铁矿区植被综合胁迫的全波段光谱特征及协同反演模型

Mineral resources exploitation will lead to environmental problems such as resource damage, environmental pollution and geological disasters etc. in mining area. Vegetation, as the typical land coverage of mining area, will be inevitably affected by the environmental problems, which will produce vegetation stress. Affected by the mining and mineral processing, water, nutrients, temperature, pollution and other stress factors can easily produce comprehensive vegetation stress. On the basis of analysis of vegetation stress mechanism in iron mining area, firstly, the diagnostic spectral characteristics and sensitive band of multi-factor vegetation stress in both visible to short-wave infrared(VSWIR) and thermal infrared(TIR) bands would be extracted by use of plant spectroscopic methods. Secondly, the effects of stress factor on the spectral characteristics would be studied on experiments. Finally, the inversion model of vegetation stress will be established by the synergy between VSWIR and TIR spectral data. Reliability and sensitivity of the model will be analyzed to enhance its applicability in mining area. The results will provide theory and methods for the hyperspectral remote sensing and information to support environmental damage analysis and management decision-making in mining area.

矿业开发会导致资源破坏、环境污染和地质灾害等矿区环境问题。植被作为矿区地表典型覆盖，其生长环境必然会受此影响，即产生植被胁迫。受采矿和选矿影响，矿区水分、养分、温度和污染等多种胁迫因子极易对植被产生综合胁迫。以铁矿为例，拟基于铁矿区植被综合胁迫机制分析，采用植物光谱学方法，重点研究铁矿区植被多因子综合胁迫在可见光-短波红外反射波段和热红外发射波段的光谱特征，提取诊断光谱特征及其敏感波段，实现光谱特征的参量化表达；实验揭示胁迫因子对光谱特征的影响规律；研究提出不同波段优势互补的植被胁迫反演模型并进行验证分析；进而，实现铁矿区植被胁迫信息的定量反演与铁矿区环境损害机理分析。研究结果可为铁矿区高光谱遥感应用提供理论与方法，推动高光谱遥感矿区应用，为铁矿区环境损害分析与治理决策提供信息支持。

矿区的矿业开发活动会导致重金属污染、干旱缺水等问题，且上述问题可能同时同地发生。作为矿区典型地表覆盖的植被也会相应地面临重金属污染胁迫、干旱胁迫及综合胁迫。目前，针对胁迫类型（尤其是综合胁迫）的光谱识别特征研究较少，遥感应用缺乏相应的理论与方法支持。为此，以铁矿区为例，对 1）矿区胁迫发生机制，2）矿区植被综合胁迫的可见光-近红外、热红外光谱识别特征，以及 3）综合胁迫的全波段协同遥感模型三方面内容进行了研究。取得了如下结果：(1) 分析表明矿区重金属污染在时间上具有持续性，但在空间分布上与污染源及扩散形式有关；干旱的发生在空间上具有区域整体性，但在时间上具有偶发性；重金属污染与干旱在时空上叠加，则对植被产生综合胁迫。(2) 实验及现场调查表明，铁毒胁迫下玉米叶片的叶绿素含量降低、光谱绿峰升高、红边蓝移；对作为矿区铁毒主因的粉尘降尘量的可见光-近红外光谱特征进行了分析，建立了精度较高的降尘量高光谱反演模型；对铜胁迫、干旱胁迫及旱铜综合胁迫的玉米植株的理化参数及光谱特征进行了研究，确定了叶绿素、叶面积、含水率作为胁迫类型识别的三个关键理化参数，在可见光-近红外波段筛选了表征三个理化参数的光谱指数REP (red-edge position)，dRE (Sum of first derivative reflectance spectra between 680 and 780 nm) 和SIWSI ((R_858.5-R_1640)/(R_858.5+R_1640 ))，在热红外波段确定了12-13μm为胁迫类型识别的敏感波段。(3) 基于可见光-近红外光谱指数和热红外发射率光谱敏感波段，以CK为参照，利用决策树方法建立了干旱、铜、旱铜综合胁迫的协同遥感识别模型。研究结果可为铁矿区植被胁迫的高光谱遥感监测提供理论与方法，从而为铁矿区环境损害分析与治理决策提供信息支持。