CN108959705B - Method for predicting subtropical forest biomass - Google Patents
Method for predicting subtropical forest biomass Download PDFInfo
- Publication number
- CN108959705B CN108959705B CN201810538163.3A CN201810538163A CN108959705B CN 108959705 B CN108959705 B CN 108959705B CN 201810538163 A CN201810538163 A CN 201810538163A CN 108959705 B CN108959705 B CN 108959705B
- Authority
- CN
- China
- Prior art keywords
- biomass
- model
- trunk volume
- sample
- forest
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S17/00—Systems using the reflection or reradiation of electromagnetic waves other than radio waves, e.g. lidar systems
- G01S17/88—Lidar systems specially adapted for specific applications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
- Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation
Abstract
The invention discloses a mixed effect model for predicting large-area subtropical forest biomass, belongs to the technical field of forest management, and relates to a forest biomass extraction technology. A technology for extracting the forest biomass of the subtropical zone in a large area is designed by utilizing a relatively convenient data acquisition way and less capital personnel investment. The technology mainly comprises five stages of extraction of laser radar data characteristic variables, estimation of ground sample plot data trunk volume, model construction and model verification based on vegetation types, and biomass calculation based on the trunk volume and a biomass relative growth model. The technology improves the biomass estimation precision of the laser radar method in large-scale forests, and provides a simplified technical solution for the airborne laser radar in forestry investigation application of large-area subtropical forests.
Description
Technical Field
The invention belongs to the technical field of forest management, relates to a forest biomass extraction technology, and particularly relates to an extraction technology for large-area subtropical forest biomass. The invention discloses a mixed effect prediction model for predicting the biomass of a large-area subtropical forest based on a laser radar and a vegetation type.
Background
Subtropical forests play an important role in the global carbon cycle, and deep understanding of their ecological functions under varying climatic conditions requires accurate estimation of their forest structure. Lidar is an advanced remote sensing technology that can provide higher accuracy to estimate forest structure information than optical or radar images. However, the application of lidar in subtropical forests is often limited to small areas, small ranges and small dimensions.
The large-area forest is more diversified in classification and structure, so that the derivative indexes of the laser radar and various forest attributes have more complex relations, and the accuracy of predicting the attributes of the large-area forest by using the airborne laser radar is reduced. As previously noted, when airborne lidar was used to estimate tropical forest biomass, the coefficients were determined to be as high as 0.85 (Proceedings of the National Academy of Sciences of the United States of America, published by Asner et al in 2010) and as low as 0.38 and as high as 50% relative error (Chrin 2015 published by Remote Sensing, published by Chen et al in 2012). The difference of research results shows that the airborne laser radar needs more research to carry out extensive tests, the research conclusion of predecessors cannot be simply extended to a new research object, and effective methods need to be developed respectively for the estimation of structural variables of different forest types. Considering that the diversity of large-area forests in classification and structure increases the difficulty of accurate estimation, forest structure variable estimation based on vegetation types may improve accuracy.
Forest biomass fixes 82.5% of the carbon reserves of land vegetation, which is not only an important index in the carbon budget assessment of land ecosystems, but also the basis for the study of many forestry and ecological problems (such as material circulation, energy flow, etc.). At present, the patent is mainly used for estimating biomass by optical remote sensing inversion, such as a forest biomass remote sensing inversion method based on spectral curve characteristic differentiation (publication number: CN 106291582A) and a desert grassland green biomass remote sensing monitoring upscaling method (publication number: CN 106294991A), and a patent of applying a laser radar to large-area forest biomass does not exist. When the laser radar is applied to biomass estimation, the trunk volume can be estimated through the laser radar based on the vegetation type, and then the biomass is estimated through a relative biomass model.
In conclusion, it is necessary to develop a mixed effect model for predicting the biomass of the large-area subtropical forest based on the laser radar and the vegetation type.
Disclosure of Invention
The purpose of the invention is as follows: the mixed effect model for predicting the forest biomass of the large-area subtropical zone based on the laser radar and the vegetation type is provided, the precision is effectively improved, and the cost is reduced.
