Dust Retention Effect On Multi-Scale Vegetation Spectra Based On Three-Dimensional Radiation Transfer Model

China is rich in mineral resources and has a huge output,so the dust pollution is serious,and the dust spatial distribution characteristics caused by different operational pollution sources in the mining area are different.Vegetation plays an important role in the ecosystem,which can absorb and block dust,prevent wind and sand,and regulate climate.Remote sensing is an effective monitoring method widely used in vegetation detection research.Radiation transmission model has become the main research method due to its clear physical meaning.The ground reflected or transmitted spectral signals acquired by the remote sensor are based on pixels.From canopy scale to pixel scale,different factors affect the radiation transmission process of vegetation at different scales.In the periphery of the open-pit mining area,the dust remaining on the surface of the vegetation canopy changes its spectral characteristics,leading to a decrease in the accuracy of remote sensing methods in assessing vegetation growth and inversion of vegetation physiological index.Therefore,it is of great significance to explore the dust retention effect of vegetation spectra in canopy and pixel scale to ensure the accuracy of vegetation monitoring by using remote sensing.In this paper,the typical crop maize in mining area is selected as the research object.The maize spectral variation in canopy and pixel scale under dust retention condition was studied.Firstly,three typical point,polyline and polygon pollution sources of stope,transportation road and tailings pond in the mining area were selected.Based on AERMOD model,dust diffusion spatial distribution was simulated,so as to determine the distribution of dust stagnation in canopy around the pollution source.Then,the three-dimensional radiation transmission model DART was used to simulate the spectral characteristics of maize canopy under different dust retention and vertical distribution conditions.Quantitative analysis of the effect of dust retention on maize canopy spectrum.Based on the dust concentration spatial distribution caused by transportation,the influence of vertical distribution difference of canopy dust on the spectrum was investigated.Finally,the spectral dust retention effect at canopy scale was extended to pixel scale.The spectral dust retention effect of pixel under different FVC was analyzed and verified by remote sensing images.The study shows:①among the three pollution sources,dust pollution caused by shoveling operation,wind erosion dust caused by tailings,and dust pollution caused by road transportation,the dust pollution caused by road transportation is the most serious according to the concentration and influence range of dust pollution,with a concentration of 47.6g/m3 and serious pollution at about 10m on both sides of the road.②At the canopy scale,within the range of 0.75-1.3um,the increased dust retention decreased the canopy reflectivity.In the band of visible light and the band of 1.3-2um,the canopy reflectivity increases as the dust fall gradually increases.At 626nm in the red band,the canopy spectral change rate reached the maximum value of 2.49.At 840nm in the near-infrared band,the highest canopy spectral change rate was-0.41.The vertical distribution difference of entrainment dust has different influence on canopy spectrum,the bottom leaves dust retention has the least influence on canopy spectrum,while the entrainment dust of the top leaves has a greater influence on canopy spectrum.③At pixel scale,spectral response degree of different bands varies with FVC variation.The spectral change rate of the pixels in the near infrared band is at most-0.14,and the spectral change rate of the pixels in the visible band is at most 0.59.The lower the vegetation coverage,the lower the influence of dust on pixel reflectivity.This study lays a foundation for removing dust retention effect and improving inversion accuracy in vegetation remote sensing monitoring.