WO2003069315A1

WO2003069315A1 - Method of estimating biomass of forests and trees by remote sensing high-resolution data

Info

Publication number: WO2003069315A1
Application number: PCT/JP2003/001460
Authority: WO
Inventors: Jiro Suekuni; Makoto Nogami; Katsushi Hatayama
Original assignee: Kansai Environmental Engineering Center Co., Ltd.; The Kansai Electric Power Co., Inc.
Priority date: 2002-02-13
Filing date: 2003-02-13
Publication date: 2003-08-21
Also published as: AU2003211938A1; JP4219819B2; JPWO2003069315A1; AU2003211938B2

Abstract

A method of estimating the biomass of forests and trees by using remote-sensing high-resolution data from high places. The biomass of specific types of trees is estimated with high precision even in the cases of forests and trees where their crowns are not closed due to natural young growth shortly after being planted or young trees, or forests or trees where the crowns of even mature trees are not closed. High-resolution to ultra-high-resolution pictures are shot (11) using high-resolution satellites, air planes and radio helicopters, the pictures are captured into computers (12), portions other than specific types of trees are masked (13), the crown sizes (areas, longer diameters, shorter diameters) of specific types of trees left after the masking are measured (14), the spectral characteristics of the specific types of trees are measured (15), and the biomass of the specific types of trees is calculated from crown sizes (areas, longer diameters, shorter diameters) and spectral characteristics (16). Then, an annual biomass accumulated amount is calculated based on this calculated biomass amounts (17), and the result is printed out (18).

