Remote Sensing Quantification Of Stand Vertical Structure Rainfall Erosion Reduction Function For Masson Pine Forests

Soil and water loss is a serious global ecological and environment problem.Vegetation coverage is one of the main measures to reduce and control soil erosion.It is also an important basis for judging and quantifying the possibility and extent of regional soil erosion.It is a key factor to determine soil and water loss based on the scale of remote sensing technology.Stand vertical structure plays two functions in the process of rainfall erosion reduction:one is the interception of rainfall by the forest canopy,which greatly reduces the direct erosion of rainfall on the soil;the other is the erosion reduction effect of the understory vegetation,especially the litter,shrub and grass layer,on the throughfall and stemflow.However,due to the spectral superiority of the stand top layer,the remote sensing images of the"2D"objective world"3D"failed to effectively express the differences in the rainfall erosion reduction function of the stand vertical structure.It is very important to find an effective description method to mine and clarify the expression mechanism of"3D"structure information of forest cover from"2D"multispectral image information,so as to improve the accuracy of the quantitative assessment of soil and water loss by remote sensing.Therefore,based on the program of NSFC—Remote sensing quantitative simulation of rainfall erosion reduction function of vertical structure of forest stands(31770760),Hetian Town was selected as the study area and Masson pine as the research object,with leaf area index(LAI)and leaf area density(LAD)as the expression parameters of stand vertical structure.With the help of the linear transmission relationship of soil basal respiration→soil microbial richness→understory vegetation abundance,the understory vegetation coverage status is indirectly expressed through soil basal respiration,so as to realize the remote sensing expression of"3D"structure information of forest cover.Through the measurement of runoff plot and field rainfall,the rainfall erosion reduction process of the vertical structure of Masson pine stand is studied,the rainfall erosion reduction law of the stand vertical structure is clarified,and the remote sensing quantitative simulation of the rainfall erosion reduction function of the stand vertical structure is realized,which provides theoretical basis and technical reference for remote sensing prediction of soil and water loss.The main results are as follows:(1)Estimation of LAI:Based on the UAV remote sensing data and the measured LAI data,the vegetation index including green normalized vegetation index(GNDVI),green ratio vegetation index(GRVI),modified soil adjusted vegetation index(MSAVI),normalized difference vegetation index(NDVI),renormalized difference vegetation index(RDVI),ratio vegetation index(RVI),structure insensitive pigment index(SIPI),and visible atmospherically resistant index(VARI)was extracted and the impact of different methods(linear regression,multiple stepwise regression,BP neural network,support vector machine,and random forest)and different image spatial resolution scales(0.08 m,0.1m,0.2 m,0.5 m,1 m,2 m,and 5 m)on the estimated LAI difference was studied.The results showed that the pixel scale is 0.5 m,and the fitting coefficient R~2 of the estimated LAI value and the measured value based on the random forest model is the highest,with an overall accuracy of81.95%.The range of the LAI value of the Masson pine forest in the experimental area is between 0-4.35,with an average of 1.46.(2)Estimation of stand LAD:Based on the UAV high repetition rate remote sensing image pair,the point cloud data of the Masson pine forest was reconstructed using structure from motion and multi view stereo.Then,the UAV remote sensing inversion of LAD of Masson pine forest with different voxel sizes(50 mm-100 mm,step length is 10 mm)was realized using contact frequency method.The results showed that the accuracy of LAD estimation of Masson pine forest is the highest when the voxel size is 90 mm,with an estimation accuracy of 87.09%.The overall LAD of Masson pine forest increased first and then decreased with canopy height.(3)Remote sensing indirect estimation of understory vegetation coverage:Based on the carbon balance theory,the modified CASA model and the CENTURY model were used to estimate the soil basal respiration(SBR)of Masson pine forest,with an estimation accuracy of84.33%.Based on this,for the first time,three indicators including soil basal respiration,nitrogen reflection index and yellow leaf factor were selected as indicators for determining the status of vegetation cover under the forest.The principal component analysis was used to construct a remote sensing indirect estimation model of the remote sensing composite evaluation index for the understory vegetation coverage,with model accuracy of 82.62%.(4)Remote sensing simulation of stand canopy rainfall interception mechanism:Based on LAI and LAD data of Masson pine forest,combined with the fractional vegetation coverage(FVC),the canopy interception(I_c)model under different rainfall(P)conditions was established(i.e.LAD-FVC):Two rainfall events(23.4mm and 47.9 mm)in May 2018 in the study area were simulated using five models(LAI,LAD,LAI-LAD,LAI-FVC,LAD-FVC)based on rainfall interception mechanism.The results showed that the LAD-FVC model has the highest estimation accuracy,with an average estimation accuracy of 94.98%,which is0.8-6.9%higher than other models(LAI,LAD,LAI-LAD,LAI-FVC).The maximum canopy interception of Masson pine forests with the rainfall of 23.4 mm and 47.9 mm were 2.71 mm and 4.23 mm,and canopy interception rate were 11.58%and 8.83%,respectively.(5)Simulation analysis of rainfall erosion reduction of undergrowth vegetation(RER):Based on the data of composite evaluation index(CEI),a quantitative remote sensing model of RER was established:RER=13.39×CEI-0.31.The model was used to simulate the RER of Masson pine forest.The simulation results showed that the range of RER in the study area is 0-13.08 t/km~2,with the average value of 3.83t/km~2.(6)Remote sensing quantification of stand vertical structure rainfall erosion reduction function for Masson pine forests:Based on the canopy interception mechanism and rainfall erosion reduction mechanism of understory vegetation of the Masson pine forests,the process of rainfall erosion reduction of stand vertical structure of Masson pine forests was studied by using runoff plot observation data.The results showed that the remote sensing quantitative technology system of rainfall erosion reduction function of stand vertical structure can effectively simulate the canopy interception mechanism of Masson pine forests,which is 27%higher than the variance explained rate of the traditional vegetation coverage model.The average rainfall erosion reductions of Masson pine forests with 23.4 mm and 47.9 mm of secondary rainfall were 6.13 t/km~2and 6.54 t/km~2,respectively.The contribution rate of understory vegetation to rainfall erosion decreases with increasing rainfall,and its value was greater than that of forest canopy on rainfall erosion reduction in scenarios where the rainfall was less than 52.5 mm.The results can provide a theoretical basis and technical guidance for the scale application of remote sensing and a more accurate and quantitative description of the effects and functions of vegetation coverage on rainfall erosion reduction.