| Phyllostachys pubescens is an important bamboo species in China.It is widely distributed in the low mountainous and hilly areas of southern China.It has the characteristics of fast growth,strong reproductive ability,and high economic and ecological benefits.Vegetation chlorophyll,as an important indicator of vegetation physiological state,plays an important role in leaf and canopy scale spectrum.Traditional vegetation parameter estimation methods are time-consuming and laborious,and remote sensing technology can quickly and easily obtain vegetation information at various scales.Therefore,effective inversion of vegetation parameters at different scales is a key factor for the successful application of remote sensing technology to vegetation information management.The study took the Phyllostachys heterocycla cv.Pubescens forest in Tianbaoyan,Yongan City,Fujian Province as the research object.The multi-spectral lens mounted on the drone was used to obtain image data in the study area,and the ground data was collected simultaneously.The leaves,canopy and pixels Screening of the spectrally sensitive characteristic parameters at three scales,the inversion models of chlorophyll content in the bamboo leaf,canopy and pixel scales were constructed by linear,non-linear model and radiation transmission model,respectively,to further determine the effective inversion of chlorophyll content at each scale Methods,and analyze the differences of Phyllostachys pubescens chlorophyll inversion at different scales,at the same time based on the measured samples to study the Phyllostachys pubescens chlorophyll"leaf-canopy","canopy-pixel" conduction mechanism.The main results are as follows:(1)Analyze the spectral characteristic expression of chlorophyll content in the bamboo leaf,canopy and pixel scale.By comparing the leaf hyperspectral with the canopy hyperspectral(400～2500nm),it is found that the overall waveforms and trends of the two hyperspectrals are similar However,the overall reflectance of the leaf spectrum is higher than that of the canopy spectrum.The analysis of the multi-spectral images of unmanned aerial vehicles in the study area and the processed leaf and canopy spectra revealed that the overall spectral trend is relatively close,and the spectral values of the blades in the visible light band are higher,among which are between the green and blue bands.The pixel spectrum and the canopy spectrum have a high degree of overlap,and the increasing trend of the pixel spectrum from the red edge band to the near infrared band is significantly higher than that of the canopy spectrum.(2)Correlation analysis of the chlorophyll content with the original spectrum(R),first-order differential spectrum(FDR)and vegetation index to screen and determine the spectrally sensitive characteristic parameters of chlorophyll content on the bamboo leaf,canopy and pixel scales.The results show that the spectral characteristic parameters on the leaf scale are R553～R687nm,R743～R775nm,FDR550～FDR580nm,FDR675～FDR690nm,medium resolution terrestrial chlorophyll index(MTCI),improved red edge ratio vegetation index(MSR705),normalization Vegetation index(NDVI705,NDVIg-b);canopy scales R480～R520nm,R600-R670nm,FDR500～FDR520nm,FDR598～FDR610,FDR645～FDR685,ratio vegetation index(RVIⅠ,RVIⅡ,SR705,MSR670);pixel scale Rblue,Rgreen,Rred,RREdge,Rnir,Green Normalized Vegetation Index(GNDVI),Structure Independent Pigment Index(SIPI),NDVI,RVI.(3)Based on the chlorophyll spectral characteristic parameters at different scales,multiple stepwise regression,support vector machine and radiation transmission model were used to invert the chlorophyll content of bamboo bamboo,canopy and pixel scales.The results showed that:The multivariate stepwise regression model(Y=54.539 × MTCI-58.503× NDVIg-b+38.929)constructed by combining feature parameters is the most ideal,and its determination coefficient R2 reaches 0.5983;on the canopy scale,using PROSAIL radiation transmission model and support vector machine The model has a high accuracy in inverting the chlorophyll content of the canopy,R2 is 0.5587 and 0.5773 respectively;on the pixel scale,the inversion accuracy of each model with a pixel size of lmlm is better.Among them,the PROSAIL model is used The look-up table has the highest accuracy when inverting chlorophyll,R2 is 0.5944.(4)A comprehensive analysis of the differences in the chlorophyll content inversion process at the bamboo leaf,canopy and pixel scales.The results show that:when performing chlorophyll inversion at different scales,the spectral characteristic expression of chlorophyll,the selection of spectral characteristic parameters and effective There are certain similarities and differences in the inversion methods.When the band is in the range of 480～780nm,the original spectrum and the first-order differential spectrum contain more effective spectral information;the vegetation characteristic parameters that are more sensitive to chlorophyll mainly involve the normalized vegetation index and the ratio vegetation index;the chlorophyll is effectively inverted at each scale In the method,the leaf scale is the best with the multiple stepwise regression model combining the spectral characteristic parameters.The canopy and pixel scales are the PROSAIL radiation transmission model and the support vector machine nonlinear model.(5)Based on the estimated differences of Phyllostachys pubescens chlorophyll at different scales,the transmission mechanism of Phyllostachys pubescens chlorophyll in "leaf-canopy" and "canopy-pixel"was studied and discussed through 40 samples.The results showed that:-There is a certain linear relationship between the "canopy" scales,and this relationship is quantified as Y=1.5723CEast+2.12Csouth-31.477;at the same time,in the "canopy-pixel",it is found that when the chlorophyll is retrieved from the image,the pixel Selecting different number of pixels(n)in scale will make the inversion result different.When the number of corresponding pixels in a canopy is 2(1m × 2m),3(1m × 3m),and 4(2m × 2m),the inversion effect is the best.And the transmission mechanism of chlorophyll "canopy-pixel" of Phyllostachys pubescens in this study area is determined as:(?)... |