Description

Description: Biomass estimation method for forests and trees by remote sensing high-resolution data technology Field of the Invention The present invention relates to a method for estimating biomass of forests and trees by remote sensing high-resolution data processing. Forests where the crown is not blocked because of the tree, or when the crown is not blocked even if it has matured.Remote sensing suitable for estimating the biomass of the tree. And a method for estimating biomass. Background Technology Various adverse effects due to global warming are regarded as problems on a global scale. On the other hand, it is said that the assimilation of carbon dioxide by forests and trees also has the effect of preventing global warming. Therefore, the need for afforestation projects is being raised to increase biomass, which is the basis for calculating carbon dioxide assimilation by forests and trees, and afforestation projects are being actively conducted. At the same time, biomass is estimated from forests' trees, but the biomass in forests.trees depends on the type of forest and trees, canopy size (area, major axis, minor axis) and their spectral characteristics. Problems to be Solved by the Invention Therefore, remote sensing using satellite photographs and aerial photographs has been implemented to measure the species of forests and trees, the crown size (area, major axis, minor axis) and spectral characteristics. I have. In the forestry field, the Landsat emperor data analysis that has been carried out so far has been performed on a pixel-by-pixel basis, that is, in the case of the RAND SAT / TM, for example, as a unit of 30 m square. This method provides sufficient accuracy for general surveys over a wide area, and shows that high-accuracy analysis can be performed even in natural forests in forests where the crowns are closed and the tree species are limited. However, the usefulness of remote sensing techniques in ecology is well established. On the other hand, not all forests and trees can be surveyed in this way. This is because the crowns of forests and trees may or may not be closed depending on the type of forest and tree. In other words, as shown in Fig. 7 (a), in general, trees grow from the seedling stage, like saplings, young trees, and mature trees, and as shown in Fig. 7 (b). Thus, the crown size (area, major axis, minor axis), crown thickness and LAI (Leaf Area Index) gradually increase, and as shown in Fig. 7 (c), the crown ratio also increases from A → → It grows larger like C → O (= C). In Fig. 7 (c), P is a pixel, J is a canopy, and E is soil and other parts other than the canopy. Therefore, in forests and trees where the canopy is closed, as shown in C and D in Fig. 7 (c), the conventional remote sensing method can be used, as shown in Fig. 7 (c). Between the seedling stage of A and the young tree stage of B, the crown J is small and not closed, so the soil and other non-canopy parts E existing between crown J and crown J, that is, It is difficult to distinguish from the ground, roads, trees of other species, vegetation, etc., and trees of tree types other than the planted specific trees, and the ground, roads, vegetation, etc., become noise. Canopy size (area, major axis, minor axis) and its spectral properties could not be measured. In addition, depending on the species, even for mature trees, the crown is not blocked. For example, in the case of planting eucalyptus trees in Australia, the canopy often does not close even when the tree matures. In particular, in Malian eucalyptus, crops such as wheat are often grown between plantations. In the case of such forests and trees, the pixel-based method that has been used up to now analyzes the value of a mixture of planted trees and the ground, roads, undergrowth and other cultivated plants, etc. It was not possible to measure the crown area and spectral characteristics of specific tree species with high accuracy, and therefore to estimate biomass with high accuracy. Moreover, since it is almost impossible to separate the information once synthesized, it was necessary to devise a method from the data acquisition stage. Therefore, the present invention can be applied to a case where the crown is not closed due to a young tree or a young tree that has a relatively short age after planting, or even a mature tree that does not have a closed crown. Canopy size (area, major axis, minor axis) and its spectral characteristics can be measured with high accuracy, and the biomass of forests and trees can be highly accurately estimated from these results using remote sensing high-resolution data. The purpose is to provide an estimation method. DISCLOSURE OF THE INVENTION The forest-tree biomass estimation method according to claim 1 of the present invention by remote sensing high-resolution data processing includes: The method is characterized in that the biomass of a specific tree is estimated from the size and spectral characteristics of the crown occupying the predetermined area by taking an image and masking a portion other than the crown portion of the specific tree within the predetermined area based on the photograph. . Here, the term “high-resolution data” is obtained by analyzing high-resolution photos by high-resolution satellites or by analyzing ultra-high-resolution photos by airplanes or wireless Helicopters. Data. The term “canopy size” includes the area, major axis, and minor axis of the canopy part, and is hereinafter referred to as canopy size (area, major axis, minor axis). However, it is not always necessary to include all of the area, the major axis and the minor axis, and only the area may be included. By measuring the major axis and the minor axis, it is possible to estimate the biomass with higher accuracy. Furthermore, the above-mentioned “masking other than the crown portion of the specific tree within the predetermined area” means that the portion other than the specific tree planted within the predetermined area (in the image in the photograph), that is, This refers to erasing trees, ground and roads, vegetation and agricultural products, etc., other than the specified tree species as image data, and cutting out the crown of the specified tree species along the edge of the crown on a photo taken on a computer. The other part says to erase the data on your computer. The work of erasing the part other than the canopy part at the time of the combi-up is to cut out the crown of a specific tree species in the picture image on the monitor while looking at the photograph actually taken by the worker, and erase the other parts Alternatively, the spectral characteristics or the like corresponding to a specific tree species may be grasped in advance, and the other spectral characteristics may be automatically deleted by the combination process. In the former case, the masking operation requires a relatively long time, but since it is possible to cut out the tree while checking the specific tree species in the actually photographed image, there is an advantage that the crown of the specific tree species can be cut out with high accuracy. On the other hand, in the latter case, although the accuracy of cutting out a specific tree species is not always high, there is an advantage that the cutting operation can be performed in a short time. Therefore, when adopting this estimation method, the former worker performs high-precision cutting, and after the cutting data is accumulated to some extent and the data of the specific tree species is obtained, the accumulated data is If automatic processing is performed by a computer based on this, high-precision extraction can be performed in a relatively short time. According to the biomass estimation method for forests and trees as described above, if the canopy is not obstructed, that is, within a specific area in the field, in other words, in the image on the photo, In addition to the above, the canopy size (area, long diameter, short diameter) of trees of a specific species, even if there are different types of trees other than the specific species, as well as the ground, roads, vegetation and crops Diameter can be measured with high accuracy, and together with the measurement of the spectral characteristics of the canopy, the biomass of a particular tree species can be estimated with high accuracy. Therefore, for example, if an investor invests in afforestation and entrusts the management of the forests and trees, the manager will decide whether the canopy size should be set annually according to the age of the plantation, or every specified age. (Area, major axis, minor axis) and / or changes in the spectral characteristics of the canopy, estimating the extent of biomass increase and annual biomass accumulation in the area, and reporting the results to investors. Can be. Investors can use the reported data to check the effects of planting their own trees on global warming, as well as forecasting the timing of logging and the expected selling price of timber at the time of logging. Also, in tree species areas such as Yururi forest where the crown does not close even after maturity, similarly to the above, high-precision measurement of canopy size (area, major axis, minor axis) and / or spectral characteristics It is possible to estimate highly accurate biomass and annual biomass accumulation by measuring the amount of biomass. The forest-tree biomass estimation method according to claim 2 of the present invention by remote sensing high-resolution data is characterized in that the photograph is a photograph taken in a specific wavelength band. According to such a forest / tree biomass estimation method, a photograph that emphasizes trees of a tree type that matches or approximates the spectral characteristics based on the spectral characteristics of a specific tree species This makes it possible to blur or easily mask other spectral characteristics, leaving only the crown of a specific tree species, which not only improves masking accuracy but also speeds up masking work. It can be implemented at low cost. The forest-tree biomass estimation method using remote sensing high-resolution data according to claim 3 of the present invention is characterized in that the photograph includes a plurality of tree crowns in one pixel. . According to such a method for estimating forest and tree biomass, it is possible to measure the spectral characteristics even in a state in which there is a gap between the crowns and the crowns are not closed, and as a result, And the use of appropriate estimation formulas, it is possible to estimate biomass and annual biomass accumulation with high accuracy. The forest-tree biomass estimation method according to the remote sensing high-resolution data set forth in claim 4 of the present invention, wherein the photograph is an ultra-high-resolution photograph, and one canopy extends over a plurality of pixels. It is characterized by the following. According to such a method of estimating forest and tree biomass, as one canopy extends over a plurality of pixels, the more pixels the canopy blocks, the larger the canopy size (area , Long diameter, short diameter) can be improved, and biomass and annual biomass accumulation can be estimated with high accuracy. However, as the number of pixels increases, the number of images increases, and the number of image processing steps increases accordingly. Therefore, if the number of pixels is set appropriately according to the crown size (area, major axis, minor axis) Good. BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a schematic flow chart in a method for estimating biomass of a forest / tree using remote sensing high-resolution data according to an embodiment of the present invention. FIG. 2 is a schematic explanatory diagram of each step in the forest / tree biomass estimation method using remote sensing high-resolution data according to the embodiment of the present invention.

