煤矿区土地复垦及复垦监测工作, 对于我国土地利用和生态环境治理具有重要意义。 微生物复垦技术能够促进植物吸收利用矿质养分和水分, 增强土壤肥力, 对矿区生态恢复具有显著作用。 监测和评价土地复垦效应对植物生长影响的传统方法, 通常采用野外采集植物和土壤样本并进行室内分析, 但这些方法不仅破坏植物根系原状土壤, 造成植株损伤, 而且耗费人力、 物力, 时效性差。 高光谱遥感技术具有数据获取速度快、 信息量大、 精度高且无须离体破坏植株等优点, 对于土地复垦监测有非常大的潜力。 目前, 土地复垦效应遥感监测相关研究仍以观测盆栽大豆、 玉米等作物的叶片光谱分析为主。 实际上, 卫星遥感数据观测到的是冠层光谱, 并非叶片光谱, 但目前还没有通过植被冠层光谱对矿区土地复垦进行监测的研究成果出现。 植被冠层光谱不仅受到叶片光谱的影响, 还受到植株长势、 下垫面等其他因素的影响, 光谱特征变化更为复杂。 矿区植被冠层光谱特征对于土地复垦效应的敏感度分析, 是对矿区植被理化参量进行定量反演的基础, 也是限制高光谱技术应用于大面积土地复垦监测的主要瓶颈。 于煤炭矿区土地复垦实验基地开展野外冠层光谱观测实验, 获取了接菌组和对照组野外植株冠层光谱数据, 并从光谱波形变化和光谱特征参量变化两方面综合分析了植被冠层光谱对土地复垦的敏感性。 冠层光谱波形方面, 分别采用标准差和光谱敏感度作为组内和组间光谱波形差异的有效指标; 冠层光谱特征参量方面, 选取了植被红边、 黄边、 蓝边、 绿峰、 红谷等典型光谱特征, 计算获取其位置、 斜率、 面积等特征参量, 并通过描述性统计和单因素方差分析研究了这些冠层光谱特征参量对土地复垦效应的敏感性, 挑选出矿区土地复垦监测的有效特征参量。 研究表明, 接菌组和对照组冠层光谱的主要波形变化趋势一致, 但接菌组植株的生长状况更稳定, 不同植株之间差异较小, 且绿峰和红谷两个特征更突出。 这说明土地复垦能够减少植株间冠层光谱差异, 增强植被典型光谱特征, 而绿峰和红谷对土地复垦有较高的光谱敏感度。 光谱特征参量方面, 绿峰、 红谷、 红边波长在土地复垦作用下显著向长波方向移动, 而此前叶片光谱研究中对土地复垦较敏感的红边、 蓝边斜率变化并不显著。 这说明, 野外植被冠层光谱分析结果与实验室植被叶片光谱分析的结果并不完全一致, 这可能和植被类型、 生长周期、 土壤背景光谱干扰等因素相关。 在采用卫星或航拍遥感数据进行矿区植被环境监测时, 所获取的都是植被冠层光谱, 因此本研究所得到的结论具有更强的参考意义和实际应用价值。

The land reclamation and its monitoring in coal mining area is of great significance to land use and ecological environment governance in China. The Microbial Reclamation Technology can promote the plant’s absorption and utilization for mineral nutrient and water, and strengthen the soil fertility, having a significant effect on ecological restoration of mining area. The traditional method for monitoring the land reclamation on plant growth is usually collecting plant and soil samples in the field for indoor analysis. But this method not only destroys rhizosphere soil, causing damages to plants, but also consumes large quantity of manpower, material resources, and time. The hyperspectral remote sensing technology has the advantages of fast data acquisition, large information, high precision and nondestructive for plants, having great potential for land reclamation monitoring. At present, the research on the monitoring of land reclamation through remote sensing still stays at the laboratory level of observing the leaf spectra of potted soybean, corn and other crops. In fact, the observation of satellite remote sensing data is the canopy spectra, not the leaf spectra, but now there is no research result on the monitoring of land reclamation in the mining area based on vegetation canopy spectra. The vegetation canopy spectrum is not only affected by the leaf spectrum, but also influenced by other factors such as plant growth condition and underlying surface, and the spectral change is more complicated. Analysis on susceptibility of vegetation canopy spectra in coal mining area to land reclamation is the basis of the quantitative inversion for the physical and chemical parameters of vegetation, and the main bottlenecks of the hyperspectral technology to be applied in large area reclamation monitoring.This research performs field experiment on vegetation canopy spectral observation in the land reclamation basement of coal mining area, obtains the wild plant canopy spectral data of reclamation group and the control group, and comprehensively analyses the spectral susceptibility of vegetation canopy spectra to land reclamation, from the aspects of spectral waveform and spectral feature parameters. In terms of the spectral waveform of canopy, standard deviation and spectral sensitivity are used as effective indicators for the difference of spectral waveform within each group and between two groups. As for the canopy spectral feature parameters, we selected the red edge, yellow edge, blue edge, red valley and green mountain as typical spectral features, calculated their parameters (such as wavelength position, slope and area), and performed descriptive statistical analysis and one-way ANOVA to investigate the canopy spectral feature parameters’ sensitivity to the effect of land reclamation. Results showed that the canopy spectral waveform trend of the reclamation group is similar with that of the control group, but the plants in the reclamation group have smaller spectral difference, and their vegetation typical features, such as green peak and red valley, are more prominent. This indicated that land reclamation can reduce the canopy spectral difference between plants and strengthen the typical vegetation spectral features, while green peak and red valley are the most sensible spectral features to land reclamation effect. In terms of the specific canopy spectral feature parameters, the wavelengths of green peak, red valley and red edge tend to shift to longer wavelength significantly under the function of land reclamation, but the slope of red edge and blue edge, which are sensitive to land reclamation changes in previous studies on leaf spectra, are not significant any more in this research. This showed that the analysis results based on vegetation canopy spectra in the field are not in consistent with previous laboratory analysis results of leaf spectra, which may be caused by vegetation types, growth cycle, the interference of soil background, etc. When monitoring the vegetation environment in mining area based on satellite or aerial remote sensing, the data obtained are canopy spectra, not leaf spectra. Therefore, the conclusions of this research have strong reference value for future practical applications.