CN117218531B - Sea-land ecological staggered zone mangrove plant overground carbon reserve estimation method - Google Patents
Sea-land ecological staggered zone mangrove plant overground carbon reserve estimation method Download PDFInfo
- Publication number
- CN117218531B CN117218531B CN202311156020.3A CN202311156020A CN117218531B CN 117218531 B CN117218531 B CN 117218531B CN 202311156020 A CN202311156020 A CN 202311156020A CN 117218531 B CN117218531 B CN 117218531B
- Authority
- CN
- China
- Prior art keywords
- mangrove
- area
- data
- research area
- satellite image
- 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
- 240000002044 Rhizophora apiculata Species 0.000 title claims abstract description 251
- OKTJSMMVPCPJKN-UHFFFAOYSA-N Carbon Chemical compound [C] OKTJSMMVPCPJKN-UHFFFAOYSA-N 0.000 title claims abstract description 70
- 229910052799 carbon Inorganic materials 0.000 title claims abstract description 70
- 238000000034 method Methods 0.000 title claims abstract description 56
- 238000011160 research Methods 0.000 claims abstract description 183
- 238000011835 investigation Methods 0.000 claims abstract description 101
- 239000002028 Biomass Substances 0.000 claims abstract description 73
- 238000009826 distribution Methods 0.000 claims abstract description 61
- 241000196324 Embryophyta Species 0.000 claims abstract description 33
- 150000003839 salts Chemical class 0.000 claims abstract description 31
- 238000012216 screening Methods 0.000 claims abstract description 20
- 238000007637 random forest analysis Methods 0.000 claims abstract description 17
- 238000010801 machine learning Methods 0.000 claims abstract description 15
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 14
- 230000003595 spectral effect Effects 0.000 claims description 33
- 238000012549 training Methods 0.000 claims description 32
- 238000003384 imaging method Methods 0.000 claims description 31
- 238000012937 correction Methods 0.000 claims description 24
- 238000001228 spectrum Methods 0.000 claims description 24
- 238000007781 pre-processing Methods 0.000 claims description 23
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims description 23
- 238000004364 calculation method Methods 0.000 claims description 19
- 238000002310 reflectometry Methods 0.000 claims description 18
- 238000006243 chemical reaction Methods 0.000 claims description 16
- 230000012010 growth Effects 0.000 claims description 10
- 210000000481 breast Anatomy 0.000 claims description 9
- 238000011156 evaluation Methods 0.000 claims description 9
- 238000013145 classification model Methods 0.000 claims description 8
- 230000002159 abnormal effect Effects 0.000 claims description 6
- 230000001419 dependent effect Effects 0.000 claims description 6
- 238000010606 normalization Methods 0.000 claims description 6
- 238000003066 decision tree Methods 0.000 claims description 5
- 230000005855 radiation Effects 0.000 claims description 5
- 230000005059 dormancy Effects 0.000 claims description 4
- 241000894007 species Species 0.000 claims description 4
- 238000013507 mapping Methods 0.000 abstract description 2
- 238000010586 diagram Methods 0.000 description 9
- 241000120622 Rhizophoraceae Species 0.000 description 6
- 238000005520 cutting process Methods 0.000 description 5
- 238000012544 monitoring process Methods 0.000 description 4
- 230000000007 visual effect Effects 0.000 description 4
- 238000004519 manufacturing process Methods 0.000 description 3
- 238000005259 measurement Methods 0.000 description 3
- 235000013399 edible fruits Nutrition 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 238000009533 lab test Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 238000005070 sampling Methods 0.000 description 2
- 238000006467 substitution reaction Methods 0.000 description 2
- 240000000249 Morus alba Species 0.000 description 1
- 235000008708 Morus alba Nutrition 0.000 description 1
- 240000003793 Rhizophora mangle Species 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 210000000038 chest Anatomy 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000034373 developmental growth involved in morphogenesis Effects 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000013467 fragmentation Methods 0.000 description 1
- 238000006062 fragmentation reaction Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 239000013049 sediment Substances 0.000 description 1
- 239000002689 soil Substances 0.000 description 1
- 238000013179 statistical model Methods 0.000 description 1
Landscapes
- Image Processing (AREA)
Abstract
The invention discloses a method for estimating the overground carbon reserves of a mangrove plant facing sea-land ecological staggered zone, which comprises the following steps: acquiring investigation data and multi-temporal satellite image data of a research area, and screening low-tide satellite images through an inter-tide area index and an inter-tide area proportion; carrying out mangrove forest distribution remote sensing identification by using a random forest machine learning algorithm; constructing a mangrove land biomass inversion regression model of a research area; estimating the overground biomass carbon reserves of the research area based on the overground biomass inversion regression model of the mangrove forest of the research area to obtain the overground carbon reserves of the mangrove forest plants of the research area. According to the invention, the automatic and accurate identification of mangrove and coastal salt biogas is realized through multi-time-phase multispectral satellite remote sensing images and field investigation data, and the remote sensing inversion automation degree of the carbon reserves on the mangrove plants is improved. The method for estimating the on-ground carbon reserves of the mangrove plants oriented to the sea-land ecological staggered zone can be widely applied to the technical field of forestry mapping remote sensing.
Description
Technical Field
The invention relates to the technical field of forestry mapping remote sensing, in particular to a method for estimating the overground carbon reserves of mangrove plants in sea-land ecological staggered zones.
Background
Mangroves are generally distributed in tropical and subtropical coastal zones, and form typical mangrove-coastal salt biogas ecological interlaces with coastal salt biogas, and sea-land ecological interlaces are considered as the most difficult areas to monitor by remote sensing due to spatial heterogeneity, dynamic property and transitivity. Firstly, land coverage types of sea-land ecological staggered belts are various, the land coverage types have high spatial heterogeneity and species diversity, and the phenomena of remote sensing of homonymous foreign matters and homonymous foreign matters are common; secondly, the environmental gradient of the edge is large, different earth surface coverage types are often distributed in a narrow strip shape or a broken plaque shape, and the problem of mixed pixels of medium-low spatial resolution vegetation remote sensing is easily caused; and thirdly, the tide fluctuates, the water level fluctuation enables the dynamic change of the environment of the sea-land ecological staggered zone to be obvious, the imaging range and the shape of the sea-land ecological staggered zone on an optical image are directly affected, and the spectral characteristics of vegetation are changed. Therefore, the complex and dynamic growth environment makes the traditional single-phase and low-spatial resolution remote sensing monitoring mode difficult to meet the requirements of mangrove forest high-precision identification and carbon reserve remote sensing monitoring of land-sea ecological staggered zones. Most of the current researches are directly based on visual interpretation of single-time-phase remote sensing images to extract the mangrove distribution range of the research area as the range of the remote sensing inversion of the mangrove carbon reserves, but the visual interpretation depends on the experience of interpretation staff and is influenced by subjective factors. Mangrove forest on land and sea ecological staggered zone overlaps with partial ecological niche of coastal salt marsh, and spectral characteristics are similar, and traditional classification method is easy to confuse mangrove forest and coastal salt marsh, and tidal fluctuation environment brings great uncertainty for range identification and carbon reserve estimation of mangrove forest.
Disclosure of Invention
In order to solve the technical problems, the invention aims to provide a method for estimating the on-ground carbon reserves of mangrove plants in a sea-land ecological staggered zone, which is used for realizing high-precision automatic identification of mangroves and coastal salt and biogas by constructing a low-tide satellite image meeting requirements through screening an inter-tide area index and an inter-tide area proportion and combining plant weather information and field investigation data in a multi-time-phase multispectral satellite remote sensing image, so that the on-ground carbon reserves of mangrove vegetation are remotely sensed and inverted to be automatic.