Figures 3 (a) to 3 (c) illustrate the relationship between the crown and the pixel size.Figure 3 (a) shows the pixel crown diagram when the crown ratio is 100%. ) Is a pixel crown diagram when the crown ratio is less than 100%, and FIG. 3 (c) is a pixel crown diagram by a high-resolution photograph when the crown ratio is less than 100%.

Figures 4 (a) to 4 (d) show the relationship between various tree factors and biomass based on the local ground-based tree survey.Figure 4 (a) shows the relationship between tree height and biomass. Figure, Figure 4 (b) is a characteristic diagram showing the relationship between canopy thickness and biomass, Figure 4 (c) is a characteristic diagram showing the relationship between average crown ratio and biomass, and Figure 4 (d) is the canopy area and biomass. FIG. 4 is a characteristic diagram showing a relationship between

Figures 5 (a) to 5 (d) show the relationship between various tree factors and biomass based on the aerial image of the site. Figure, Figure 5 (b) is a characteristic diagram showing the relationship between canopy area and biomass, and Figure 5 (c) is a distribution diagram of the difference between the biomass estimation result using the average canopy diameter and the measured value from the field survey. 5 (d) is the distribution map of the difference between the biomass estimation result using the canopy area and the actual measurement value from the field survey.

FIG. 6 is a correlation estimation characteristic diagram of the evaluation index IR * G / R and biomass.

