use in agricultural monitoring in China

[J]. Remote Sensing, 2021, 13(6):1221.

[本文引用: 1]

Mulder V L, de Bruin S, Schaepman M E, et al.

The use of remote sensing in soil and terrain mapping: A review

[J]. Geoderma, 2011, 162(1/2):1-19.

[本文引用: 1]

贡璐, 韩丽, 任曼丽, 等.

塔里木河上游典型绿洲土壤水盐空间分异特征

[J]. 水土保持学报, 2012, 26(4):251-255,278.

[本文引用: 1]

Gong L, Han L, Ren M L, et al.

Spatial variability of soil water-salt in a typical oasis on the upper reaches of the Tarim River

[J]. Journal of Soil and Water Conservation, 2012, 26(4):251-255,278.

[本文引用: 1]

Deng K, Ding J, Yang A, et al.

Modeling of the spatial distribution of soil profile salinity based on the electromagnetic induction technique

[J]. Acta Ecologica Sinica, 2016, 36(20):6387.

[本文引用: 1]

Kahaer Y, Yang S, Tashpolat N, et al.

Hyperspectral estimation of soil electrical conductivity based on fractional order differentially optimised spectral indices

[J]. Acta Ecologica Sinica, 2019, 39(19):7237-7248.

[本文引用: 1]

Xu C, Zeng W, Huang J, et al.

Prediction of soil moisture content and soil salt concentration from hyperspectral laboratory and field data

[J]. Remote Sensing, 2016, 8(1):42.

[本文引用: 2]

蒲智, 于瑞德, 尹昌应, 等.

干旱区典型盐碱土壤含盐量估算的最佳高光谱指数研究

[J]. 水土保持通报, 2012, 32(6):129-133.

[本文引用: 1]

Pu Z, Yu R D, Yin C Y, et al.

Optimal hyperspectral indices for soil salt content estimation on typical saline soil in arid areas

[J]. Bulletin of Soil and Water Conservation, 2012, 32(6):129-133.

[本文引用: 1]

Song C, Ren H, Huang C.

Estimating soil salinity in the Yellow River Delta,Eastern China: An integrated approach using spectral and terrain indices with the generalized additive model

[J]. Pedosphere, 2016, 26(5):626-635.

[本文引用: 6]

Abbas A, Khan S, Hussain N, et al.

Characterizing soil salinity in irrigated agriculture using a remote sensing approach

[J]. Physics and Chemistry of the Earth,Parts A/B/C, 2013, 55:43-52.

[本文引用: 2]

王飞, 丁建丽, 魏阳, 等.

基于Landsat系列数据的盐分指数和植被指数对土壤盐度变异性的响应分析——以新疆天山南北典型绿洲为例

[J]. 生态学报, 2017, 37(15):5007-5022.

[本文引用: 2]

Wang F, Ding J L, Wei Y, et al.

Sensitivity analysis of soil salinity and vegetation indices to detect soil salinity variation by using Landsat series images:Applications in different oases in Xinjiang,China

[J]. Acta Ecologica Sinica, 2017, 37(15):5007-5022.

[本文引用: 2]

郭鹏, 李华, 陈红艳, 等.

基于光谱指数优选的土壤盐分定量光谱估测

[J]. 水土保持通报, 2018, 38(3):193-199,205.

[本文引用: 1]

Guo P, Li H, Chen H Y, et al.

Quantitative spectral estimation of soil salinity based on optimum spectral indices

[J]. Bulletin of Soil and Water Conservation, 2018, 38(3):193-199,205.

[本文引用: 1]

Khan N, Rastoskuev V V, Shalina E, et al.

Mapping salt-affected soils using remote sensing indicators:A simple approach with the use of GIS IDRISI

[C]// 22nd Asian Conference on Remote Sensing, 2001, 5(9).

[本文引用: 2]

Tripathi N K, Rai B K, Dwivedi P.

Spatial modeling of soil alkalinity in GIS environment using IRS data

[C]// Proceedings of the 18th Asian Conference on Remote Sensing,Kuala Lumpur,Malaysia. 1997:20-24.

[本文引用: 1]

Triki F H, Bouaziz M, Benzina M, et al.

Modeling of soil salinity within a semi-arid region using spectral analysis

[J]. Arabian Journal of Geosciences, 2015, 8(12):11175-11182.

[本文引用: 2]

Huete A R.

A soil-adjusted vegetation index (SAVI)

[J]. Remote Sensing of Environment, 1988, 25(3):295-309.

[本文引用: 1]

Ivushkin K, Bartholomeus H, Bregt A K, et al.