The first technical scheme adopted by the invention is as follows: a method for estimating the overground carbon reserves of a mangrove plant facing sea-land ecological staggered zone comprises the following steps:
Acquiring investigation data and multi-temporal satellite image data of a research area, performing data preprocessing on the investigation data, and screening and data preprocessing the multi-temporal satellite image data to obtain preprocessed investigation data and preprocessed multi-temporal satellite image data;
Preparing a classification training sample based on the preprocessed investigation data, taking the preprocessed multi-temporal satellite image data as classification basic data, and carrying out remote sensing recognition on the mangrove distribution range through a random forest machine learning algorithm to obtain the mangrove distribution range of a research area;
Acquiring corresponding medium-high spatial resolution multispectral satellite images according to the distribution range of the mangrove in the research area, and constructing an inversion regression model of the biomass on the mangrove ground in the research area by combining the preprocessed investigation data;
And acquiring a carbon conversion coefficient of the biomass on the mangrove plant ground, and estimating the carbon reserve of the mangrove plant ground in the research area by combining the middle-high spatial resolution multispectral satellite image of the distribution range of the mangrove in the research area and the inversion regression model of the biomass on the mangrove plant ground in the research area.
Further, the step of acquiring survey data and multi-time satellite image data of the research area, performing data preprocessing on the survey data, and screening and data preprocessing on the multi-time satellite image data to obtain preprocessed survey data and preprocessed multi-time satellite image data specifically includes:
Setting a mangrove study area range and acquiring investigation data of a corresponding study area;
performing data preprocessing on investigation data of a research area to obtain preprocessed investigation data;
shooting the mangrove research area range through a satellite multispectral imager to obtain multi-temporal satellite image data;
And screening and preprocessing the multi-temporal satellite image data to obtain preprocessed multi-temporal satellite image data.
Further, the step of performing data preprocessing on the survey data of the investigation region to obtain preprocessed survey data specifically includes:
Shooting a mangrove research area range through an unmanned aerial vehicle to acquire first-class investigation data of the research area;
The first type of investigation data of the investigation region is photos with different ground object types and shooting geographic position information, wherein the different ground object types comprise ground object types and vegetation types;
Setting a mangrove ecological survey sample position based on a research area range, measuring mangrove plants in the mangrove ecological survey sample position, and obtaining plant height, breast diameter and base diameter information of the mangrove plants in the mangrove ecological survey sample;
Estimating the overground biomass of mangrove plants of different species of mangrove ecological investigation sample sides through a mangrove abnormal growth equation according to the plant height, breast diameter and base diameter information of the mangrove plants, obtaining the overground biomass of the mangrove plants of the mangrove ecological investigation sample sides, and obtaining second-type investigation data of a study area;
Integrating the first type of investigation data of the investigation region and the second type of investigation data of the investigation region to obtain preprocessed investigation data.
Further, the step of screening and data preprocessing the multi-temporal satellite image data to obtain preprocessed multi-temporal satellite image data specifically includes:
selecting satellite image data corresponding to the cloud cover less than 20% by considering cloud cover interference, and selecting satellite images imaged in the growing period and the dormant period of the coastal salt biogas plant in a research area to obtain preliminarily screened multi-temporal satellite image data;
sequentially performing geometric correction, radiometric calibration and atmospheric correction on the preliminarily screened multi-temporal satellite image data to obtain corrected multi-temporal satellite image data;
The geometric correction is used for correcting geometric distortion caused in the imaging process of the sensor;
The radiometric calibration is used for converting an electronic signal recorded by a remote sensing image sensor in the data input unit into the radiance or reflectivity of the ground surface feature;
the atmosphere correction is used for eliminating the influence of atmosphere refraction and scattering and converting the radiation brightness or reflectivity into the actual reflectivity of the ground object;
based on the mangrove study area range and combined with photos with geographical position information in the investigation data of the study area, making a study area range vector;
Clipping the corrected multi-temporal satellite image data through a research area range vector to obtain clipped multi-temporal satellite image data;
And (3) taking the influence of tidal fluctuation on the remote sensing imaging of the ground object spectrum and the distribution range thereof into consideration, and screening the cut multi-temporal satellite image data to obtain the preprocessed multi-temporal satellite image data.
Further, the step of screening the cut multi-temporal satellite image data to obtain preprocessed multi-temporal satellite image data, which specifically comprises the steps of:
taking the influence of tidal fluctuation on the remote sensing imaging of the ground object spectrum and the distribution range thereof into consideration, and acquiring the area of the investigation region when the lowest tide is greater than 0 and the area of the investigation region when the highest tide is less than 0 and the normalized water body index;
constructing an intertidal zone area index and an intertidal zone area ratio according to the area of the study area when the lowest tide is the time and the normalized vegetation index is larger than 0, the area of the study area when the highest tide is the time and the normalized water body index is smaller than 0 and the area of the study area;
And screening the preliminarily screened multi-temporal satellite image data according to the area indexes of the intertidal zones and the area ratio of the intertidal zones, and selecting an image with the area ratio of the intertidal zones being more than 10% and the largest area index of the intertidal zones as the preprocessed multi-temporal satellite image data based on the growing period and the dormant period of the coastal salt biogas plant in the research area.
Further, the expression of the inter-tidal zone area index and the inter-tidal zone area ratio is specifically as follows:
；
；
In the above-mentioned method, the step of, Represents the area index of the intertidal zone,Represents the area proportion of the intertidal zone,Representing normalized vegetation index in a research area during satellite image imagingThe area of the region is such that,Representing the area of the investigation region,Representing normalized water index in research area during satellite image imagingThe area of the region is such that,Image normalization vegetation index representing lowest tide imaging of research areaThe area of the region is such that,Image normalization water index for representing imaging of highest tide in research areaArea of the region.
Further, the step of preparing a classification training sample based on the preprocessed investigation data, using the preprocessed multi-temporal satellite image data as classification basic data, and performing remote sensing recognition on the mangrove distribution range through a random forest machine learning algorithm to obtain the mangrove distribution range of the research area, specifically comprises the following steps:
Calculating spectral features and texture features of the multi-phase satellite image data based on the preprocessed multi-phase satellite image data;
based on first-class investigation data of a research area, making classification training samples of different ground objects;
Setting random forest machine learning algorithm parameters, setting the number of decision trees to be generated as 100, setting the number of split nodes as the square root of all features, randomly selecting 70% of sample data in classified training samples as training samples, and using the remaining 30% of sample data as precision evaluation samples;
Spectral features and texture features of multi-temporal satellite image data corresponding to training samples are used as input data and are input into a random forest machine learning algorithm to construct a ground feature classification model of a research area;
and taking the spectral characteristics and the texture characteristics of the preprocessed multi-temporal satellite data as input data, and inputting the input data into a ground object classification model of a research area to classify the ground objects, so as to obtain the mangrove distribution range of the research area.
Further, the step of acquiring spectral features and texture features of the multi-phase satellite image data based on the preprocessed multi-phase satellite image data specifically includes:
The method comprises the steps of selecting a wave band of preprocessed multi-temporal satellite image data, and selecting a red wave band, a green wave band, a blue wave band, a near infrared wave band and a short wave infrared wave band of the preprocessed multi-temporal satellite image data;
carrying out vegetation index calculation based on red light wave bands, green light wave bands, blue light wave bands, near infrared wave bands and short wave infrared wave bands of the preprocessed multi-time phase satellite image data, wherein the vegetation index calculation comprises normalized vegetation indexes, enhanced vegetation indexes, difference vegetation indexes and simple ratio vegetation indexes;
integrating the red light wave band, the green light wave band, the blue light wave band, the near infrared wave band and the short wave infrared wave band of the preprocessed multi-time-phase satellite image data and the normalized vegetation index, the enhanced vegetation index, the difference vegetation index and the simple ratio vegetation index of the preprocessed multi-time-phase satellite image data to construct the spectral characteristics of the multi-time-phase satellite image data;
calculating standard deviation of normalized vegetation indexes in a circular window with the radius of 5 pixels, and constructing texture features of multi-phase satellite image data.
