Research On Leaf Area Index Measurement By Hemispherical Photography And Uncertainty

Leaf area index(LAI),defined as half of the surface area of all leaves per unit area,is an important parameter of vegetation structure that affects photosynthesis,transpiration and rainfall interception processes.With the increasing investment in ecological environment protection and agricultural production in our country,the quantitative observation of LAI has become a research hotspot in the field of ecology and agriculture.Among many leaf area index measurement methods,the digital hemispheric photography(DHP)based measurement method is widely used because of the advantages of easy operation,low cost,and permanent recording of canopy structure,but the method is affected by various factors in actual use,resulting in biased measurement results.This study addresses a the uncertainties in the practical application of digital hemispheric photography to measure LAI,combining simulation scenarios and real scenario experiments.The specific research contents are as follows:(1)The theory of leaf area index measurement by hemispheric photography is studied based on both the Lambert-Beer law and probabilistic model.It was found that the LAI measurement results were underestimated by up to 68% due to the leaf clumping effect.For this problem,the LX method and the LXG method were used to correct them,and the experiments showed that the LXG method had better correction effects in both crop and woodland vegetation types.For the problem of choosing the ring width for the single-angle method,For the single angle method ring width selection problem,through the experimental analysis,comprehensive accuracy and calculation quantity,determining the ring width is 55°～ 60°.(2)The LAI measurement method was optimized.The single angle method and Miller method were found to have their own applicability ranges,and this study proposes an adaptive selection method according to the canopy opening levels.The experiments show that the accuracy of the measurement results of the optimized method is improved compared with both the single-angle method and the Miller method.(3)Analysis of the causes of LAI underestimation due to automatic exposure mode from the perspective of imaging.Improved image quality and reduced LAI underestimation through the use of polarizers.Combining digital hemispheric photography and multi-exposure fusion algorithm,multiple canopy images with different exposure times are fused to reconstruct the information which is lost in vegetation canopy images in automatic exposure mode.Using the LAI-2200 measurement results as a reference,compared with the automatic exposure mode measurements,the coefficient of determination improved by 0.139 and the RMSE decreased by 0.5,indicating that the method can significantly improve the accuracy of LAI measurement by hemispheric photography.(4)Proposed an optimal exposure method for digital hemispheric photography based on light intensity.For the problem that the optimal exposure time of hemispheric photography method is difficult to determine,the thesis constructs a model of optimal exposure time and light intensity under the canopy.Through experimental verification,the results show that using the LAI-2200 measurement results as a reference,the coefficient of determination of the model in the thesis reaches 0.806 and the RMSE is0.37,which improves the coefficient of determination by 0.324 and reduces the RMSE by 0.37 compared with the automatic exposure mode measurement results.This method conveniently determines the optimal exposure time for digital hemispherical photography and reduces the dependence of hemispherical photography on imaging light conditions.so that the digital hemispheric photography method can be used in a wider range,which is a guideline for determining the optimal exposure time for automatic measuring instruments.