Satellite thermography for soil salinity assessment of cropped areas in Uzbekistan

[J]. Land Degradation & Development, 2017, 28(3):870-877.

[本文引用: 2]

Qi G, Zhao G, Xi X.

Soil salinity inversion of winter wheat areas based on satellite-unmanned aerial vehicle-ground collaborative system in coastal of the Yellow River Delta

[J]. Sensors, 2020, 20(22):6521.

[本文引用: 4]

Lyon J G, Yuan D, Lunetta R S, et al.

A change detection experiment using vegetation indices

[J]. Photogrammetric Engineering and Remote Sensing, 1998, 64(2):143-150.

[本文引用: 1]

McFeeters S K.

The use of the normalized difference water index (NDWI) in the delineation of open water features

[J]. International Journal of Remote Sensing, 1996, 17(7):1425-1432.

[本文引用: 1]

Sandholt I, Rasmussen K, Andersen J.

A simple interpretation of the surface temperature/vegetation index space for assessment of surface moisture status

[J]. Remote Sensing of Environment, 2002, 79(2/3):213-224.

[本文引用: 1]

Wan Z, Wang P, Li X.

Using MODIS land surface temperature and normalized difference vegetation index products for monitoring drought in the southern Great Plains,USA

[J]. International Journal of Remote Sensing, 2004, 25(1):61-72.

[本文引用: 1]

Tziolas N, Tsakiridis N, Ben-Dor E, et al.

A memory-based learning approach utilizing combined spectral sources and geographical proximity for improved VIS-NIR-SWIR soil properties estimation

[J]. Geoderma, 2019, 340:11-24.

[本文引用: 1]

Sidike A, Zhao S, Wen Y.

Estimating soil salinity in Pingluo County of China using QuickBird data and soil reflectance spectra

[J]. International Journal of Applied Earth Observation and Geoinformation, 2014, 26:156-175.

[本文引用: 3]

Fan X, Liu Y, Tao J, et al.

Soil salinity retrieval from advanced multi-spectral sensor with partial least square regression

[J]. Remote Sensing, 2015, 7(1):488-511.

[本文引用: 2]

Zhang T T, Zeng S L, Gao Y, et al.

Using hyperspectral vegetation indices as a proxy to monitor soil salinity

[J]. Ecological Indicators, 2011, 11(6):1552-1562.

[本文引用: 2]

Abdullah A Y M, Biswas R K, Chowdhury A I, et al.

Modeling soil salinity using direct and indirect measurement techniques:A comparative analysis

[J]. Environmental Development, 2019, 29:67-80.

DOI:10.1016/j.envdev.2018.12.007      [本文引用: 1]

Soil salinization is a major problem for low-elevation countries like Bangladesh and is expected to worsen due to global warming and associated sea level rise. A constant monitoring of salinity affected areas is imperative to prevent land degradation, agricultural and livelihood losses. With this aim, we developed three soil salinity models with direct, indirect and a combination of both of these soil salinity measurement techniques, for south-western Bangladesh. The five salinity indexes (direct) and eleven environmental variables (indirect) were integrated using Principal Component Analysis, and the predictability of the models was evaluated against ground-based soil salinity measurements from soil survey and land cover maps generated during the model development. Delineation based solely on salinity indexes yielded results that contradicted with models developed from indirect variables and combining all the variables. Results suggest that the salinity model developed by combining direct and indirect techniques has the highest prediction capacity and can also be explained in terms of land cover changes. Therefore, an integrated approach yielded better delineation of salt-affected areas, characterized by active hydrological processes and vegetation cover. The findings and maps produced from this study would provide a new contextual planning tool to policymakers, for devising adaptation strategies in affected areas of Bangladesh.

Khan N M, Rastoskuev V V, Sato Y, et al.

Assessment of hydrosaline land degradation by using a simple approach of remote sensing indicators

[J]. Agricultural Water Management, 2005, 77(1/2/3):96-109.

[本文引用: 1]

Allbed A, Kumar L, Aldakheel Y Y.

Assessing soil salinity using soil salinity and vegetation indices derived from IKONOS high-spatial resolution imageries:Applications in a date palm dominated region

[J]. Geoderma, 2014, 230:1-8.

[本文引用: 1]

张思源, 岳楚, 袁国礼, 等.

基于ENDVI-SI3特征空间的盐渍化反演模型及风险评估

[J]. 自然资源遥感, 2022, 34(4):136-143.

[本文引用: 1]

Zhang S Y, Yue C, Yuan G L, et al.