Further, the step of obtaining corresponding mid-high spatial resolution multispectral satellite images according to the mangrove distribution range of the research area, obtaining the overground biomass of the vegetation survey sample side mangrove plant by combining the survey data of the research area, and constructing an inversion regression model of the overground biomass of the mangrove in the research area comprises the following steps:
Obtaining corresponding medium-high spatial resolution multispectral satellite images according to the positions of the mangrove ecological survey sample sides and the survey time;
Performing spectrum band selection and vegetation index calculation on the mid-high spatial resolution multispectral satellite image of the mangrove ecological survey sample position to obtain remote sensing spectrum band data and vegetation index data corresponding to the mangrove ecological survey sample position and survey time;
taking the mangrove plant overground biomass of the mangrove ecological survey sample side in the second kind of survey data of the research area as a dependent variable, and taking the remote sensing spectrum band data and the vegetation index data of the mangrove ecological survey sample side as independent variables to construct a mangrove overground biomass inversion regression model of the research area.
Further, the step of obtaining a carbon conversion coefficient of the mangrove plant overground carbon reserve in the research area by combining the mid-high spatial resolution multispectral satellite image of the mangrove distribution range in the research area and the inversion regression model of the mangrove plant overground carbon reserve in the research area comprises the following steps:
acquiring spectrum band data and vegetation index data of a mid-high spatial resolution multispectral satellite image in a mangrove forest distribution range of a research area;
based on spectral band data and vegetation index data of the mid-high spatial resolution multispectral satellite image in the distribution range of the mangrove in the research area, carrying out mangrove ground biomass inversion by using a mangrove ground biomass inversion regression model in the research area, and obtaining the mangrove ground biomass in the research area;
Acquiring biomass carbon conversion coefficients of mangrove forests on the research area by a literature research method;
Multiplying the above-ground biomass of the mangrove in the research area by the above-ground biomass carbon conversion coefficient to obtain the above-ground carbon reserve of the mangrove plant in the research area.
The method has the beneficial effects that: according to the invention, investigation data and multi-temporal satellite image data of a research area are obtained, and a low-tide satellite image meeting the requirements is screened out by constructing an inter-tide area index and an inter-tide area proportion, so that the automatic accurate identification of mangrove forests and coastal salt and biogas in the land-sea ecological staggered zone is realized by utilizing the difference of vegetation weathers of mangroves and coastal salt and biogas, a multi-spectrum satellite remote sensing image and on-site ecological investigation sample side data in the distribution range of the identified mangrove forests are based, a mangrove plant on-site carbon reserve remote sensing inversion model is constructed, the on-site carbon reserve remote sensing inversion automation degree of the mangrove forests is improved, the manual auxiliary workload is reduced, and technical support is provided for estimating the on-site carbon reserve of the mangrove forests in the sea-land ecological staggered zone with high spatial heterogeneity, and the method can be applied to investigation and monitoring of on-site carbon reserve distribution of the mangrove forests in a long time sequence and a large range.
Drawings
FIG. 1 is a flow chart of the steps of a method for estimating the carbon reserves on the land of a sea-land oriented ecological staggered zone mangrove plant according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of a satellite image data screening and preprocessing process according to an embodiment of the present invention;
FIG. 3 is a flowchart illustrating steps for performing mangrove distribution remote sensing identification in accordance with an embodiment of the present invention;
FIG. 4 is a flow chart of the steps for estimating the carbon reserves on the surface of mangrove plants according to an embodiment of the present invention;
FIG. 5 is a plot of a range vector data distribution diagram for an embodiment of the present invention;
FIG. 6 is a region diagram of a cut pre-processed sentinel satellite number two image and a normalized vegetation index greater than 0 acquired in accordance with an embodiment of the present invention;
FIG. 7 is a graph of intertidal range mask data for research areas made from satellite images of guard No. two with less than 20% cloud cover in accordance with an embodiment of the present invention;
FIG. 8 is a diagram of a training sample distribution diagram for classification of ground features in a research area in accordance with an embodiment of the present invention;
FIG. 9 is a diagram of classification of ground features in a research area in accordance with an embodiment of the present invention;
FIG. 10 is a mangrove distribution diagram of a research area in accordance with an embodiment of the present invention;
FIG. 11 is a diagram of a remote sensing inversion distribution of biomass on the mangrove land in a research area in accordance with an embodiment of the present invention;
FIG. 12 is a remote sensing inversion distribution diagram of the above-ground carbon reserves of mangrove plants in a research area in accordance with an embodiment of the present invention.
Detailed Description
The invention will now be described in further detail with reference to the drawings and to specific examples. The step numbers in the following embodiments are set for convenience of illustration only, and the order between the steps is not limited in any way, and the execution order of the steps in the embodiments may be adaptively adjusted according to the understanding of those skilled in the art.
The conventional mangrove carbon reserve remote sensing method generally obtains the mangrove distribution range of a research area as a range of the mangrove carbon reserve remote sensing inversion directly based on visual interpretation of remote sensing images, but the visual interpretation depends on experience of an interpreter and is influenced by subjective factors. Moreover, mangrove-coastal salt marsh ecological staggered zones are often distributed in coastal areas, the uneven distribution of the fragmentation of the mangrove and the similar spectral characteristics of the coastal salt marsh and the mangrove bring great challenges to the accurate identification of the distribution range of the mangrove, and the environment of tidal fluctuation brings great range uncertainty to the inversion of carbon reserves on the ground of the subsequent mangrove.
Explanation of technical terms of the present invention:
Carbon reserves refer to the total amount of organic carbon in a certain area of plant material and a certain depth of sediment. Above-ground biomass refers to the weight of all living plants above the ground surface expressed as dry plant weight per unit area, including above-ground stems, leaves, flowers, fruits, and the like. Subsurface biomass refers to the weight of all living plants expressed in terms of dry weight per unit area, including rhizomes, straight roots, fibrous roots, and the like. The above-ground carbon reserve of mangrove plant refers to the total amount of organic carbon in mangrove plant body above the ground surface in a certain area, and comprises above-ground stems, leaves, flowers, fruits and the like.
Based on the method, the invention provides a method for estimating the overground carbon reserves of the mangrove plants facing the sea-land ecological staggered zone. Firstly, combining multi-time-phase multi-spectrum satellite remote sensing images and field investigation data, screening out low-tide satellite images meeting requirements by constructing an inter-tide area index and an inter-tide area proportion, and realizing automatic and accurate identification of mangrove and coastal salt biogas by using a machine learning method by utilizing the vegetation weather difference of the mangrove and the coastal salt biogas; then, based on the multispectral satellite remote sensing image, the on-site ecological investigation sample side data and the mangrove abnormal speed growth equation in the identified mangrove forest distribution range, a mangrove forest plant on-ground carbon reserve remote sensing inversion model is constructed, the degree of automation of the mangrove forest plant on-ground carbon reserve investigation is improved, the manual auxiliary workload is reduced, technical method support is provided for the estimation of the mangrove forest plant on-ground carbon reserves in the sea-land ecological staggered zone with high spatial heterogeneity, and the method can be applied to investigation and monitoring of long-time sequence and large-range mangrove forest plant on-ground carbon reserves distribution.