Figures 7 (a) to 7 (c) show the growth stages of trees after planting, Figure 7 (a) is a side view of trees showing the growth stages of trees after planting, and Figure 7 (b) is Tree crown diagram at the growth stage of trees after planting, and Fig. 7 (c) is a pixel crown diagram showing the change in crown ratio with the growth stage of trees after planting. BEST MODE FOR CARRYING OUT THE INVENTION Hereinafter, an embodiment of a method for estimating biomass of forests and trees by remote sensing high resolution data of the present invention will be described with reference to the drawings. FIG. 1 shows a remote sensing high-resolution image according to the first embodiment of the present invention. FIG. 1 is an overall schematic flowchart according to a forest / tree biomass estimation method 10 according to the present invention. In Fig. 1, first, photographs of the forests and trees are taken from a high place using a Landsat satellite or a high-resolution satellite (IKONOS), or an aerial vehicle or a radio helicopter. It can take pictures with a resolution of 30 m on each side, and a high-resolution satellite (IKONOS) can take high-resolution pictures with a side of about 4 m. In the case of one, it is possible to take an ultra-high resolution photograph with a side of several cn! ~ 20 cm.In this photographing, use an observation device such as a scanner-type optical sensor or expose to a specific wavelength band When filming is used, trees other than the specific tree species involved in afforestation, as well as the ground, roads, vegetation, and agricultural crops can be removed by combi-evening processing based on their spectral characteristics. In this process, it is easy to cut the crown of a specific tree species, and avoid photographing in the morning or evening when there are many long wavelengths, for example, between 10 am and 2 pm By doing so, the tree image according to the tree species becomes clear under a light beam with a small wavelength deviation, and the subsequent cutting operation can be performed easily and with high accuracy. There is no shadow on the tree canopy due to the oblique incidence of the sun or the sunset, which has the advantage that the subsequent cutting work can be performed with high precision. 1 2) Projected on a monitor At this time, if the photographed image is a digital photograph, it can be directly imported to a computer, or if it is a film analog photograph, a scanner Then, digitize the image using a digital camera and then import it into the combi.Next, based on the high-altitude photo projected on the monitor, the part other than the specific tree species, that is, Mask trees, ground, roads, vegetation, agricultural crops, etc. (13) As described above, this cutting work is performed when this biomass estimation method is adopted or when sampling is performed. When it is not long after use, it is desirable to carry out the inspection while looking at pictures actually taken by experienced workers. After a certain amount of data has been accumulated, it is desirable to perform automatic processing by combination based on the accumulated data. Next, the canopy size (area, major axis, minor axis) of a particular tree species obtained by masking is measured (14). In measuring the canopy size, only the area of the canopy may be measured, but by measuring the major and minor diameters of the canopy, more accurate analysis can be performed. Next, the spectral characteristics of the canopy of the specific tree species cut out as described above are measured (15). The measurement of the spectral characteristics is performed, for example, for each of R, G, B, and IR. Based on this spectral characteristic, a well-known vegetation index such as NVDI (Normalized Difference Vegetation Index) is calculated. This NVD I is one of the indexes of vitality in forest trees, called the “Normalized Difference Vegetation Index”. The reference data and remote sensing data are used in the infrared (R) and near infrared (NIR) regions. The initial vegetation index RVI (= NIR / R), which is the ratio between NIR and R, is normalized based on the spectral characteristics of the two bands.NDVI = (NIR-R) / (NIR + R). In addition to the above-mentioned NDVI, the evaluation index of the vitality of a tree is PVI (Perpendicular Vegetation Index) = (NiR — AxR—?) /, 1 + hi ² ) and ² can be used. Where and /? Are the slope and intercept of the soil line. It is calculated by MSAVI (Modified Soil-Adjusted Vegetation Index) = (1 + L) X (NI RR) / (NIR + R + L), which is an index based on SAV I with reduced soil background. The vegetation index can be used. Where L = l-2 a xND VIX (NIR— «xR). In addition, GEMI (Global Environment Monitoring Index) = rj, which is an index that reduces the effects of soil background and atmospheric costs, is calculated as: (R-0.125) / (1-R). The calculated vegetation index can be used. Here, V = [2 (NIR ² −R ² ) +1.5 NIR + 0.5 R]. (NI + R + 0.5). Next, the biomass of the specific tree species is calculated from the above measurement results of the crown size (area, major axis, minor axis) and spectral characteristics (vegetation index) (16). The term “biomass” refers to the dry weight of stems, branches and leaves. The larger the canopy size (area, major axis, minor axis), the larger the size and the more mature trees. Next, based on the above biomass, calculate the annual biomass accumulation of the specific tree species (17). This is the difference between biomass measured in different years and divided by both periods. Next, the above calculation result is printed (18). Based on the results of this printout, the biomass and annual biomass accumulation of the forest's trees will be grasped and used for future afforestation management, as a reference for logging plans, and provided to investors. FIG. 2 is a schematic diagram of each step in the method for estimating the biomass of forests and trees using high-resolution remote sensing data according to the present invention. That is, FIG. 2A shows a situation when a high-resolution color infrared photograph (for example, a scale of 1 / 7,000) of the tree 21 is taken by the aircraft 20 or the like. (B) shows the capture of a photograph into a computer. In the example shown, an analog photograph 22 taken by film is read by a scanner 23 and digitized (for example, 1,200 dpi). Shows the case of taking in (C) is a pre-analysis process such as geometric correction that corrects image distortion caused by the inclination of the attitude of the aircraft during observation. is there. (D) is the extraction of the crown information. For example, by blackening the part 26 other than the crown 25 of the specific tree on a computer, only the crown 25 of the specific tree remains as an image. Based on the crown information, the size (area, major axis, minor axis) of the crown and its spectral characteristics are measured. (E) is a model formula creation process based on the canopy information obtained as described above and on-site data. The horizontal axis is the vegetation index, and the vertical axis is the biomass. When these processes are completed, the analysis results are output as shown in (f), and the biomass per area and the annual biomass accumulation can be found. Figure 3 shows the difference in crown ratio depending on the size of the crown.Figure 3 (a) shows the state of a mature forest in which the crown 25 is closed in the pixel 30. However, its spectral characteristics can be measured. Fig. 3 (b) shows the number of pixels in a tree 3

