Research On Forest Height Retrieval And Monitoring Method Using Space-borne LiDAR Data

Forests provide ecological services such as climate regulation and water conservation as the main body of the terrestrial ecosystem.It is crucial to the improvement of the ecological environment and the maintenance of ecological equilibrium.Monitoring forest canopy height and change is critical for understanding the forest carbon cycle and global climate change.Li DAR,being an active remote sensing technology,can gather forest vertical structure information rapidly and precisely,and it has unique benefits in forest height inversion and change monitoring.Large observation range and regular repeated observation are two characteristics of satellite remote sensing.Spaceborne Li DAR has a large coverage area.It has a lot of advantages when it comes to quantitative inversion of forest metrics on a regional basis.Because spaceborne Li DAR data has such a broad coverage,it has become one of the greatest options for quantitative inversion of forest vertical structural characteristics across enormous areas.The current generation of spaceborne Li DAR GEDI,for example,has a tiny spot diameter and a high sample density.It can precisely calculate the height of the forest canopy,however the laser points have a discontinuous distribution.As a result,various satellite remote sensing pictures(such as landsat-8,sentinel-2,and others)must be integrated to achieve large-scale high-resolution forest height mapping,which allows for forest height change monitoring.The purpose of this paper is to carry out precision verification,error analysis and monitoring of forest height change for the new generation of Li DAR GEDI data.There are two specific objectives:1)The influence of location error,terrain slope,slope direction,vegetation coverage,azimuth,acquisition time,beam type and different forest type factors on the accuracy of GEDI data is discussed;2)A method for monitoring vegetation height change based on forest height map and spaceborne Li DAR GEDI data.The main contents and findings are as follows:(1)Forest height accuracy evaluation and error analysis.The terrain feature parameters are extracted using the parameters of Li DAR data GEDI and ASTER GDEM data,The accuracy and error analysis of GEDI elevation and forest height are investigated,as well as the airborne Li DAR data as validation data.The results demonstrate that by addressing geolocation inaccuracies,GEDI elevation and canopy height accuracy can be greatly improved,and the RMSE value may be lowered by around 50%;Vegetation coverage is the most important element determining the accuracy of forest height extraction,followed by slope;The slope aspect and slope are the two most important elements determining the accuracy of ground elevation extraction.The data accuracy is higher when the vegetation covering is greater than 25%;In mild slope locations with a slope of 0-5°,the accuracy of ground elevation and forest height is the highest.The findings will serve as a solid foundation for screening and applying GEDI data.(2)The study area’s forest height mapping and change monitoring.The forest height extrapolation model is constructed using the random forest algorithm,and two spatial distribution maps of forest height in the study area are created using the spaceborne Li DAR GEDI data as the main data source,combined with the characteristic parameters extracted from Landsat-8 and Sentinel-2 data,respectively.The forest height map R~2 coupled with Landsat-8 was 0.59 and 0.56,with RMSE of 3.66m and 2.70m,respectively,and the forest canopy height map R~2 combined with Sentinel-2 was 0.57 and 0.47,with RMSE of 3.73m and2.86m,respectively.Second,the change in forest height is estimated using the forest height map from 2019 as well as aerial Li DAR data from 2014.The R~2 values in the JERC experimental area are 0.46 and 0.41,respectively,with RMSE values of 4.85m and 4.93m,and the R~2 values in the OSBS experimental area are 0.32 and 0.34,respectively,with RMSE values of 3.36m and 3.34m.(3)GEDI Li DAR data and aerial Li DAR data are being studied for monitoring strategies.The forest height parameters(RH98)extracted from spaceborne Li DAR GEDI data in 2019and CHM data in 2014 are used to track the change in forest height.The findings reveal that forest height change is quite accurate(JERC and OSBS had R~2 values of 0.70 and 0.48,respectively,RMSE values of 2.75m and 1.61m).As a result,this approach can accurately reflect the change in forest height in the experimental area.(4)Making a map of the change in forest height.Through the analysis of the results of(2)and(3),in the JERC and OSBS experimental areas.The tree height change result is used as the primary data source,combined with the characteristic parameters extracted from landsat-8data,the distribution map of forest height change in the experimental area is made.The R~2values of JERC and OSBS experimental areas are 0.40 and 0.38,0.52 and 0.47respectively,The RMSE values of JERC and OSBS experimental areas are 3.97 and 2.04m,2.83m and 1.49m respectively.It serves as a foundation for creating a global forest height change distribution map.There are 48 figures,8 tables and 71 references in this paper.