Referring to fig. 1, the invention provides a method for estimating the carbon reserves on the land of a mangrove plant facing sea-land ecological staggered zone, which comprises the following steps:
S1, acquiring investigation data and multi-time satellite image data of a research area, respectively carrying out data preprocessing on the investigation data, and screening and data preprocessing on the multi-time satellite image data to obtain preprocessed investigation data and preprocessed multi-time satellite image data;
Specifically, setting a mangrove study area range; acquiring investigation data of a research area and preprocessing the data, wherein the process comprises the steps of shooting a mangrove research area range through an unmanned aerial vehicle to acquire first-class investigation data of the research area; the first type of investigation data of the investigation region is photos with different ground object types and shooting geographic position information, wherein the different ground object types comprise ground object types and vegetation types; setting a mangrove ecological survey sample position based on a research area range, measuring mangrove plants in the mangrove ecological survey sample position, obtaining plant height, breast diameter and base diameter of the mangrove plants in a mangrove ecological survey sample side, estimating overground biomass of different types of mangrove plants in the mangrove ecological survey sample side through a mangrove differential speed growth equation, and obtaining second type survey data of the research area; Integrating the first type of investigation data of the research area and the second type of investigation data of the research area to obtain preprocessed investigation data; carrying out multiple shooting processing on the mangrove research area range by a satellite multispectral imager to obtain multi-temporal satellite image data; selecting multi-temporal satellite image data corresponding to the cloud cover less than 20% by considering cloud cover interference, and selecting satellite images imaged in the growing period and the dormant period of the coastal salt biogas plant of a research area to obtain preliminarily screened multi-temporal satellite image data; sequentially performing geometric correction, radiometric calibration and atmospheric correction on the preliminarily screened multi-temporal satellite image data to obtain corrected multi-temporal satellite image data; The geometric correction is used for correcting geometric distortion caused in the imaging process of the sensor; the radiometric calibration is used for converting an electronic signal recorded by a remote sensing image sensor in the data input unit into the radiance or reflectivity of the ground surface feature; the atmospheric correction is used for eliminating the influence of atmospheric refraction and scattering and converting the radiation brightness or reflectivity into the actual reflectivity of the ground object; creating a research area range vector based on the mangrove research area range and in combination with photographs having photographed geographical location information in survey data of the research area; clipping the corrected multi-temporal satellite image data through a research area range vector to obtain clipped multi-temporal satellite image data; Screening the cut multi-temporal satellite image data by considering the influence of the tide fluctuation on the remote sensing imaging of the ground object spectrum and the distribution range thereof to obtain preprocessed multi-temporal satellite image data, wherein the area of the lowest tide of the research area and the normalized vegetation index of the highest tide of the research area and the area of the normalized water index of the highest tide of the research area and the normalized water index of the highest tide of the research area are smaller than 0by considering the influence of the tide fluctuation on the remote sensing imaging of the ground object spectrum and the distribution range thereof; constructing an intertidal zone area index and an intertidal zone area ratio according to the area of the study area when the lowest tide is the time and the normalized vegetation index is larger than 0, the area of the study area when the highest tide is the time and the normalized water body index is smaller than 0 and the area of the study area; And screening the preliminarily screened multi-temporal satellite image data according to the inter-tidal zone area index and the inter-tidal zone area ratio, and selecting an image with the inter-tidal zone area ratio being more than 10% and the largest inter-tidal zone area index as the screened and preprocessed multi-temporal satellite image data based on 2 time periods of the growing period and the dormancy period of the coastal salt biogas plant in the research area.
Further, as shown in fig. 2, multi-temporal satellite image data of the coastal salt marsh plants in the research area, which have a cloud cover of less than 20%, were selected as multi-temporal satellite image data of the cloud cover, and an intertidal zone area index and an intertidal zone area ratio were constructed. Considering that the tidal fluctuation has a large influence on the remote sensing imaging of the ground object spectrum and the distribution range thereof, and the research area is mainly outside the coastline, constructing an inter-tidal zone area index (IAI) and an inter-tidal zone area ratio (IAR), wherein the calculated expression of the inter-tidal zone area index is as follows:
；
the calculation expression of the area ratio of the intertidal zone is as follows:
；
In the above-mentioned method, the step of, Represents the area index of the intertidal zone,Represents the area proportion of the intertidal zone,Representing normalized vegetation index in a research area during satellite image imagingThe area of the region is such that,Representing the area of the investigation region,Representing normalized water index in research area during satellite image imagingThe area of the region is such that,Image normalization vegetation index representing lowest tide imaging of research areaThe area of the region is such that,Image normalization water index for representing imaging of highest tide in research areaThe area of the region;
it is worth mentioning that, AndMay be calculated based on historical satellite data for 1 year or more. The first method is to combine the local tide table of the research area, the local transit time of the satellite and the actual shooting time of the historical satellite data to obtain the satellite images imaged at the lowest tide and the highest tide of the research area; the second method is to make inter-tide band range data imaged at the time of the near-lowest tide and the near-highest tide of the satellite image shooting of the year based on historical satellite data of at least 1 year, and the specific operation is as follows:
(1) Collecting historical satellite data of a research area for at least 1 year, screening out satellite images with cloud cover less than 20%, and preprocessing to obtain preprocessed historical satellite images;
(2) Calculating normalized vegetation index And normalizing the water index；
(3) For preprocessed historical satellite imagesData set, productionAll mask data are combined to obtain the mask data of the inter-tidal zone range of the research area near the lowest tide time, and the inter-tidal zone range can be calculated；
(4) For preprocessed historical satellite imagesData set, productionCalculating mask data set, wherein all mask data are intersected to obtain the mask data of the range of the intertidal zone of the research area in the time of the near highest tide。
Wherein, the calculation expression for the normalized vegetation index is:
；
In the above-mentioned method, the step of, Represents the normalized vegetation index(s),Indicating the reflectivity in the near infrared band,Representing the reflectivity of the red band;
wherein, the calculation expression for the normalized water index is:
；
In the above-mentioned method, the step of, Represents the normalized water body index,Indicating the reflectivity in the near infrared band,Representing the reflectivity of the green band;
IAR and IAI indices were used to screen low tide satellite images. For example, a batch of satellite images of the growing period (generally also the time of field investigation) of the coastal salt biogas plants in the research area, and 1 scene images with IAR greater than 10% and maximum IAI index are selected as the classification base images of the first time phase and the base images of inversion of the overground carbon reserves of the mangrove plants; and (3) a batch of satellite images imaged in the dormancy period of the coastal salt biogas plants in the research area (autumn and winter in the current year of field investigation), selecting 1-scene images with IAR more than 10% and maximum IAI index as classification base images of a second time phase, and obtaining preprocessed multi-time-phase satellite image data through cloud cover, intertidal area index and intertidal area ratio screening and various preprocessing operations.
In this embodiment, further, in sunny days in the spring or summer, unmanned aerial vehicle, on-site photographing and other modes are used to obtain photos of different land feature types in the research area, and the land feature positions are recorded to obtain on-site investigation data of the land feature types and vegetation types, wherein the land feature types comprise mangrove forests, coastal salt marsh, light beaches, water bodies and the like. For multi-temporal satellite images, selecting imaging seasons of a research area as spring festival and winter or summer and winter, selecting multi-temporal satellite image data corresponding to cloud cover less than 20% in consideration of cloud cover interference, and acquiring a medium-high spatial resolution satellite image data set imaged at the lowest tide time of the growing period and the dormancy period of coastal salt marsh plants of the research area as basic data for distinguishing mangroves from coastal salt marsh in a sea-land ecological staggered zone in consideration of the influence of tidal fluctuation on the remote sensing imaging of ground object spectrum and the distribution range of the ground object spectrum. The preprocessing of satellite images comprises geometric correction, radiation correction and atmospheric correction, and the range vector of a research area is used for clipping the original data of the high-spatial-resolution multispectral remote sensing image.
