关键词: 高光谱遥感, 滞尘能力, 光谱特征, 绿化植物, 广州市

Abstract:

Hyperspectral remote sensing provides highly efficient, convenient, and non-destructive technical means for the quantitative study of the dust retention content of plants. Based on common greening plants in Guangzhou and hyperspectral remote sensing technology, this study designed a semi-closed greenhouse experiment scheme and explored the spectral characteristics of different plants under dust pollution. The research results show that: 1) Dust-retention time is an important factor that affects the accumulated dust retention content of greening plants. The amount of dust retention content increases with time but decreases after reaching saturation. There are differences in the amount and time of saturation of different plants, and the order of saturation of dust retention from large to small is: Cordyline fruticosa(L.) A. Cheval > Loropetalum chinensis(R.Br) Oliv. Var rubrum Yieh > Ficus microcarpa L. f. cv ‘Golden leaves’. In the specific green vegetation screening, Cordyline fruticosa(L.)A. Cheval can be given priority consideration. 2) In the visible wavelengths affected by foliar dust, the spectral reflectance of the leaves increases, with increases in the amount of dust retention; in the near-infrared wavelengths, the spectral reflectance is inversely proportional to the amount of dust retention. 3) Dust retention content directly or indirectly affects the spectral characteristics of plant leaves and has nothing to do with the three sides of the plant. The dust retention content is directly proportional to the red valley reflectance and green peak reflectance and has no significant relationship with the red valley and green peak positions. There is a negative correlation between the dust retention content and the Normalized Difference Vegetation Index (NDVI). The NDVI values of different plants are affected by dust retention content to varying degrees; in order from large to small, these are: Cordyline fruticosa(L.)A. Cheval > Loropetalum chinensis(R.Br) Oliv. Var rubrum Yieh > Ficus microcarpa L. f. cv ‘Golden leaves’. The model established by the NDVI and dust retention content can quickly and non-destructively evaluate the dust retention capability of plants, which can provide a scientific basis for urban air quality monitoring. The experimental scheme of the semi-enclosed greenhouse designed in this study provides a new idea for the long-term sustainability study of the dust retention effect of plants. Furthermore, the dust retention effect of plants differs according to their environments, requiring further research. Important future research areas include the inversion band selection of spectral estimation models, the construction of high-precision spectral estimation models, the influence of different plant species on spectral estimation models, and the inversion of dust retention on a large area scale by combining spectral estimation models of dust retention content with the remote sensing images of hyperspectral UAV, satellite, etc.

Key words: Hyperspectral Remote Sensing, dust retention ability, spectral characteristics, green plants, Guangzhou