In order to achieve the purpose, the invention adopts the following technical scheme:
a mixed effect model for predicting the trunk volume of a large-area subtropical forest based on a laser radar and a vegetation type is characterized by comprising the following steps:
1) Acquiring laser radar data, preprocessing the data and extracting characteristic variables by using a sensor;
2) Obtaining ground sample plot data and estimating a trunk volume in a design sample plot of an area to be measured, and determining a vegetation type of the sample plot level;
3) Taking the trunk volume information obtained by ground actual measurement and summary as a dependent variable, taking the characteristic variable of the laser radar as an independent variable, and establishing a mixed effect model of an optimal nonlinear parameter regression model based on the vegetation type;
4) Mixed effect model performance was evaluated using 10-fold Cross Validation (CV);
5) Biomass was calculated based on the trunk volume estimated by lidar and the relative growth model of biomass established from the parse wood, B = aV + B.
In the step 1), a sensor is adopted to obtain LiDAR point cloud data in a to-be-detected area of a forest in a subtropical hill, and the sensor records complete laser pulse return waveform information. Processing the laser radar point cloud data by using Tiffs (laser radar data filtering and forest tool kit) software to separate ground echoes, and manually editing and perfecting classification results; and then generating a 1-meter Digital Terrain Model (DTM) by interpolating the ground echoes. The height of each laser spot is calculated by subtracting the height of the corresponding DTM cell below it from the Z coordinate of each laser spot. Calculating 14 characteristic variables of the laser point height according to LiDAR point cloud data: 3 statistical variables (mean H) mean Standard deviation H std And kurtosis H kurt ) (ii) a 10 height percentile variables：H ptc10 、H ptc20 、…、H ptc100 (ii) a Mean square value H of 1 mean height qm (high points have higher weight). The lidar indices are calculated from echoes of all altitudes to characterize horizontal and vertical canopy structures.
In the step 2), a plurality of square sample plots are arranged in the range of the area to be measured, and in the sample plot investigation process, for trees with the diameter at breast height of more than or equal to 5 cm and bamboos with the diameter at breast height of more than or equal to 2 cm, the tree species, the tree height and the diameter at breast height of a single tree are measured one by one; estimating the volume of the trunk according to the single tree survey data and the different-speed growth model; and summarizing the sample plot information to obtain the unit trunk volume (cubic meters per hectare) of each sample plot. Establishment of land-level vegetation types according to the chinese national vegetation classification system. Including evergreen broad-leaved forest, coniferous and broadleaf mixed forest and bamboo forest.
In step 3), the trunk volume information collected by ground actual measurement is used as a dependent variable, the prediction index obtained by the laser radar is used as an independent variable, and a mixed effect model of nonlinear parameter regression is established based on the vegetation type, wherein the mixed effect model comprises a Simple Power Model (SPM) and a multi-power model (MPM). Both have the same model formula:
wherein H i Are the individual lidar indices described in step (1) generated from a set of lidar points. n is the number of lidar measurements in the model. For the simple power model, n =1; for the multivariate power model, n is greater than or equal to 2.
To develop a nonlinear multivariate power model to predict trunk volume, we selected the lidar criteria as follows: first, the trunk volume and lidar indices are logarithmically transformed, and then forward stepwise regression is used on a logarithmic scale to select statistically significant variables. The p-value thresholds for the inclusion and exclusion variables are set to 0.2 and 0.1, respectively, intentionally higher than the typical value of 0.05 in the statistical software package to reduce the risk of removing important lidar indicators. When necessary, the model was redesigned by removing any statistically insignificant (p-value > 0.05) variables.
In step 4), the model performance was evaluated using 10-fold Cross Validation (CV), which steps were as follows: (1) randomly dividing the sample plot into 10 parts; (2) The trunk volume observed for any 9 samples was used to calibrate the model to predict the trunk volume for the remaining 1 sample; (3) Repeating the previous steps for 10 times, and sequentially predicting to obtain the remaining 9 stock volumes of the sample prescription. (4) Calculating a cross-validation determination factor (R) based on observed and predicted trunk volume 2 ) And the Relative Root Mean Square Error (RRMSE) is as follows:
wherein, y i Is the volume of trunk material observed by the sample plot,
is the trunk volume predicted by the model, n is the number of squares (n = 140), ->
Is the average trunk volume observed for 140 samples. When using a mixed effect model, if the prediction model does not include the vegetation type present in the calibration model plot, then the mixed effect model without the random term is used for the prediction.