S2, making classification training samples by the preprocessed investigation data, taking preprocessed multi-temporal satellite image data as classification basic data, and carrying out mangrove distribution remote sensing recognition by a random forest machine learning algorithm to obtain a mangrove distribution range of a research area;
Specifically, referring to fig. 3, spectral characteristics and texture characteristics of the multi-temporal satellite image data are calculated based on the preprocessed multi-temporal satellite image data, wherein a band selection is performed on the preprocessed multi-temporal satellite image data, red, green, blue, near-infrared and short-wave infrared bands of the preprocessed multi-temporal satellite image data are selected, vegetation index calculation is performed based on the red, green, blue, near-infrared and short-wave infrared bands of the preprocessed multi-temporal satellite image data, the vegetation index calculation comprises normalized vegetation indexes, enhanced vegetation indexes, differential vegetation indexes and simple ratio vegetation indexes, the normalized vegetation indexes, the enhanced vegetation indexes, the differential vegetation indexes and the simple ratio vegetation indexes of the preprocessed multi-temporal satellite image data are integrated, the normalized vegetation indexes, the enhanced vegetation indexes, the differential vegetation indexes and the simple ratio vegetation indexes of the preprocessed multi-temporal satellite image data of the integrated multi-temporal satellite image data are constructed, and the standard vegetation indexes with the radius of the normalized vegetation indexes within a circular window of 5 pixels are calculated;
setting random forest machine learning algorithm parameters, setting the number of decision trees to be generated as 100, setting the number of split nodes as the square root of all features, randomly selecting 70% of sample data in classified training samples as training samples, and using the remaining 30% of sample data as precision evaluation samples; spectral features and texture features of multi-temporal satellite image data corresponding to training samples are used as input data and are input into a random forest machine learning algorithm to construct a ground feature classification model of a research area; and taking the spectral characteristics and the texture characteristics of the preprocessed multi-temporal satellite data as input data, and inputting the input data into a ground object classification model of a research area to classify the ground objects, so as to obtain the mangrove distribution range of the research area.
In the embodiment, the classification training sample is prepared, and the classification training sample of the main ground object type of the research area is prepared based on the field investigation data of the main ground object type of the research area;
The process of selecting the wave band and calculating the vegetation index of the preprocessed multi-time phase satellite image data comprises the steps of selecting the red of the medium-high resolution multi-spectrum satellite data ) Green (green)) ' Lanzhu) Near infrared%) Short wave infrared ray) The band identification mangrove distribution information, 4 common vegetation indexes, namely a normalized vegetation index (Normalized Difference Vegetation Index, NDVI), an enhanced vegetation index (Enhance Vegetation Index, EVI), a differential vegetation index (DIFFERENCE VEGETATION INDEX, DVI), a simple ratio vegetation index (Simple Ratio Index, SRI), are used for vegetation extraction, and the texture features are represented by standard deviations of NDVI in a circular window with a radius of 5 pixels;
the calculation expression of the enhanced vegetation index is as follows:
；
In the above-mentioned method, the step of, Represents the index of the enhanced vegetation,Representing the reflectivity of the blue band;
The calculation expression of the differential vegetation index is as follows:
；
In the above-mentioned method, the step of, Representing a differential vegetation index;
the calculation expression of the simple ratio vegetation index is as follows:
；
In the above-mentioned method, the step of, Representing a simple ratio vegetation index;
And classifying the types of ground features such as mangrove forests, coastal salt marsh, light beaches, water bodies and the like in the research area by using a random forest machine learning algorithm, and obtaining the accurate data of the mangrove forests in the research area. Setting parameters of a random forest classification method, setting the number of decision trees to be generated as 100, and setting the number of split nodes as the square root of all features. And randomly selecting 70% of sample data from the classification training samples as training samples of the classification model, wherein the rest 30% of sample data can be used as precision evaluation samples for classification effect evaluation. The data input by training the random forest classification model are the spectrum and texture feature data of the multi-temporal satellite images corresponding to different ground object types in the research area. Wherein the spectral features include a plurality of spectral bands and spectral indices of the Sentinel-2 image: red, green, blue, near infrared and short wave infrared bands; NDVI, EVI, DVI and SRI index. Texture features are represented by standard deviations of NDVI within a circular window of radius 5 pixels.
S3, acquiring corresponding medium-high spatial resolution multispectral satellite images according to the distribution range of the mangrove in the research area, and constructing an inversion regression model of the biomass on the mangrove ground in the research area by combining the preprocessed investigation data;
Specifically, referring to fig. 4, based on the mangrove distribution range of the research area, a mangrove ecological survey sample position is set; measuring mangrove plants in the mangrove ecological investigation sample position, obtaining the plant height, breast diameter and base diameter of the mangrove plants in the mangrove ecological investigation sample side, and estimating the overground biomass of different types of mangrove plants in the mangrove ecological investigation sample side through a mangrove differential speed growth equation; acquiring corresponding medium-high spatial resolution multispectral satellite images based on the mangrove ecological survey sample position; performing spectrum band selection and vegetation index calculation on the mid-high spatial resolution multispectral satellite images of the mangrove ecological survey sample position to obtain remote sensing spectrum band data and vegetation index data corresponding to the mangrove ecological survey sample position; taking the mangrove plant overground biomass of the mangrove ecological investigation sample side as a dependent variable, taking remote sensing spectrum band data and vegetation index data of the mangrove ecological investigation sample side as independent variables, and constructing a mangrove overground biomass inversion regression model of a research area.
In this embodiment, the site mangrove ecological survey data is pre-processed with satellite data. According to the range of the research area and the distribution characteristics of mangroves, arranging mangrove ecological investigation sample sides, wherein the area of each sample side is 10m Í and 10m, measuring the plant height, breast diameter and base diameter of mangrove plants in each sample side, and estimating the aboveground biomass of different types of mangrove plants according to the different-speed growth equation of the mangrove;
selecting a middle-high spatial resolution multispectral satellite image with imaging time closest to mangrove ecological investigation time (the time interval is preferably within 1 month) and imaging at low tide, performing geometric correction, radiation correction and atmospheric correction on the satellite image, and cutting the satellite image through a mangrove distribution vector range of a research area;
band selection and vegetation index calculation; selecting the red of the medium-high resolution multispectral satellite data ) Green (green)) ' Lanzhu) Near infrared%) Short wave infrared ray) The wave band, NDVI, EVI, DVI and SRI vegetation index are taken as dependent variables, the mangrove overground biomass value of a mangrove ecological investigation sample party serving as a training sample is taken as independent variables, the spectral reflectivity of the sample party on an image and the vegetation index are taken as independent variables, and a mangrove overground biomass inversion regression model of a research area is constructed.
In this embodiment, 70% of the mangrove ecological survey sample is further randomly selected as a training sample of the above-ground biomass regression model, and the remaining 30% of the sample can be used as a sample for evaluating the accuracy of the regression model. Taking the mangrove aerial biomass numerical value counted by the mangrove ecological investigation sampling party serving as a training sample as a dependent variable, taking the image spectral reflectivity and the vegetation index of the position of the sampling party as independent variables, and constructing a mangrove aerial biomass inversion regression model of a research area. The accuracy of the biomass inversion model was evaluated using root mean square error (Root Mean Square Error, RMSE). RMSE is a measure of the deviation between the predicted value and the measured value, the smaller the value, the higher the accuracy of the prediction model;
the calculation formula of RMSE is as follows:
；
In the above-mentioned method, the step of, The root mean square error is indicated as,Representing a sampleThe predicted value of the current value,Representing a sampleIs used for the measurement of (a),Representing the number of samples.
S4, acquiring a carbon conversion coefficient of the mangrove plant overground carbon, and estimating the research area mangrove plant overground carbon reserve by combining the middle-high spatial resolution multispectral satellite image of the research area mangrove distribution range and the research area mangrove plant overground carbon reserve by using a regression model of the research area mangrove plant overground biomass inversion.
Specifically, based on the spectral characteristics of multi-temporal satellite image data, performing mangrove land biomass inversion by using a mangrove land biomass inversion regression model of a research area to obtain mangrove land biomass of the research area; obtaining the carbon conversion coefficient of the overground biomass of the mangrove plant according to the methods of literature investigation or laboratory experiments and the like; multiplying the above-ground biomass of the mangrove in the research area by the above-ground biomass carbon conversion coefficient to obtain the above-ground carbon reserve of the mangrove plant in the research area.