Salinization inversion model based on ENDVI-SI3 characteristic space and risk assessment

[J]. Remote Sensing for Natural Resources, 2022, 34(4):136-143.

[本文引用: 1]

丁建丽, 瞿娟, 孙永猛, 等.

基于MSAVI-WI特征空间的新疆渭干河-库车河流域绿洲土壤盐渍化研究

[J]. 地理研究, 2013, 32(2):223-232.

[本文引用: 3]

Ding J L, Qu J, Sun Y M, et al.

The retrieval model of soil salinization information in arid region based on MSAVI-WI feature space:A case study of the delta oasis in Weigan-Kuqa watershed

[J]. Geographical Research, 2013, 32(2):223-232.

[本文引用: 3]

Tajgardan T, Shataee S, Ayoubi S.

Spatial prediction of soil salinity in the arid zones using ASTER data,Case study:North of Ag ghala,Golestan Province,Iran

[C]// 28th Asian Conference on Remote Sensing 2007,ACRS 2007, 2007, 3:1712-1717.

[本文引用: 2]

Howari F M, Goodell P C, Miyamoto S.

Spectral properties of salt crusts formed on saline soils

[J]. Journal of Environmental Quality, 2002, 31(5):1453-1461.

PMID:12371161      [本文引用: 1]

Rapid identification and large-scale mapping of salt-affected lands will help improve salinity management in watersheds and ecosystems. This study was conducted to examine spectral reflectance of soils treated with saline solutions containing NaCl, NaHCO3, Na2SO4, and CaSO4.2H2O. Spectral reflectance was measured upon salt crusts formed on two soils (Torrifluvents) subirrigated with saline solutions of 500, 1000, and 1500 mmolc L-1 with a spectroradiometer in the visible and near-infrared region (400-2500 nm). Spectral analyses revealed that samples of gypsum crusts have diagnostic absorption features near 1023, 1225, 1457, 1757, 1800, and 2336 nm, whereas halite crusts have diagnostic absorption features near 1442, 1851, 1958, and 2226 nm. Several broad absorption features were seen in the spectra of the crusts of sodium bicarbonate at 1243, 1498, 1790, 1988, and 2356 nm. The spectrum of soils treated with sodium sulfate exhibited absorption features at 1243, 1472, 1677, 1774, 1851, 1968, and 2245 nm. Crystal size or salt concentrations did not affect the positions of the absorption bands of the salt crusts. However, reflectance increased as particle sizes decreased or with increasing presence of salt crusts. Spectroscopy can be used under certain conditions to identify the presence of primary diagnostic spectral features of gypsum, nahcolite, thenardite, and halite crusts.

Bouaziz M, Matschullat J, Gloaguen R.

Improved remote sensing detection of soil salinity from a semi-arid climate in Northeast Brazil

[J]. Comptes Rendus Geoscience, 2011, 343(11/12):795-803.

[本文引用: 2]

陈文娇, 翁永玲, 范兴旺, 等.

基于光谱转换的土壤盐分反演与动态分析

[J]. 东南大学学报(自然科学版), 2017, 47(6):1233-1238.

[本文引用: 2]

Chen W J, Weng Y L, Fan X W, et al.

Soil salinity retrieval and dynamic analysis based on spectral band inter-calibration

[J]. Journal of Southeast University (Natural Science Edition), 2017, 47(6):1233-1238.

[本文引用: 2]

Bai L, Wang C, Zang S, et al.

Mapping soil alkalinity and salinity in northern Songnen Plain,China with the HJ-1 hyperspectral imager data and partial least squares regression

[J]. Sensors, 2018, 18(11):3855.

[本文引用: 2]

Zhang X, Huang B.

Prediction of soil salinity with soil-reflected spectra:A comparison of two regression methods

[J]. Scientific Reports, 2019, 9(1):5067.

[本文引用: 2]

Wang J, Wang W, Hu Y, et al.

Soil moisture and salinity inversion based on new remote sensing index and neural network at a Salina-alkaline wetland

[J]. Water, 2021, 13(19):2762.

[本文引用: 2]

Zhang Z, Niu B, Li X, et al.

Estimation and dynamic analysis of soil salinity based on UAV and sentinel-2A multispectral imagery in the coastal area,China

[J]. Land, 2022, 11(12):2307.

[本文引用: 2]

Guan X, Wang S, Gao Z, et al.

Dynamic prediction of soil salinization in an irrigation district based on the support vector machine

[J]. Mathematical and Computer Modelling, 2013, 58(3/4):719-724.