0 indicates that the crown 25 is not closed and includes the ground and vegetation 26. In low-resolution data, the size (area, major axis, minor axis) of the crown and its Although it was impossible to measure spectral characteristics, use of ultra-high resolution data according to the present invention

{Fig. 3 (c)} makes it possible to measure the size (area, major axis, minor axis) of the crown and its spectral characteristics with high accuracy without being affected by the ground or vegetation. Example The accuracy of biomass estimation by the above method of estimating the biomass of forests and trees by remote sensing high-resolution data was verified by actually conducting on-site tree surveys and analysis based on aerial images at the site. This will be described below. Case 1 Analysis at 2.5 cm resolution [Ultra high resolution data]

1. Survey Area

Mangrove forest at the mouth of the Urauchi River in Iriomote Island, Oki-Ref prefecture

2. Survey method 2.1 Field survey

(1) Establishment of a rectangular area (50mx50m)

(2) Tree survey (DBH = breast height diameter, tree height, major and minor diameters of crown, crown thickness)

(3) Surveying (obtain location information within each tree square)

(4) Aerial photography (shooting a color infrared film using a small wireless helicopter)

2.2 Data analysis

(1) Aggregation of ground survey results

(2) Image processing and analysis of aerial photographs

(3) Verification of biomass estimation accuracy Creation of biomass estimation formula

A biomass estimation formula was created from the biomass measurement values in the field survey and tree crown information based on aerial images. Here, the “biomass measurement value in the field survey” refers to the value obtained by measuring the DBH (breast height diameter) of all target trees in the field and calculating the biomass individually from this value using the relative growth formula. .