Furthermore, the invention takes a sentinel second satellite image as an example to provide a mangrove plant overground carbon reserve estimation method applicable to sea-land ecological staggered zones, and the steps of the invention are illustrated by examples;
Performing field investigation;
On cloudless sunny days (generally in spring festival or summer, namely the growing period of coastal salt marsh plants in a research area), taking photos of different ground object types (including mangrove forests, coastal salt marsh, light beaches, water bodies and the like) in the research area by using unmanned aerial vehicles, on-site cameras and other modes, and recording the ground object positions; according to the range of the research area and the distribution characteristics of the mangrove, arranging mangrove ecological investigation sample sides, wherein the area of each sample side is 10m Í and 10m, measuring the plant height, breast diameter and base diameter of mangrove plants in each sample side, and obtaining the ground object type, vegetation type and field investigation data of the mangrove ecological investigation sample sides of the research area;
Carrying out data preprocessing on the second satellite image of the sentinel;
The second sentinel is an important optical remote sensing satellite in the "gothic" plan series jointly implemented by the european commission and the european space agency, has 13 spectral bands, and the spatial resolutions of different bands are slightly different, as shown in table 1, and the L2A grade product (L2A grade product is data after orthographic correction, radiometric calibration and atmospheric correction are performed) of the second sentinel satellite imaged by selecting a first time period (i.e. time of field ecological investigation, generally spring or summer) and a second time period (autumn or winter of the first time period of field ecological investigation) from the gothic open center. The multispectral imager carried by the sentinel second satellite can provide images of vegetation, soil and water coverage, inland waterway, coastal areas and the like;
Table 1 Path No. two satellite camera spectral band parameter list
；
In this example, before preprocessing the original data, the initial sentinel second satellite data satisfying a certain condition needs to be selected, which specifically includes: selecting a first time period (namely the time of on-site ecological investigation, namely the growing period of coastal salt biogas plants in the research area, spring or summer, here 2022 year 4 month images) and a second time period (namely the dormant period of the coastal salt biogas plants in the research area, here 2022 year 11 month images) of imaging a sentinel second satellite L2A level product (the L2A level product is data after orthographic correction, radiometric calibration and atmospheric correction are carried out), enabling an initial remote sensing image to be in a preset cloud amount range, enabling the system to cover the whole research area in low tide, selecting a sentinel second multispectral data L2A level product, carrying out orthographic correction, radiometric calibration and atmospheric correction on the product, and cutting the sentinel second multispectral image data by using the research area range vector data shown in fig. 5 to obtain multi-spectral remote sensing image data after pretreatment of the research area;
In a research area in Guangdong, 4 months of 2022, field investigation is carried out, and the second satellite image of the sentinel with the cloud content of less than 20% in the area is only 2 scenes, namely imaging data of 12 days of 3 months and 6 days of 4 months in the current year. For the two-scene images, an IAI index is calculated, and the image of 4 months and 6 days with the largest IAI is selected as the basic data for classification and inversion of the carbon reserves on the mangrove plant ground, specifically as shown in the following table 2:
Table 2 study area 3 months to 5 months cloud cover less than 20% guard number two satellite dataset IAR and IAI index
FIG. 6 shows a pre-processed sentinel satellite image number two after cutting out on month 4 and 6 of study area 2022 FIG. 6 (a) and an area FIG. 6 (b) with normalized vegetation index greater than 0; pre-processing a sentinel second satellite image after 2022, 3 and 12 days of cutting in a research area 6 (c) and a region map with a normalized vegetation index greater than 0 (d);
Fig. 7 shows the inter-tidal zone range mask data of the investigation region at the time of the year near the lowest tide prepared based on the satellite image of the sentinel No. two with the cloud cover of less than 20% in the investigation region 2021 fig. 7 (a) and the inter-tidal zone range mask data of the investigation region at the time of the highest tide fig. 7 (b).
Manufacturing a classification training sample;
Aiming at main ground object types (generally including mangrove forest, coastal salt marsh, light beach, water body and the like) of a research area, based on collected unmanned aerial vehicle images, a classification training sample of different ground object types is obtained, namely, shown in fig. 8;
Spectral band selection and vegetation index calculation;
Red of mid-high resolution multispectral satellite data is selected (B4, ) The color of green (B3,) Blue (B2),) Near infrared (B5-B8,) And short-wave infrared (B11,) The band identifies mangrove distribution information. 4 common vegetation indices, namely normalized vegetation index (Normalized Difference Vegetation Index, NDVI), enhanced vegetation index (Enhance Vegetation Index, EVI), differential vegetation index (DIFFERENCE VEGETATION INDEX, DVI), simple ratio vegetation index (Simple Ratio Index, SRI), for vegetation extraction, texture features are represented by standard deviations of NDVI within a circular window of radius 5 pixels;
classifying to obtain the mangrove forest distribution range of the sea-land ecological staggered zone;
The method comprises the steps of classifying the types of ground features such as mangrove, coastal salt marsh, light beach, water body and the like in a research area through a random forest machine learning algorithm, namely, the method is shown in fig. 9, and extracting the information of the distribution range of the mangrove in the research area, namely, the information of the distribution range of the mangrove in the research area is shown in fig. 10. And randomly selecting 70% of sample data from the classification training samples as training samples of the classification model, wherein the rest 30% of sample data can be used as precision evaluation samples for classification effect evaluation. The random forest classification model training input data uses satellite image data based on study area 2022, 4 and 11 months to calculate image spectral features and texture feature data for different terrain type training samples. Wherein, spectral feature includes a plurality of spectral bands and spectral index of sentinel second number image: red, green, blue, near infrared and short wave infrared 8 wave bands; NDVI, EVI, DVI and SRI index. Texture features are represented by standard deviations of NDVI within a circular window of radius 5 pixels. Setting parameters of a random forest classification method, setting the number of decision trees to be generated as 100, and setting the number of split nodes as the square root of all features;
preprocessing mangrove ecological sample side investigation data and satellite remote sensing data;
Based on parameters such as plant height, breast diameter and the like of mangrove plants in each sample side obtained by the investigation of mangrove ecological sample sides, the aboveground biomass of each mangrove is estimated according to a mangrove abnormal speed growth equation, and then the aboveground biomass of all mangroves in each sample side is obtained. Randomly taking 70% of the number of the vegetation sample of the mangrove forest to be investigated as a training sample, and taking 30% of the number of the remaining sample as an accuracy evaluation sample. Obtaining carbon conversion coefficients of the overground biomass of different mangrove plants according to the methods of literature investigation or laboratory experiments and the like. The mangrove plants in the research area of the example are mainly mulberry leaves without petals, and according to the data of references and the like, the carbon conversion coefficient of the overground biomass is 0.43, and the abnormal growth equation is as follows:
；
；
In the above-mentioned method, the step of, Represents the chest diameter (cm),Represents the plant height (m),Represents the above-ground biomass (kg),Represents the maximum diameter (cm) of the sample tree,The decision coefficient of the abnormal growth equation is an index for measuring the fitting degree of the statistical model and the actual data,Representing the number of mangrove plants used in the construction of the differential growth equation;
Selecting a sentinel second multispectral satellite image L2A grade product with imaging time closest to mangrove ecological investigation time (the time interval is preferably within 1 month) and imaging at low tide and with less cloud shielding, and cutting the satellite image according to the mangrove distribution vector range of the research area obtained above.
