Estimation Of Leaf Area Index Of Typical Forest Types In Guangxi Based On Multi-Source Data

Forest ecosystem is the largest ecosystem on land and plays an irreplaceable ecological function in regulating global climate,inter-echelon material cycle,energy exchange and maintaining biodiversity.Leaf area index(LAI)is a key parameter to quantify forest productivity and community ecological processes by simulating rainfall interception,photosynthesis,respiration and transpiration in forests.The methods of measuring LAI are mainly divided into direct and indirect methods.The direct method is relatively accurate and can be regarded as the truth LAI,but it is time consuming and difficult to repeat.The indirect method mainly measures LAI by some optical instruments or remote sensing technology,which is convenient and fast,but there is a large uncertainty in the measurement process,and the results are mostly effective LAI,which requires comprehensive comparison and cross-validation.In addition,different forest types have significant differences in stand structure,and the applicabilit y and validity of different indirect methods still need to be further verified.Therefore,the comparative analysis of multiple LAI measurement methods in different forest types and the quantitative assessment of error sources and differences can provide a scientific basis for improving the accuracy of indirect method estimation and rapid and accurate determination of forest LAI.In this study,the natural secondary forest of Fangcheng Yellow Carnellia National Natural Reserve in Guangxi and the Eucalyptus Plantation in Guangxi state-owned Gaofeng Forest Farm were used as the research objects,and the LAI measured by various methods such as plant canopy analyzer(LAI-2200),tracing radiation and architecture of canopies(TRAC),digital hemispheral photography(DHP),terrestrial laser scanning(TLS)were used as the data sources.Based on this,the effects of wood composition and clumping effect on its measurement process were fully considered,and the corresponding correction process was carried out to comprehensively analyze the degree of differences in LAI estimation by different measurement methods and the changes of LAI before and after correction.At the same time,based on the stand factor data and the geodetector method,the differences in the effects of different forest types on different LAI measurement methods were analyzed,and the results showed that:(1)The average LAI of natural secondary forest was 5.40 ± 1.19.The accuracy of LAI estimation by different indirect methods was: LAI-2200 > TLS multi-station > TRAC > TLS single-station > DHP.LAI-2200 and TRAC were overestimated.TLS multi-station and TLS single-station were underestimated,and TLS single-station was underestimated to a greater extent.The LAI measured by DHP was the largest.The clumping effect of natural secondary forest was generally greater than the effect of woody components on LAI measurements.The corrected TRAC measurements were no longer significantly different from the control values,but the DHP,TLS multi-station and TLS single-station measurements were still significantly different from the control values.(2)The average LAI of Eucalyptus plantation was 2.73 ± 0.71.The LAI estimation accuracy of different indirect methods was: TRAC > LAI-2200 > DHP > TLS,which were underestimated to varying degrees.The linear correlation between the LAI measured by TRAC and the control value was weak.LAI-2200 took into account the estimation accuracy and trend correlation at the same time.The LAI measured by DHP and TLS was seriously underestimated.The clumping effect of Eucalyptus plantation was less than that of wood components on LAI measurement.The corrected measured values of LAI-2200,TRAC,DHP,and TLS were further underestimated.(3)In natural secondary forests,LAI-2200 was mainly influenced by mean DBH and canopy depression in the determination of LAI.The different stand factors had less explanatory power for LAI determined by TRAC as well as TLS single station.Mean DBH had more influence on LAI determined by DHP and TLS multi-station.The mean DBH had a greater effect on LAI-2200 determination of LAI when interacted with other factors.The interaction between different factors still had a weak effect on LAI determined by TRAC,and the influence on its spatial distribution was more consistent.The interaction between mean tree height and mean DBH was the dominant effect on the LAI determined by DHP,and there was no significant difference between the other factors on the spatial distribution of LAI determined by them.The interaction between the factors of TLS multi-station was greater than that of TLS single-station,and there was no significant difference between the effects of different stand factors on the spatial distribution of LAI determined by them.(4)In Eucalyptus plantations,the explanatory power of different stand factors for LAI-2200-measured LAI was larger,and the interaction between factors further strengthened the explanatory power.The explanatory power of different stand factors for LAI measured by TRAC and TLS was smaller,and the effect of interactions between factors was still weak.For DHP,it was mainly influenced by mean DBH and mean canopy area,and the interaction between different factors had a greater effect on their measured LAI.The effects of mean tree height and mean DBH on the spatial distribution of LAI measured by LAI-2200 were similar.The effects of different stand factors on the spatial distribution of LAI determined by TRAC and TLS were all more consistent.For DHP,except for mean tree height and canopy density,mean DBH and canopy density,there were also no significant differences among the other factors on the spatial distribution of the measured LAI.