In the step 5), the establishment of the biomass relative growth model comprises the following steps: 1) Measuring the biomass B of the analytic wood; 2) Obtaining an observed trunk volume V according to the single tree investigation method and the different-speed growth model in the step (2); 3) From the measured biomass B and the observed trunk volume B, a biomass relative growth model B = aV + B is established. The trunk volume V estimated by the laser radar from the first four steps was substituted into the established biomass relative growth model B = aV + B to calculate the biomass B.
The invention aims at 19000km of southeast China 2 The forest tree trunk volume is predicted and simulated based on the airborne laser radar data, and then the biomass is estimated by using a biomass relative growth model. The method is implemented by taking large-area subtropical forests in hilly areas in western Fujian province as implementation objects, firstly extracting 14 characteristic variables from laser radar data, simultaneously observing the volume of the forest trunk on the ground, then establishing a mixed effect model of the two under the premise of combining vegetation types, and finally estimating the biomass of the whole research area by combining a biomass relative growth model. The invention provides a method flow from data acquisition to forest biomass acquisition by utilizing an airborne laser radar technology, improves the biomass estimation precision of the laser radar method in large-scale forests, and provides a simplified technical solution for the airborne laser radar in forestry investigation application of large-area subtropical forests.
Compared with the prior art, the invention has the advantages that the field sample area exceeds 19000km 2 The area of (a), is 100-1000 times larger than other typical studies; in addition, the large area of the study area means an increase in vegetation diversity and a challenge for model prediction, so the present invention uses a mixed effect model including vegetation types, as compared to the prior art. The invention is implemented in large area for the first time in subtropical forest regions. Test results show that the trunk volume estimation of the large-area subtropical forest in China by the method obtains higher precision: r 2 ＝0.49，RRMSE＝64.3％。
Drawings
FIG. 1 is a plot of test zones and plots.
Fig. 2 is a scatter plot of the predicted value of the trunk volume based on the mixed effect model versus the measured values in the overlapping area (dotted line is 1 line; solid line is the linear regression line between the observed value and the predicted value).
Detailed Description
The present invention will be described in further detail with reference to examples.
Example 1
A mixed effect model for predicting the trunk volume of a large-area subtropical forest based on a laser radar and a vegetation type comprises the following steps:
1) General description of test area
The research area is located in Fujian province and is the whole Longyan city. The whole province has more mountains and wide breadth. In Longyan, the majority of the land is mountainous and hilly. .
2) Lidar data acquisition and preprocessing
Onboard lidar data were collected on a Cessna208 aircraft using LeicaALS70 sensors (11-12 months in 2013), with a flying height of 3500 meters and a flying speed of 230-280 km/h. The sensor laser emission used a pulse repetition frequency of 79-234kHz with a scan rate of 12-32Hz and a scan angle of 50 degrees. The minimum side cleavage was 25%, and the average dot density was 1.2 dots/m 2 And each laser return point comprises a three-dimensional coordinate value, an intensity value and return type information.
3) Acquisition of ground survey data
TABLE 1 type of vegetation in the study area
140 square plots (size: 25.82 m.times.25.82 m) were set in the Longyan region in 2013. The southwest corner is marked and measured similarly with GNSS (global navigation satellite system) with an error of less than 10 meters. The other corners use a compass and a tape in a clockwise order. The measurement of the side length takes into account the influence of the side slope (side length =25.82m/cos (θ), where θ is the terrain slope). The closing error of the four corners and two sides of the square is less than 1%. In the sample plot investigation process, for trees with the diameter at breast height being more than or equal to 5 cm and bamboos with the diameter at breast height being more than or equal to 2 cm, the tree species, the tree height and the diameter at breast height of a single tree are measured one by one; obtaining an observed trunk volume according to the single-tree survey data and the different-speed growth model; and summarizing the unit trunk volume of each sample land according to the sample land information. The type of vegetation at the level of the plot was established according to the chinese national vegetation classification system (table 1).
4) Feature variable extraction
TABLE 2 LiDAR laser point cloud height feature variable summary
14 feature variables were calculated from the LiDAR point cloud data: 3 statistical variables of height (mean H) mean Standard deviation H std Deviation H skew And kurtosis H kurt ) (ii) a 10 height percentile variables: h ptc10 、H ptc20 、…、H ptc100 (ii) a Mean square value of average height H qm (high points have higher weight). The meanings and calculation formulas of the 14 characteristic variables are shown in Table 2.