Number of tree samples n: 72

Total biomass in trees for creating the estimation formula: 799.6 kg Figures 4 (a) to (d) show the relationship between biomass and the results of each tree survey by field survey. Fig. 4 (a) is a characteristic diagram showing the relationship between tree height and biomass. Figure 4 (b) is a characteristic diagram showing the relationship between canopy thickness and biomass. Fig. 4 (c) is a characteristic diagram showing the relationship between the average crown diameter and biomass.The average crown diameter has a higher correlation with biomass than the tree height in Fig. 4 (a) (the crown thickness in Fig. 4 (b). You can see that. Fig. 4 (d) is a characteristic diagram showing the relationship between canopy area and biomass, and it can be seen that there is also a high correlation. Analysis results based on aerial images

Figures 5 (a) and 5 (b) show the analysis results based on aerial images in the same rectangular area, focusing on the average crown diameter and crown area, which were highly correlated in each tree survey. Figure 5 (a) FIG. 4 is a characteristic diagram showing the relationship between the average crown diameter and biomass, and is in good agreement with the characteristic diagram of the average crown diameter and biomass in FIG. Fig. 5 (b) is a characteristic diagram showing the relationship between canopy area and biomass, and also agrees well with the characteristic diagram of canopy area and biomass shown in Fig. 4 (d). Therefore, an equation for calculating biomass was created from the average canopy diameter and canopy area, which have a high correlation with biomass.

> Verify accuracy of estimation formula

Next, the accuracy of the biomass estimation formula was verified by comparing the biomass measurement value obtained from the field survey with the biomass obtained by the above estimation formula. The verification was performed using trees different from the trees used in the estimation formula creation.

Number of tree samples n: 72

Total biomass in the test trees: 1, 06 1.9 kg The distribution of the difference between the biomass estimation result using the average canopy diameter and the actual measurement value from the field survey is shown in Figure 5 (c). It is concentrated on small parts and shows high correlation. In addition, the distribution of the difference between the biomass estimation result using the canopy area and the actual measurement value from the field survey is concentrated on the part with a small error, as shown in Fig. 5 (d). ing. As described above, the total biomass in the test tree was 1,061.9 kg, whereas the biomass obtained from the estimation formula using the average crown diameter was 1,036.8 kg. The estimation accuracy based on the average crown diameter is 97.6%. The biomass obtained from the estimation formula using the canopy area is 1,055.3 kg, and the estimation accuracy based on the canopy area is 94.7%. Thus, it was found that both the average crown diameter and the crown area have high estimation accuracy. Therefore, the method for estimating the biomass of forests and trees using the remote sensing high-resolution data of the present invention is not sufficient. It is practical and can be expected to have a significant effect on improving the global environment by improving the accuracy of biomass estimation at the growth stage of forests and trees in the future. Case 2 Analysis at a resolution of 4m [When the crown is not cut out]

1. Study Area

Rizofora abikiyura forest in Thailand (rat)

2. Survey method

2.1 Field survey

(1) 7 squares (20mx20m) (1 year, 3 years, 5 years, 7 years, 9 years, 11 years, 13 years)

(2) Tree survey (DBH = breast height diameter, tree height, major and minor diameters of the crown, crown thickness)

(3) Aerial photography (shooting a color infrared film using a large wireless helicopter)

() Measurement of spectral characteristics

2.2 Data analysis

(1) Aggregate ground survey results ''

(2) Image processing and analysis of aerial photographs

(3) Verification of biomass estimation accuracy In regions where the dry and rainy seasons are clear, such as in Southeast Asia, the seasonal and spectral characteristics (NDVI) were used to check for differences in biomass estimation accuracy between the dry and rainy seasons. Investigating the relationship between the two, the results shown in Table 1 were obtained.