[本文引用: 2]

Judkins G, Myint S.

Spatial variation of soil salinity in the Mexicali Valley,Mexico:Application of a practical method for agricultural monitoring

[J]. Environmental Management, 2012, 50(3):478-489.

[本文引用: 1]

吴霞, 王长军, 樊丽琴, 等.

基于多光谱遥感的盐渍化评价指数对宁夏银北灌区土壤盐度预测的适用性分析

[J]. 国土资源遥感, 2021, 33(2):124-133.doi:10.6046/gtzyyg.2020210.

[本文引用: 1]

Wu X, Wang C J, Fan L Q, et al.

An applicability analysis of salinization evaluation index based on multispectral remote sensing to soil salinity prediction in Yinbei irrigation area of Ningxia

[J]. Remote Sensing for Land and Resources, 2021, 33(2):124-133.doi:10.6046/gtzyyg.2020210.

[本文引用: 1]

Alqasemi A S, Ibrahim M, Fadhil Al-Quraishi A M, et al.

Detection and modeling of soil salinity variations in arid lands using remote sensing data

[J]. Open Geosciences, 2021, 13(1):443-453.

[本文引用: 1]

Zare S, Fallah Shamsi S R, Ali Abtahi S.

Weakly-coupled geo-statistical mapping of soil salinity to stepwise multiple linear regression of MODIS spectral image products

[J]. Journal of African Earth Sciences, 2019, 152:101-114.

[本文引用: 1]

Jantaravikorn Y, Ongsomwang S.

Soil salinity prediction and its severity mapping using a suitable interpolation method on data collected by electromagnetic induction method

[J]. Applied Sciences, 2022, 12(20):10550.

[本文引用: 1]

蒋烨林, 王让会, 李焱, 等.

艾比湖流域不同土地覆盖类型土壤养分高光谱反演模型研究

[J]. 中国生态农业学报, 2016, 24(11):1555-1564.

[本文引用: 1]

Jiang Y L, Wang R H, Li Y, et al.

Hyper-spectral retrieval of soil nutrient content of various land-cover types in Ebinur Lake basin

[J]. Chinese Journal of Eco-Agriculture, 2016, 24(11):1555-1564.

[本文引用: 1]

Liu Y, Qian J, Yue H.

Comparison and evaluation of different dryness indices based on vegetation indices-land surface temperature/albedo feature space

[J]. Advances in Space Research, 2021, 68(7):2791-2803.

[本文引用: 1]

王飞, 丁建丽, 伍漫春.

基于NDVI-SI特征空间的土壤盐渍化遥感模型

[J]. 农业工程学报, 2010, 26(8):168-173,8.

[本文引用: 1]

Wang F, Ding J L, Wu M C.

Remote sensing monitoring models of soil salinization based on NDVI-SI feature space

[J]. Transactions of the Chinese Society of Agricultural Engineering, 2010, 26(8):168-173,8.

[本文引用: 1]

Guo B, Yang F, Fan Y, et al.

Dynamic monitoring of soil salinization in Yellow River Delta utilizing MSAVI-SI feature space models with Landsat images

[J]. Environmental Earth Sciences, 2019, 78(10):308.

[本文引用: 1]

冯娟, 丁建丽, 魏雯瑜.

基于Albedo-MSAVI特征空间的渭库绿洲土壤盐渍化研究

[J]. 中国农村水利水电, 2018(2):147-152.

[本文引用: 1]

选取新疆塔里木河流域渭库绿洲为研究区,以landsat8 OLI和野外调查为基础数据,分析盐渍化土壤与反照率(Albedo)和修改型土壤调节植被指数(MSAVI)的关系,并提出Albedo-MSAVI特征空间概念,构建监测土壤盐渍化不同指数模型SDI。结果表明：1、Albedo-MSAVI特征空间分布呈现明显的规律性,即地表反照率与修改型土壤调节植被指数在第一象限呈双曲线分布,且盐渍化与非盐渍化土壤分异明显；2、SDI模型与土壤含盐量的决定系数为0.96,能够较好的反应绿洲盐渍化情况；3、不同盐渍化程度的SDI平均差异与非盐渍化SDI平均差异特征明显,其差值在0.12~0.34之间,因此SDI指数可以作为盐渍化信息快速提取的指标,且SDI模型的构建有利于干旱区大尺度定量监测和评价土壤盐渍化工作。

Feng J, Ding J L, Wei W Y.

A study of soil salinization in Weigan and Kuqa Rivers Oasis based on albedo-MSAVI feature space

[J]. China Rural Water and Hydropower, 2018(2):147-152.