Band selection and vegetation index calculation;
taking the mangrove aerial biomass value of a mangrove ecological investigation sample party serving as a training sample as a dependent variable, taking the spectral reflectivity and vegetation index of the sample party on satellite images as independent variables, and constructing a mangrove aerial biomass inversion regression model of a research area. Performing accuracy evaluation on the biomass inversion model by using root mean square error (Root Mean Square Error, RMSE), wherein the RMSE is the deviation between a measurement value and a measurement value, and the smaller the value is, the higher the accuracy of the prediction model is;
Estimating the overground carbon reserves of mangrove plants;
Based on satellite image spectral reflectivity and vegetation index data in the distribution range of the mangrove in the research area, the constructed regression model is used for inverting the biomass on the mangrove in the research area, namely the biomass is shown in figure 11. When mangrove plants in the research area are mainly one kind of mangrove plants, multiplying the inverted overground biomass of the mangrove plants in the research area by the overground biomass carbon conversion coefficient of the mangrove plants, namely the overground carbon reserves of the mangrove plants in the research area are shown in figure 12; when the main distribution mangrove vegetation species in the research area is more than one, there are two treatment modes: (a) Assuming that the carbon conversion coefficient of the biomass on the local mangrove plant land is 50%, multiplying the biomass on the mangrove plant land obtained by inversion by 50%, and obtaining the biomass on the mangrove plant land in the research area; (b) The method comprises the steps of finely classifying the distribution range of mangrove plant types of main mangrove vegetation types in a research area by using a machine learning algorithm, obtaining distribution data of the mangrove plant types in the research area, and multiplying the above-ground biomass carbon conversion coefficient of each mangrove vegetation type by the above-ground biomass of the mangrove forest of the type to obtain the above-ground biomass carbon reserves of the mangrove plant type.
While the preferred embodiment of the present application has been described in detail, the application is not limited to the embodiment, and various equivalent modifications and substitutions can be made by those skilled in the art without departing from the spirit of the application, and these equivalent modifications and substitutions are intended to be included in the scope of the present application as defined in the appended claims.
Claims (3)
1. The method for estimating the overground carbon reserves of the mangrove plant facing the sea-land ecological staggered zone is characterized by comprising the following steps of:
Acquiring investigation data and multi-temporal satellite image data of a research area, performing data preprocessing on the investigation data, screening and data preprocessing on the multi-temporal satellite image data to obtain preprocessed investigation data and preprocessed multi-temporal satellite image data, wherein the method comprises the following steps of:
Setting a mangrove study area range and acquiring investigation data of a corresponding study area;
data preprocessing is carried out on investigation data of a research area to obtain preprocessed investigation data, and the method comprises the following steps:
Shooting a mangrove research area range through an unmanned aerial vehicle to acquire first-class investigation data of the research area;
The first type of investigation data of the investigation region is photos with different ground object types and shooting geographic position information, wherein the different ground object types comprise ground object types and vegetation types;
Setting a mangrove ecological survey sample position based on a research area range, measuring mangrove plants in the mangrove ecological survey sample position, and obtaining plant height, breast diameter and base diameter information of the mangrove plants in the mangrove ecological survey sample;
Estimating the overground biomass of mangrove plants of different species of mangrove ecological investigation sample sides through a mangrove abnormal growth equation according to the plant height, breast diameter and base diameter information of the mangrove plants, obtaining the overground biomass of the mangrove plants of the mangrove ecological investigation sample sides, and obtaining second-type investigation data of a study area;
integrating the first type of investigation data of the research area and the second type of investigation data of the research area to obtain preprocessed investigation data;
shooting the mangrove research area range through a satellite multispectral imager to obtain multi-temporal satellite image data;
Screening and data preprocessing are carried out on the multi-temporal satellite image data to obtain preprocessed multi-temporal satellite image data, wherein the method comprises the following steps:
selecting satellite image data corresponding to the cloud cover less than 20% by considering cloud cover interference, and selecting satellite images imaged in the growing period and the dormant period of the coastal salt biogas plant in a research area to obtain preliminarily screened multi-temporal satellite image data;
sequentially performing geometric correction, radiometric calibration and atmospheric correction on the preliminarily screened multi-temporal satellite image data to obtain corrected multi-temporal satellite image data;
The geometric correction is used for correcting geometric distortion caused in the imaging process of the sensor;
The radiometric calibration is used for converting an electronic signal recorded by a remote sensing image sensor in the data input unit into the radiance or reflectivity of the ground surface feature;
the atmosphere correction is used for eliminating the influence of atmosphere refraction and scattering and converting the radiation brightness or reflectivity into the actual reflectivity of the ground object;
based on the mangrove study area range and combined with photos with geographical position information in the investigation data of the study area, making a study area range vector;
Clipping the corrected multi-temporal satellite image data through a research area range vector to obtain clipped multi-temporal satellite image data;
Taking into consideration the influence of tidal fluctuation on remote sensing imaging of ground object spectrum and distribution range thereof, screening the cut multi-temporal satellite image data to obtain preprocessed multi-temporal satellite image data, wherein the method comprises the following steps:
taking the influence of tidal fluctuation on the remote sensing imaging of the ground object spectrum and the distribution range thereof into consideration, and acquiring the area of the investigation region when the lowest tide is greater than 0 and the area of the investigation region when the highest tide is less than 0 and the normalized water body index;
constructing an intertidal zone area index and an intertidal zone area ratio according to the area of the study area when the lowest tide is the time and the normalized vegetation index is larger than 0, the area of the study area when the highest tide is the time and the normalized water body index is smaller than 0 and the area of the study area;
wherein the expression of the inter-tidal zone area index and the inter-tidal zone area ratio is specifically as follows:
；
；
In the above-mentioned method, the step of, Represents the area index of the intertidal zone,Represents the area proportion of the intertidal zone,Representing normalized vegetation index in a research area during satellite image imagingThe area of the region is such that,Representing the area of the investigation region,Representing normalized water index in research area during satellite image imagingThe area of the region is such that,Image normalization vegetation index representing lowest tide imaging of research areaThe area of the region is such that,Image normalization water index for representing imaging of highest tide in research areaThe area of the region;
Screening the preliminarily screened multi-temporal satellite image data according to the area indexes of the intertidal zones and the area ratio of the intertidal zones, and selecting an image with the area ratio of the intertidal zones being more than 10% and the largest area index of the intertidal zones as the preprocessed multi-temporal satellite image data based on the growing period and the dormancy period of the coastal salt biogas plants in the research area;
Making a classification training sample based on the preprocessed investigation data, taking the preprocessed multi-temporal satellite image data as classification basic data, and carrying out remote sensing recognition on the mangrove distribution range through a random forest machine learning algorithm to obtain the mangrove distribution range of a research area, wherein the method comprises the following steps:
Calculating spectral features and texture features of the multi-phase satellite image data based on the preprocessed multi-phase satellite image data;
based on first-class investigation data of a research area, making classification training samples of different ground objects;
Setting random forest machine learning algorithm parameters, setting the number of decision trees to be generated as 100, setting the number of split nodes as the square root of all features, randomly selecting 70% of sample data in classified training samples as training samples, and using the remaining 30% of sample data as precision evaluation samples;
Spectral features and texture features of multi-temporal satellite image data corresponding to training samples are used as input data and are input into a random forest machine learning algorithm to construct a ground feature classification model of a research area;
Taking the spectral characteristics and the texture characteristics of the preprocessed multi-temporal satellite data as input data, and inputting the input data into a ground object classification model of a research area to classify the ground objects, so as to obtain the mangrove distribution range of the research area;
Acquiring corresponding medium-high spatial resolution multispectral satellite images according to the distribution range of the mangrove in the research area and combining the preprocessed investigation data to construct an inversion regression model of the biomass on the mangrove ground in the research area, wherein the method comprises the following steps:
Obtaining corresponding medium-high spatial resolution multispectral satellite images according to the positions of the mangrove ecological survey sample sides and the survey time;
Performing spectrum band selection and vegetation index calculation on the mid-high spatial resolution multispectral satellite image of the mangrove ecological survey sample position to obtain remote sensing spectrum band data and vegetation index data corresponding to the mangrove ecological survey sample position and survey time;
Taking the mangrove plant overground biomass of the mangrove ecological survey sample side in the second type of survey data of the research area as a dependent variable, and taking remote sensing spectrum band data and vegetation index data of the mangrove ecological survey sample side as independent variables to construct a mangrove overground biomass inversion regression model of the research area;
And acquiring a carbon conversion coefficient of the biomass on the mangrove plant ground, and estimating the carbon reserve of the mangrove plant ground in the research area by combining the middle-high spatial resolution multispectral satellite image of the distribution range of the mangrove in the research area and the inversion regression model of the biomass on the mangrove plant ground in the research area.