Table 1 Relationship between seasons and spectral characteristics (NDVI)

* Values in the table indicate the correlation coefficient (r ² ) between LAI (leaf area index) and NDVI

R.a. Rizov; ¾ | -Rahzophora apiculata

R.m. Rhizophora mucronata

C on Ceriobus tagal (Ceriops tagal)

B.g. Bruguiera gymnor niza

In Table 1, Trat is the northern part of the Gulf of Thailand, Chumphon is the central part, and Khanom is the southern part. From the results in Table 1, it was found that the relationship between NDVI and LAI had a higher correlation in the rainy season. In addition, when the relationship between mangrove age and biomass was investigated, the results shown in Table 2 were obtained.

3¾ For L 弒 (Rule 26) Table 2 Relationship between mangrove age and biomass

From the results in Table 2, it can be seen that the biomass increases as the mangrove grows older and the trees grow. In addition, when the correlation coefficient with biomass in various evaluation indices was investigated, the results shown in Table 3 were obtained.

Replacement paper (Rule 26) Table 3 Relationship between vegetation index and biomass

From Table 3, NDVI (IR.R = 0.818), GR (= 0.892), IR * G (= 0.889), IR (= 0.951), IG / R (= 0.955) ) Was found to have a high correlation, and in particular, IR * G / R (= 0.955) was found to have the highest correlation. FIG. 6 is a characteristic diagram of an equation for estimating the relationship between the evaluation index IR * G / R (r ² = 0.955) showing the highest correlation value in Table 3 and biomass. For mangrove forests 1 to 13 years old, aerial images were obtained for each age, and three new simulation images were created based on those images. Images composed of forests of all ages from 1 to 13 years of age are classified as A class, images composed of young forests of 1 to 5 years of age are classified into B class, and older forests of 7 to 13 years of age. The image composed of was designated as C class. The forests of each age that make up each class are included in the image at the same rate. Table 4 shows the results of estimating biomass for these classes A to C,

Table 4 Estimation accuracy of biomass in mangrove plantations

(Analysis result at 4m resolution, without masking)

From Table 4, the accuracy was 53.1% in class B, which consisted only of young forests 1 to 5 years old, and the lowest value was shown in three classes. Conversely, class C, which consists of only old forests aged 7 to 13 years, shows the highest value of 87.6%, and class A, which includes all ages, is 1 to 13 years old. Shows the accuracy between the two. As can be seen from this example, the accuracy of estimation is low when analyzing a young forest with a clear canopy at a resolution of 4 m. On the other hand, even at a resolution of 4 m, it was shown that if the canopy is closed, it is possible to estimate with high accuracy by using an appropriate estimation formula. In other words, in order to perform high-precision estimation without masking, it is important that the canopy is closed first. At the same time, it is necessary to use an appropriate estimation formula. Effects of the Invention As described above, the present invention captures a photograph of a forest or a tree from a high place, and masks a portion other than a crown portion of a specific tree within a predetermined area based on the photograph to obtain the above-described image. The feature is to estimate the pyomas of a specific tree from the size (area, major axis, minor axis) and spectral characteristics of the canopy occupying a given area, so that the crown is not blocked shortly after planting. Or even mature trees, the crown will not be blocked

Replacement form (Rule 26) Even in the case of tree plantations, masking noise information between the crowns of the trees enables the crown size (area, major axis, minor axis) of the planted specific tree and its spectral characteristics to be measured with high accuracy. As a result, biomass can be estimated with high accuracy, and the annual biomass accumulation can be calculated with high accuracy.

Claims

The scope of the claims

1. Photographs of forests and trees are taken from high places, and based on the photographs, masking is performed on areas other than the crown of the specific tree within a predetermined area, and the specific tree is determined from the crown size and spectral characteristics occupying the predetermined area. A method for estimating forest and tree biomass using remote sensing high-resolution data, characterized by estimating the biomass of forests.

2. The method for estimating biomass of a forest or a tree using high-resolution remote sensing data according to claim 1, wherein the photograph is an image photographed in a specific wavelength band.

3. The biomass estimation method according to claim 1 or 2, wherein the photograph includes a plurality of tree crowns in one pixel.

4. The photograph according to claim 1 or 2, wherein the photograph is an ultra-high-resolution photograph, and one canopy extends over a plurality of pixels. Biomass estimation method.