5) Statistical modeling
The method comprises the steps of taking trunk volume information obtained through ground actual measurement and summary as a dependent variable, taking a prediction index obtained by a laser radar as an independent variable, and establishing a mixed effect model of a nonlinear parameter regression model based on vegetation types, wherein the mixed effect model comprises a Simple Power Model (SPM) and a multi-power model (MPM). Both have the same model formula:
wherein H i Are the individual lidar indices described in (4) generated from a set of lidar points. N is the number of lidar measurements in the model. For the simple power model, n =1; for the multivariate power model, n is greater than or equal to 2.
To develop a nonlinear multivariate power model to predict trunk volume, we selected the lidar criteria as follows: first, the trunk volume and lidar indices are logarithmically transformed, and then forward stepwise regression is used on a logarithmic scale to select statistically significant variables. The p-value thresholds for the inclusion and exclusion variables are set to 0.2 and 0.1, respectively, intentionally higher than the typical value of 0.05 in the statistical software package to reduce the risk of removing important lidar indicators. When necessary, the model was redesigned by removing any statistically insignificant (p-value > 0.05) variables.
6) Model validation
Model performance was evaluated using 10-fold Cross Validation (CV) with the following steps: (1) randomly dividing the sample plot into 10 parts; (2) The trunk volume observed for any 9 samples was used to calibrate the model to predict the trunk volume for the remaining 1 sample; (3) Repeating the previous steps for 10 times, and sequentially predicting to obtain the remaining 9 stock volumes of the sample prescription. (4) Calculating a determination coefficient (R) for cross-validation based on observed and predicted trunk volume 2 ) And the Relative Root Mean Square Error (RRMSE) is as follows:
wherein, y i Is the volume of trunk material observed by the sample plot,
is the trunk volume predicted by the model, n is the number of squares (n = 140), ->
Is the average trunk volume observed for 140 samples. When using a mixed effect model, if the prediction model does not include the vegetation type present in the calibration model plot, then the mixed effect model without the random term is used for the prediction.
TABLE 3 evaluation of Mixed Effect model accuracy
The accuracy of the mixed effect model was evaluated as shown in table 3. A comparison scatter plot of predicted values of the trunk volume model and actual values of the overlap area plots and 1 line plot are shown in fig. 2.
7) Acquisition of 242 pieces of parsed wood data
TABLE 4 Biomass relative growth model
Note: b is aboveground biomass, kg/plant; v is the volume of the trunk material, m 3 A plant.
A biomass relative growth model was obtained by plot survey and 242 parse trees (table 4).
8) Operation result of method
Test results show that the trunk volume estimation of the large-area subtropical forest in China by the method is carried out in H pct30 The higher precision is obtained: r 2 And (4) the biomass of the forest is estimated according to the biomass relative growth model, wherein the RRMSE is 0.49 percent and 64.3 percent, and the biomass of the forest in the large area subtropics in China is finally estimated according to the biomass relative growth model.
Claims (1)
1. A method for predicting subtropical forest biomass based on a laser radar and a vegetation type is characterized by comprising the following steps:
1) Acquiring laser radar data by using a sensor, and preprocessing the acquired data and extracting characteristic variables;
the method for acquiring the laser radar data by using the sensor specifically comprises the following steps: acquiring laser radar point cloud data in a to-be-detected area of a forest in a subtropical hilly area by using a sensor, and recording complete laser pulse return waveform information by using the sensor;
preprocessing the acquired data, which specifically comprises the following steps: processing the laser radar point cloud data by using laser radar data filtering and forest tool kit Tiffs software to separate ground echoes; generating a 1-meter Digital Terrain Model (DTM) by interpolating ground echoes, and calculating the height of each laser point by subtracting the height of a corresponding DTM unit below each laser point from the Z coordinate of each laser point;
the extraction of the characteristic variables specifically comprises the following steps: the 14 characteristic variables of the height of the laser spot are extracted: the 3 statistical variables include the mean value H mean Standard deviation H std And kurtosis H kurt (ii) a 10 pieces high hundredQuantile variables: h ptc10 、H ptc20 、…、H ptc100 (ii) a Mean square value H of 1 mean height qm Wherein high points have higher weight;