[本文引用: 1]

Yao Y, Ding J, Wang S.

Soil salinization monitoring in the Werigan-Kuqa Oasis,China,based on a three-dimensional feature space model with machine learning algorithm

[J]. Remote Sensing Letters, 2021, 12(3):269-277.

[本文引用: 1]

Mutanga O, Adam E, Cho M A.

High density biomass estimation for wetland vegetation using WorldView-2 imagery and random forest regression algorithm

[J]. International Journal of Applied Earth Observation and Geoinformation, 2012, 18:399-406.

[本文引用: 1]

Sui H, Chen D, Yan J, et al.

Soil salinity estimation over coastal wetlands based on random forest algorithm and hydrological connectivity metric

[J]. Frontiers in Marine Science, 2022, 9:895172.

[本文引用: 1]

胡婕. 基于多源遥感的干旱地区土壤盐分反演研究[D]. 杭州: 浙江大学, 2019.

[本文引用: 1]

Hu J. Estimation of soil salinity in arid area based on multi-source remote sensing[D]. Hangzhou: Zhejiang University, 2019.

[本文引用: 1]

Tan K, Ye Y Y, Du P J, et al.

Estimation of heavy metal concentrations in reclaimed mining soils using reflectance spectroscopy

[J]. Spectroscopy and Spectral Analysis, 2014, 34(12):3317-3322.

PMID:25881431      [本文引用: 1]

A selection of soil samples from reclaimed mining areas were prepared to establish the quantitative inversion models of the soil heavy metal (As, Zn, Cu, Cr, and Pb) concentrations. The concentrations of the soil heavy metals and the visible and near-infrared spectra of the soil samples were obtained in a darkroom. Firstly, smoothing processing was used to smooth the noise in the original spectra, and the spectral transformation techniques of first derivative (FD), continuum removal (CR), and standard normal variate (SNV) were used to promote the model stability and the accuracy of the prediction. Through correlation analysis, the feature bands of the different transformed spectra were extracted. Finally, three different inversion models were adopted and compared, i. e., traditional multiple linear regression (MLR), partial least squares regression (PLSR), and least squares support vector machines (LS-SVM) modeling. The results indicated that: (1) the stability and accuracy of the inversion models established by the different transformed spectra was high, in which LS-SVM was better than PLSR, and PLSR was better than MLR (except for a few cases); and (2) the spectral features extracted from the different transformed spectra had a certain influence on the inversion model, in which the results based on CR transformation and SNV transformation were better than the FD transformation. Therefore, the quantitative estimation of heavy metal concentrations by the use of reflectance spectroscopy is feasible, and the pre-processing is essential to improve the accuracy of the model.

Cai S, Zhang R, Liu L, et al.

A method of salt-affected soil information extraction based on a support vector machine with texture features

[J]. Mathematical and Computer Modelling, 2010, 51(11/12):1319-1325.

[本文引用: 1]

Nawi N M, Khan A, Rehman M Z. A new back-propagation neural network optimized with cuckoo search algorithm[M]//Lecture Notes in Computer Science. Berlin,Heidelberg: Springer Berlin Heidelberg, 2013:413-426.

[本文引用: 1]

Wang X, Zhang F, Kung H T, et al.

Extracting soil salinization information with a fractional-order filtering algorithm and grid-search support vector machine (GS-SVM) model

[J]. International Journal of Remote Sensing, 2020, 41(3):953-973.

[本文引用: 1]

Du P, Xia J, Chanussot J, et al.

Hyperspectral remote sensing image classification based on the integration of support vector machine and random forest

[C]// 2012 IEEE International Geoscience and Remote Sensing Symposium.Munich,Germany.IEEE, 2012:174-177.

[本文引用: 1]

王飞, 杨胜天, 丁建丽, 等.

环境敏感变量优选及机器学习算法预测绿洲土壤盐分

[J]. 农业工程学报, 2018, 34(22):102-110.

[本文引用: 1]

Wang F, Yang S T, Ding J L, et al.

Environmental sensitive variable optimization and machine learning algorithm using in soil salt prediction at oasis

[J]. Transactions of the Chinese Society of Agricultural Engineering, 2018, 34(22):102-110.

[本文引用: 1]

/

〈 〉

京ICP备05055290号-2
版权所有 © 2015 《自然资源遥感》编辑部
地址:北京学院路31号中国国土资源航空物探遥感中心 邮编：100083
电话:010-62060291/62060292 E-mail:zrzyyg@163.com
本系统由北京玛格泰克科技发展有限公司设计开发