2. The method for estimating the aboveground carbon reserves of the sea-land oriented ecological staggered zone mangrove plant according to claim 1, wherein the step of acquiring the spectral features and the texture features of the multi-phase satellite image data based on the preprocessed multi-phase satellite image data specifically comprises the following steps:
The method comprises the steps of selecting a wave band of preprocessed multi-temporal satellite image data, and selecting a red wave band, a green wave band, a blue wave band, a near infrared wave band and a short wave infrared wave band of the preprocessed multi-temporal satellite image data;
carrying out vegetation index calculation based on red light wave bands, green light wave bands, blue light wave bands, near infrared wave bands and short wave infrared wave bands of the preprocessed multi-time phase satellite image data, wherein the vegetation index calculation comprises normalized vegetation indexes, enhanced vegetation indexes, difference vegetation indexes and simple ratio vegetation indexes;
integrating the red light wave band, the green light wave band, the blue light wave band, the near infrared wave band and the short wave infrared wave band of the preprocessed multi-time-phase satellite image data and the normalized vegetation index, the enhanced vegetation index, the difference vegetation index and the simple ratio vegetation index of the preprocessed multi-time-phase satellite image data to construct the spectral characteristics of the multi-time-phase satellite image data;
calculating standard deviation of normalized vegetation indexes in a circular window with the radius of 5 pixels, and constructing texture features of multi-phase satellite image data.
3. The method for estimating the above-ground carbon reserves of the mangrove plant in the sea-land oriented ecological staggered zone according to claim 2, wherein the step of obtaining the above-ground biomass carbon conversion coefficient of the mangrove plant in the research area and combining the mid-high spatial resolution multispectral satellite image of the distribution range of the mangrove plant in the research area and the inversion regression model of the above-ground biomass of the mangrove plant in the research area to obtain the above-ground carbon reserves of the mangrove plant in the research area specifically comprises the following steps:
acquiring spectrum band data and vegetation index data of a mid-high spatial resolution multispectral satellite image in a mangrove forest distribution range of a research area;
based on spectral band data and vegetation index data of the mid-high spatial resolution multispectral satellite image in the distribution range of the mangrove in the research area, carrying out mangrove ground biomass inversion by using a mangrove ground biomass inversion regression model in the research area, and obtaining the mangrove ground biomass in the research area;
Acquiring biomass carbon conversion coefficients of mangrove forests on the research area by a literature research method;
Multiplying the above-ground biomass of the mangrove in the research area by the above-ground biomass carbon conversion coefficient to obtain the above-ground carbon reserve of the mangrove plant in the research area.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311156020.3A CN117218531B (en) | 2023-09-08 | 2023-09-08 | Sea-land ecological staggered zone mangrove plant overground carbon reserve estimation method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311156020.3A CN117218531B (en) | 2023-09-08 | 2023-09-08 | Sea-land ecological staggered zone mangrove plant overground carbon reserve estimation method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117218531A CN117218531A (en) | 2023-12-12 |
| CN117218531B true CN117218531B (en) | 2024-06-28 |
Family
ID=89047394
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311156020.3A Active CN117218531B (en) | 2023-09-08 | 2023-09-08 | Sea-land ecological staggered zone mangrove plant overground carbon reserve estimation method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117218531B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118111928B (en) * | 2024-04-30 | 2024-07-12 | 安徽建工生态科技股份有限公司 | Automatic carbon sink determination system and method for greenbelt ecosystem |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115561773A (en) * | 2022-12-02 | 2023-01-03 | 武汉大学 | Forest carbon reserve inversion method based on ICESat-2 satellite-borne LiDAR data and multispectral data |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111767865A (en) * | 2020-06-30 | 2020-10-13 | 北部湾大学 | Method for inverting mangrove forest biomass by using aerial image and laser data |
| CN112713923B (en) * | 2020-11-30 | 2022-07-15 | 上海航天控制技术研究所 | Dynamic scheduling method for satellite integrated electronic data link |
| CN114445719A (en) * | 2021-11-29 | 2022-05-06 | 南京林业大学 | Remote sensing data-based vegetation classification and biomass inversion method |
| CN115082809B (en) * | 2022-06-23 | 2023-02-17 | 宁波大学 | New tidal flat evolution monitoring method based on remote sensing image big data |
| CN116188793A (en) * | 2022-11-09 | 2023-05-30 | 江苏省农业科学院 | Astragalus sinicus planting area monitoring method based on satellite remote sensing image |
| CN116128353A (en) * | 2023-01-30 | 2023-05-16 | 北京林业大学 | Dominant tree species BOCs emission estimation method based on remote sensing biomass inversion |
-
2023
- 2023-09-08 CN CN202311156020.3A patent/CN117218531B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115561773A (en) * | 2022-12-02 | 2023-01-03 | 武汉大学 | Forest carbon reserve inversion method based on ICESat-2 satellite-borne LiDAR data and multispectral data |
Non-Patent Citations (1)
| Title |
|---|
| 基于高分辨影像机器学习算法的落叶松人工林提取研究;李煜;《中国优秀硕士学位论文全文数据库农业科技辑》;20220215;第8-16页、第22-24页、第48页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117218531A (en) | 2023-12-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105678281B (en) | Remote sensing monitoring method for mulching film farmland based on spectrum and texture characteristics | |
| CN111242224B (en) | Multi-source remote sensing data classification method based on unmanned aerial vehicle extraction classification sample points | |
| CN106372592B (en) | A kind of winter wheat planting area calculation method based on winter wheat area index | |
| CN112800973B (en) | Spartina alterniflora extraction method based on vegetation phenological feature decision | |
| Al-Ali et al. | A comparative study of remote sensing classification methods for monitoring and assessing desert vegetation using a UAV-based multispectral sensor | |
| Sun et al. | Classification mapping and species identification of salt marshes based on a short-time interval NDVI time-series from HJ-1 optical imagery | |
| CN108458978B (en) | Sensitive waveband and waveband combination optimal tree species multispectral remote sensing identification method | |
| Kordi et al. | Crop classification based on phenology information by using time series of optical and synthetic-aperture radar images | |
| US20120155714A1 (en) | Vegetation indices for measuring multilayer microcrop density and growth | |
| CN108932521B (en) | Deep learning-based crop classification method and system | |
| CN105758806B (en) | Remote sensing monitoring method for mulching film farmland based on spectral characteristics | |
| CN109816542B (en) | Method and system for reducing yield and settling claim of crops | |
| CN113221765B (en) | Vegetation phenological period extraction method based on digital camera image effective pixels | |
| CN115810155B (en) | Tidal wetland classification method | |
| CN117218531B (en) | Sea-land ecological staggered zone mangrove plant overground carbon reserve estimation method | |
| US20220392215A1 (en) | System and Method for Mapping Land Cover Types with Landsat, Sentinel-1, and Sentinel-2 Images | |
| CN109784300A (en) | A kind of crops science survey production method and system | |
| CN114049564B (en) | Pine wood nematode disease level prediction model construction method based on hyperspectral remote sensing image | |
| Manakos et al. | Comparison between atmospheric correction modules on the basis of worldview-2 imagery and in situ spectroradiometric measurements | |
| CN116824384A (en) | Soybean identification method based on standard curve | |
| CN114778483A (en) | Method for correcting terrain shadow of remote sensing image near-infrared wave band for monitoring mountainous region | |
| CN113887493B (en) | Black and odorous water body remote sensing image identification method based on ID3 algorithm | |
| CN105678280B (en) | Mulching film mulching farmland remote sensing monitoring method based on textural features | |
| CN114612794A (en) | Remote sensing identification method for land covering and planting structure in finely-divided agricultural area | |
| Pauly | Applying conventional vegetation vigor indices to UAS-derived orthomosaics: issues and considerations |
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 |