2) Obtaining ground sample plot data and estimating a trunk volume in a designed sample plot in a forest to be measured in a subtropical hilly area, and determining the vegetation type of the sample plot;
specifically, a plurality of square sample plots are arranged in a to-be-detected area range of a forest in a subtropical hilly area, and tree species, tree heights and breast diameters of single trees are detected one by one for trees with the breast diameter being more than or equal to 5 cm and bamboos with the breast diameter being more than or equal to 2 cm in the sample plot investigation process; estimating the trunk volume of the single tree according to the single tree survey data; collecting the trunk volume of each sample land according to the sample land information; the vegetation types comprise evergreen broad-leaved forests, coniferous and broad mixed forests and bamboo forests;
3) Taking the trunk volume of each sample plot obtained by summarizing the sample plot information as a dependent variable, taking the characteristic variable of the height of the laser point as an independent variable, and establishing a nonlinear parameter regression mixed effect model based on the vegetation type;
the formula of the nonlinear parametric regression mixed effect model is as follows:
wherein, when n =1, it is a simple power model; when n is more than or equal to 2, the model is a multi-element power model;
wherein a multivariate power model is developed to predict the trunk volume, H is selected i The steps are as follows:
trunk volume sum H for each plot i Performing logarithmic transformation, then using forward stepwise regression on a logarithmic scale to select statistically significant variables, setting p-value thresholds for inclusion and exclusion variables to 0.2 and 0.1, respectively;
4) Using 10-fold cross-validation CV evaluation and calibrating a mixed effect model of nonlinear parametric regression;
the method comprises the following steps: (1) randomly dividing the sample plot into 10 sample squares; (2) Use of any of 9Calibrating the trunk volume of each sample plot of the sample shares to a mixed effect model of nonlinear parametric regression to predict the trunk volume of the remaining 1 sample plot; (3) Repeating the step (2) for 9 times, and sequentially predicting to obtain the remaining 9 parts of tree trunk volume of the sample prescription; (4) Calculating a determination coefficient R of cross validation based on the trunk volume of each sample plot summarized from the sample plot information and the predicted trunk volume of each sample plot 2 And the relative root mean square error RRMSE is as follows:
where n is the number of the same, n =140,
is the average trunk volume of the trunk volumes of the plots summarized by the 140 plot information;
5) Establishing a biomass relative growth model by using the analytical wood, comprising the following steps of: 1) Measuring the biomass of the analytic wood; 2) Establishing a biomass relative growth model B = aV + B according to the measured biomass and trunk volume of the analytic wood;
and substituting the predicted trunk volume V into the established biomass relative growth model B = aV + B to calculate the biomass B.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810538163.3A CN108959705B (en) | 2018-05-30 | 2018-05-30 | Method for predicting subtropical forest biomass |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810538163.3A CN108959705B (en) | 2018-05-30 | 2018-05-30 | Method for predicting subtropical forest biomass |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108959705A CN108959705A (en) | 2018-12-07 |
| CN108959705B true CN108959705B (en) | 2023-04-07 |
Family
ID=64492604
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810538163.3A Active CN108959705B (en) | 2018-05-30 | 2018-05-30 | Method for predicting subtropical forest biomass |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108959705B (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109710889B (en) * | 2018-12-13 | 2023-02-03 | 北京林业大学 | Sampling method for accurately estimating forest productivity based on tree ring |
| CN110162872B (en) * | 2019-05-17 | 2022-11-29 | 中国科学院城市环境研究所 | Biomass estimation method fusing sample plot data and forest resource clearing data |
| CN110245420A (en) * | 2019-06-14 | 2019-09-17 | 海南省林业科学研究所 | The method of coast protection forest biomass and organic C storage monitoring and metering |
| CN110276757A (en) * | 2019-06-25 | 2019-09-24 | 北京林业大学 | One kind carrying out high canopy density artificial forest region single tree biomass draughtsmanship based on tilted photograph |
| CN111814397B (en) * | 2020-07-03 | 2023-03-24 | 中国科学院城市环境研究所 | Forest biomass estimation method combining source analysis and machine learning |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1502047A (en) * | 2001-02-09 | 2004-06-02 | �����ѧ��ҵ�о���֯ | Lidar system and method |
| CN105913016A (en) * | 2016-04-08 | 2016-08-31 | 南京林业大学 | Strip LiDAR data upscaling-based forest biomass estimating method |
| CN106408608A (en) * | 2016-09-30 | 2017-02-15 | 信阳师范学院 | Method for extracting trunk diameter from ground laser radar point cloud data |
| CN106931883A (en) * | 2017-03-30 | 2017-07-07 | 信阳师范学院 | A kind of stem volume acquisition methods based on laser point cloud data |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP2583252B1 (en) * | 2010-06-16 | 2023-11-01 | Yale University | Forest inventory assessment using remote sensing data |
-
2018
- 2018-05-30 CN CN201810538163.3A patent/CN108959705B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1502047A (en) * | 2001-02-09 | 2004-06-02 | �����ѧ��ҵ�о���֯ | Lidar system and method |
| CN105913016A (en) * | 2016-04-08 | 2016-08-31 | 南京林业大学 | Strip LiDAR data upscaling-based forest biomass estimating method |
| CN106408608A (en) * | 2016-09-30 | 2017-02-15 | 信阳师范学院 | Method for extracting trunk diameter from ground laser radar point cloud data |
| CN106931883A (en) * | 2017-03-30 | 2017-07-07 | 信阳师范学院 | A kind of stem volume acquisition methods based on laser point cloud data |
Non-Patent Citations (3)
| Title |
|---|
| Above ground biomass and tree species richness estimation with airborne lidar in tropical Ghana forests;Gaia Vaglio Laurin et al;《International Journal of Applied Earth Observation and Geoinformation》;20160719;全文 * |
| 亚热带森林参数的机载激光雷达估测;付甜 等;《遥感学报》;20110925;全文 * |
| 基于激光雷达和外推模型的森林地上生物量估算研究;郑朝菊;《中国优秀博硕士学位论文全文数据库(电子期刊) 农业科技辑》;20180215;全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108959705A (en) | 2018-12-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108959705B (en) | Method for predicting subtropical forest biomass | |
| CN108921885B (en) | Method for jointly inverting forest aboveground biomass by integrating three types of data sources | |
| CN104656098B (en) | A kind of method of remote sensing forest biomass inverting | |
| CN109031344B (en) | Method for jointly inverting forest structure parameters by full-waveform laser radar and hyperspectral data | |
| CN104457626B (en) | A kind of plant leaf area assessment of indices method based on laser radar point cloud | |
| Chen et al. | Estimating basal area and stem volume for individual trees from lidar data | |
| CN105913017A (en) | Corresponding period double high resolution remote sensing image-based forest biomass estimation method | |
| CN109212505A (en) | A kind of forest stand characteristics inversion method based on the multispectral high degree of overlapping image of unmanned plane | |
| CN105913016A (en) | Strip LiDAR data upscaling-based forest biomass estimating method | |
| Pennypacker et al. | Seeing the fields and forests: Application of surface-layer theory and flux-tower data to calculating vegetation canopy height | |
| CN104502919A (en) | Method for utilizing airborne laser radar point cloud to extract urban vegetation three-dimensional coverage map | |
| CN108872964A (en) | Ginkgo artificial forest canopy density extracting method based on unmanned plane LiDAR data | |
| CN103954567A (en) | Soil salinity determining method based on continuum removal method | |
| CN110109118B (en) | Forest canopy biomass prediction method | |
| CN107479045B (en) | Method and system for eliminating short vegetation based on full-waveform laser radar point cloud data | |
| CN109948106B (en) | Method for calculating forest stand height accumulated percentage by using laser point cloud | |
| CN104062644A (en) | Method for extracting tree height from laser radar Gaussian echo data | |
| CN108896021A (en) | Method based on aerophotogrammetry data reduction plantation stand structural parameters | |
| Lim et al. | Examining the effects of sampling point densities on laser canopy height and density metrics | |
| CN109212553A (en) | The method that unmanned plane LiDAR and random forest extract ginkgo biological physical characteristic | |
| Ediriweera et al. | LiDAR remote sensing of structural properties of subtropical rainforest and eucalypt forest in complex terrain in North-eastern Australia | |
| CN112924401B (en) | Semi-empirical inversion method for chlorophyll content of vegetation canopy | |
| Popescu | Estimating plot-level forest biophysical parameters using small-footprint airborne lidar measurements | |
| CN116881812A (en) | Method for estimating forest carbon reserves based on multisource remote sensing data and random forest method | |
| CN109917387A (en) | A kind of method for building up of the forest biomass model based on full-polarization SAR data |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |