CN107766799A - The analysis method and system of multi- source Remote Sensing Data data source remittance landscape based on scale effect - Google Patents
The analysis method and system of multi- source Remote Sensing Data data source remittance landscape based on scale effect Download PDFInfo
- Publication number
- CN107766799A CN107766799A CN201710899736.0A CN201710899736A CN107766799A CN 107766799 A CN107766799 A CN 107766799A CN 201710899736 A CN201710899736 A CN 201710899736A CN 107766799 A CN107766799 A CN 107766799A
- Authority
- CN
- China
- Prior art keywords
- landscape
- analysis
- index
- source
- classification
- 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.)
- Granted
Links
- 238000004458 analytical method Methods 0.000 title claims abstract description 56
- 230000000694 effects Effects 0.000 title claims abstract description 31
- 238000009826 distribution Methods 0.000 claims abstract description 58
- 238000000611 regression analysis Methods 0.000 claims abstract description 21
- 239000000443 aerosol Substances 0.000 claims abstract description 18
- 230000003287 optical effect Effects 0.000 claims abstract description 15
- 238000010219 correlation analysis Methods 0.000 claims description 15
- 239000003643 water by type Substances 0.000 claims description 5
- 230000021615 conjugation Effects 0.000 claims description 2
- 240000007594 Oryza sativa Species 0.000 claims 1
- 235000007164 Oryza sativa Nutrition 0.000 claims 1
- 235000009566 rice Nutrition 0.000 claims 1
- 238000005259 measurement Methods 0.000 abstract description 2
- 230000009466 transformation Effects 0.000 abstract description 2
- 238000011160 research Methods 0.000 description 6
- 230000008859 change Effects 0.000 description 5
- 230000000875 corresponding effect Effects 0.000 description 5
- 238000000034 method Methods 0.000 description 5
- 230000002596 correlated effect Effects 0.000 description 4
- 239000000523 sample Substances 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 3
- 238000012417 linear regression Methods 0.000 description 3
- 150000001875 compounds Chemical class 0.000 description 2
- 230000001419 dependent effect Effects 0.000 description 2
- 238000009792 diffusion process Methods 0.000 description 2
- 239000003344 environmental pollutant Substances 0.000 description 2
- 238000003306 harvesting Methods 0.000 description 2
- 239000000203 mixture Substances 0.000 description 2
- 238000010606 normalization Methods 0.000 description 2
- 231100000719 pollutant Toxicity 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 230000009467 reduction Effects 0.000 description 2
- 230000011218 segmentation Effects 0.000 description 2
- 238000013179 statistical model Methods 0.000 description 2
- 125000000391 vinyl group Chemical group [H]C([*])=C([H])[H] 0.000 description 2
- 229920002554 vinyl polymer Polymers 0.000 description 2
- 241001269238 Data Species 0.000 description 1
- VAYOSLLFUXYJDT-RDTXWAMCSA-N Lysergic acid diethylamide Chemical compound C1=CC(C=2[C@H](N(C)C[C@@H](C=2)C(=O)N(CC)CC)C2)=C3C2=CNC3=C1 VAYOSLLFUXYJDT-RDTXWAMCSA-N 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 239000003054 catalyst Substances 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000004069 differentiation Effects 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 239000004744 fabric Substances 0.000 description 1
- 239000002803 fossil fuel Substances 0.000 description 1
- 239000007789 gas Substances 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 239000002245 particle Substances 0.000 description 1
- 239000002243 precursor Substances 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000004445 quantitative analysis Methods 0.000 description 1
- VMXUWOKSQNHOCA-UKTHLTGXSA-N ranitidine Chemical compound [O-][N+](=O)\C=C(/NC)NCCSCC1=CC=C(CN(C)C)O1 VMXUWOKSQNHOCA-UKTHLTGXSA-N 0.000 description 1
- 150000003839 salts Chemical class 0.000 description 1
- 230000001568 sexual effect Effects 0.000 description 1
- 239000004071 soot Substances 0.000 description 1
- 238000013517 stratification Methods 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 239000002912 waste gas Substances 0.000 description 1
- 239000002023 wood Substances 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
- G06V20/188—Vegetation
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N15/00—Investigating characteristics of particles; Investigating permeability, pore-volume, or surface-area of porous materials
- G01N15/06—Investigating concentration of particle suspensions
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/46—Descriptors for shape, contour or point-related descriptors, e.g. scale invariant feature transform [SIFT] or bags of words [BoW]; Salient regional features
- G06V10/462—Salient features, e.g. scale invariant feature transforms [SIFT]
- G06V10/464—Salient features, e.g. scale invariant feature transforms [SIFT] using a plurality of salient features, e.g. bag-of-words [BoW] representations
-
- G01N15/075—
Abstract
The analysis method and system of a kind of multi- source Remote Sensing Data data source remittance landscape based on scale effect are first, satellite-based remote sensing image obtains ground mulching classification results, the overall landscape heterogeneity index of ground mulching under different scale is calculated, selects optimal scale of the heterogeneous maximum yardstick as source remittance landscape Analysis；On this basis, the measurement of air haze pollution level is used as by the use of the aerosol optical depth AOD products of MODIS data, based on the less classification level landscape index of autocorrelation, the landscape that converged using Geographical Weighted Regression analysis model to source carries out local regression analysis with AOD, obtains the Geographical Weighted Regression coefficient of analysis distribution of source remittance landscape.Implement the present invention, analyzing for the air haze of source remittance landscape that can be to relevant position, draw the air haze situation of this area, help preferably to carry out the allocation plan and transformation in area.
Description
Technical field
The present invention relates to atmospheric environment field, more specifically to a kind of multi- source Remote Sensing Data data based on scale effect
The analysis method and system of source remittance landscape.
Background technology
In recent years, developing rapidly with urban economy, urban atmospheric pollution is on the rise, and haze weather occurs again and again.Haze
The aggregate of a large amount of small dusts in air, soot or salt grain is suspended in, is to be less than 80% in relative humidity of atomsphere, it is horizontal
Low visibility is in a kind of 10 kilometers of weather phenomenon." source " " remittance " landscape theories are researched and proposed based on atmosphere pollution, for
Air haze pollutes, and landscape ecology thinks, " source " landscape refers to the landscape for increasing haze concentration, and " remittance " landscape refers to suppress haze concentration
Increase, the landscape for reducing atmosphere pollution.In Process of Urbanization, how under the conditions of limited land resource all kinds of scapes of reasonable Arrangement
It is most important to see the spatial distribution of type, " source " " remittance " landscape skewness weighs, and can influence the diffusion velocity of pollutant in city
And concentration distribution, the change for causing urban atmosphere to form, so as to cause the change of city atmospheric environment.Make rational planning for city " source "
" remittance " landscape pattern, it is possible to reduce air haze pollutes, and reduces the negative changes of urban construction caused atmospheric environment in itself.
At present, the research of air haze pollution focuses mostly in research meteorologic factor to haze formation and the influence developed, the matter of haze
Measure concentration and chemical composition and this 3 aspects of transmission locus of continuation haze pollution.And for city " source " " remittance " landscape lattice
The research of relation is abundant not enough between office and the pollution of urban atmosphere haze.The major reason that air haze pollutes occurs for city
Urban landscape pattern is unreasonable, the distribution of " source " " remittance " landscape is unbalance, urban inner pollution is difficult to spread in time, therefore accurate differentiation
" source " " remittance " landscape, the dependency relation of analysis city " source " " remittance " landscape pattern and the pollution of air haze, city of making rational planning for " source "
" remittance " landscape pattern, it is the effective way for reducing the pollution of air haze.Then at present, do not have related technology on the market to realize
The analysis of the air haze of " source " " remittance " landscape.
The content of the invention
The technical problem to be solved in the present invention is, is realized for the above-mentioned existing technology for not having correlation on the market
The technological deficiency of the analysis of the air haze of " source " " remittance " landscape, there is provided converge in a kind of multi- source Remote Sensing Data data source based on scale effect
The analysis method and system of landscape.
According to the wherein one side of the present invention, the present invention is its technical problem of solution, there is provided one kind is based on scale effect
Multi- source Remote Sensing Data data source converge landscape analysis method, comprise the following steps：
S1, satellite-based remote sensing image data are classified to obtain ground mulching distribution map to ground mulching, based on institute
The remote sensing aerosol optical depth data of satellite are stated to calculate air haze concentration profile；
S2, under multiple different scale-values, ground mulching distribution map and air haze concentration profile are divided into respectively
Grid, when dividing every time, ground mulching distribution map and air haze concentration distribution use identical yardstick；
S3, calculate overall landscape heterogeneity index respectively in each grid of ground mulching distribution map, and calculate respectively
The concentration average of each grid of air haze concentration profile, on each yardstick respectively by overall landscape heterogeneity index with it is dense
Spend average and carry out correlation analysis, select correlation highest yardstick as optimal scale；
S4, in optimal scale, calculate each grid of ground mulching distribution map in classification level landscape index, then by class
Other level landscape index carries out correlation analysis with air haze concentration, obtains the source remittance landscape classification of air haze concentration；
Correlation between the classification level landscape index of source remittance landscape classification between S5, analysis grid, is selected
The minimum classification level landscape index of correlation, Geographical Weighted Regression analysis is carried out with air haze concentration, obtains the institute of each grid
The Geographical Weighted Regression coefficient of analysis of source remittance landscape classification is stated, so as to obtain the analysis of the Geographical Weighted Regression of ground mulching distribution map
Coefficient is distributed.
In the analysis method that landscape is converged in the multi- source Remote Sensing Data data source based on scale effect of the present invention, point in step S1
Class refers to be classified according to building, forest land, waters, shrub and arable land.
In the analysis method that landscape is converged in the multi- source Remote Sensing Data data source based on scale effect of the present invention, remote sensing in step S1
Image data derives from satellite Landsat 8, and remote sensing aerosol optical depth data are MODIS aerosol optical depth products
MOD04 data.
In the analysis method that landscape is converged in the multi- source Remote Sensing Data data source based on scale effect of the present invention, entering in step S3
The overall landscape heterogeneity index selected during row choice of optimal scale is patch density, contagion index, Shannon diversity index
And Shannon evenness index.
It is more described in step S2 in the analysis method that landscape is converged in the multi- source Remote Sensing Data data source based on scale effect of the present invention
Individual different scale-value be 2000 meters, 3000 meters, 4000 meters, 5000m, 6000 meters, 7000 meters, 8000 meters, 9000 meters.
In the analysis method that landscape is converged in the multi- source Remote Sensing Data data source based on scale effect of the present invention, carried out in step S5
When correlation analysis differentiates to carry out source remittance landscape classification, selected classification level landscape index is plaque type area percentage
Than, patch density, fractal dimension, patch conjugation and maximum plaque index.
In the analysis method that landscape is converged in the multi- source Remote Sensing Data data source based on scale effect of the present invention, selected in step S5
The classification level landscape index of the correlation minimum gone out is plaque type area percentage and patch density.
In the analysis method that landscape is converged in the multi- source Remote Sensing Data data source based on scale effect of the present invention, selected in step S5
Go out the minimum classification level landscape index of correlation, carrying out Geographical Weighted Regression analysis with air haze concentration is specially：
Geographical weight back is carried out on optimal scale with air haze concentration with plaque type area percentage and patch density
Return analysis.
According to another aspect of the present invention, the present invention additionally provides one kind and is based on scale effect to solve its technical problem
Multi- source Remote Sensing Data data source converge landscape analysis system, comprising：
Distribution map acquisition module, ground mulching is classified to obtain earth's surface for satellite-based remote sensing image data and covered
Lid distribution map, air haze concentration profile is calculated based on the remote sensing aerosol optical depth data of the satellite；
Mesh generation module, it is respectively that ground mulching distribution map and air haze is dense under multiple different scale-values
Degree distribution map is divided into grid, and when dividing every time, ground mulching distribution map and air haze concentration distribution use identical yardstick；
Optimal scale computing module, it is heterogeneous for calculating overall landscape respectively in each grid of ground mulching distribution map
Sex index, and the concentration average of each grid of air haze concentration profile is calculated respectively, respectively will be overall on each yardstick
Landscape heterogeneity index carries out correlation analysis with concentration average, selects correlation highest yardstick as optimal scale；
Converge landscape classification acquisition module in source, in optimal scale, calculating in each grid of ground mulching distribution map
Classification level landscape index, classification level landscape index and air haze concentration are then subjected to correlation analysis, obtain air haze concentration
Source converge landscape classification；
Coefficient be distributed generation module, for analyze the source between grid converge landscape classification classification level landscape index it
Between correlation, select the minimum classification level landscape index of correlation, Geographical Weighted Regression analysis carried out with air haze concentration,
The Geographical Weighted Regression coefficient of analysis of the source remittance landscape classification of each grid is obtained, so as to obtain ground mulching distribution map
Geographical Weighted Regression coefficient of analysis is distributed.
Implement the present invention multi- source Remote Sensing Data data source based on scale effect converge landscape analysis method and system first, base
Ground mulching classification results are obtained in the remote sensing image of satellite, the overall landscape heterogeneity for calculating ground mulching under different scale refers to
Number, select optimal scale of the heterogeneous maximum yardstick as source remittance landscape Analysis；On this basis, with the gas of MODIS data
Measurement of the colloidal sol optical thickness AOD products as air haze pollution level, based on the less classification level landscape index of autocorrelation,
The landscape that converged using Geographical Weighted Regression analysis model to source carries out local regression analysis with AOD, and the geography for obtaining source remittance landscape adds
Weigh the distribution of regression analysis coefficient.The present invention can to relevant position the air haze of " source " " remittance " landscape analyze, draw
The air haze situation of this area, help preferably to carry out the allocation plan and transformation in area.
Brief description of the drawings
Below in conjunction with drawings and Examples, the invention will be further described, in accompanying drawing：
Fig. 1 is the flow chart of the analysis method of the remittance of the multi- source Remote Sensing Data data source based on the scale effect landscape of the present invention；
Fig. 2 is Wuhan City's landscape classification result figure；
Fig. 3 is Wuhan City's AOD distribution maps；
Fig. 4 is the heterogeneity index after being normalized under different scale；
Fig. 5 (a), Fig. 5 (b), Fig. 5 (c) are common field and vinyl house field, paddy field and dry land, growth period and harvesting respectively
Phase field schematic diagram；
Fig. 6 (a), Fig. 6 (b) are building classification PLAND and PD exponential distribution figure respectively；
Fig. 7 (a), Fig. 7 (b) be respectively PD regression coefficients in building classification Geographical Weighted Regression analysis result and
PLAND regression coefficient figures；
Fig. 8 is the distribution of three kinds of urban function regions；
Fig. 9 (a), Fig. 9 (b) shrub classification PLAND and PD exponential distributions；
Figure 10 (a), Figure 10 (b) be respectively PD regression coefficients in shrub classification Geographical Weighted Regression analysis result and
PLAND regression coefficient figures；
Figure 11 (a), Figure 11 (b) are forest land classification PLAND and PD exponential distribution respectively；
PD regression coefficients and PLAND in Figure 12 (a), Figure 12 (b) difference forest land classification Geographical Weighted Regression analysis result
Regression coefficient figure.
Embodiment
In order to which technical characteristic, purpose and the effect of the present invention is more clearly understood, now compares accompanying drawing and describe in detail
The embodiment of the present invention.
The present invention is by taking Wuhan City as an example, based on two kinds of remotely-sensed datas of Landsat8 and MODIS aerosol optical depths product,
City " source " " remittance " landscape pattern and the dependency relation of air haze pollution are analyzed using Geographical Weighted Regression Model, and in industry
Difference of this compound earth's surface key element of building to air haze pollution effect is analyzed in area, shopping centre and 3 kinds of residential block functional areas,
To for the making rational planning for of city " source " " remittance " landscape pattern, mitigate the pollution of air haze, improve city atmospheric environment theory is provided
With reference to.
With reference to figure 1, it is the stream of the analysis method of the remittance of the multi- source Remote Sensing Data data source based on the scale effect landscape of the present invention
Cheng Tu, the idiographic flow of this analysis method are as follows.
S1, satellite-based remote sensing image data are classified to obtain ground mulching distribution map to ground mulching, based on institute
The remote sensing aerosol optical depth data of satellite are stated to calculate air haze concentration profile
Ground mulching data are that the image classifications of Landsat 8 post-process data, including building, forest land, waters, arable land, filling
The atural object classification such as wood, as shown in Figure 2.
Aerosol optical depth data use MODIS aerosol optical depth product MOD04, by geometric correction, inlay,
The processing such as cutting, band math, obtain Wuhan City's Mean aerosol optical thickness of year in 2015 (reject extreme weather, it is such as overcast and rainy,
Strong wind, sandstorm etc.), as shown in Figure 3.
AOD average and standard deviation are as shown in table 1 corresponding to every kind of Land cover types.
AOD average and standard deviation corresponding to 1 every kind of Land cover types of table
S2, under multiple different scale-values, ground mulching distribution map and air haze concentration profile are divided into respectively
Grid, when dividing every time, ground mulching distribution map and air haze concentration distribution use identical yardstick.
S3, calculate overall landscape heterogeneity index respectively in each grid of ground mulching distribution map, and calculate respectively
The concentration average of each grid of air haze concentration profile, on each yardstick respectively by overall landscape heterogeneity index with it is dense
Spend average and carry out correlation analysis, select correlation highest yardstick as optimal scale.
The arrangement spatially of the different landscape feature of size and shape and combination form landscape pattern, including scape in earth's surface
See type, number and the spatial distribution configuration of component units.Landscape index is highly concentrated its structure of landscape pattern's message reflection
The simple quantitative index of composition and some aspect features of space configuration, it is adapted between quantitative expression landscape pattern and ecological process
The spacial analytical method of association.
The calculating of landscape index can be carried out on 3 patch, classification and overall landscape levels.Wherein on patch level
Landscape index yardstick is too small, does not have very big value to describing overall landscape pattern；Class hierarchy index can accordingly calculate one
A little statistics indexs；On overall stratification of landscapes, various diversity indices and concentration class index etc. can be calculated to reflect overall scape
The heterogeneous degree of sight.
Landscape pattern has scale effect.Scale effect is shown as, the result of calculation of landscape index with space scale change
Change and change.The size selection of yardstick determines the result of study of landscape pattern, and Research scale is excessive or too small can all reduce scape
See and the correlation of environment and explanatory.The heterogeneity of landscape determines the research importance of Spatial Patterns of Landscapes.For air
It is heterogeneous better in certain survey region for environment, show that each landscape types distribution is more uniform, be more advantageous to air
The flowing and propagation of middle material, energy etc., so as to keep excellent atmospheric environment.Therefore it is overall under present invention analysis different scale
The heterogeneous degree of landscape, heterogeneous highest yardstick is selected to carry out the Brownish haze effect study of landscape pattern.
S4, in optimal scale, calculate each grid of ground mulching distribution map in classification level landscape index, then by class
Other level landscape index carries out correlation analysis with air haze concentration, obtains the source remittance landscape classification of air haze concentration.
Specifically, overall landscape heterogeneity feature can be reflected from different perspectives, be easy to landscape pattern to analyze comprehensively by choosing
Landscape index carry out Optimal Scaling Technique (such as table 2), 2~9 km space scales calculate landscape index (landscape heterogeneity refers to
Number).
Landscape index and meaning selected by table 2
Each landscape index linear normalization of overall landscape heterogeneity will be reflected, final meter of each index under different scale
It is as shown in Figure 4 to calculate result.
From fig. 4, it can be seen that the heterogeneity index of the overall landscape after normalization reaches maximum except SHEI at 5km
Outside, other indexes reach maximum at 6km.In general, for each index in 6km, the Heterogeneity of overall landscape is most
To be obvious, therefore the present invention studies the dependency relation of " source " " remittance " landscape pattern and AOD based on 6km yardsticks.
Classification level landscape index can reflect the arranged distribution feature of all kinds of landscapes, and the present invention is based on classification level landscape index
To distinguish " source " " remittance " landscape of air haze pollution.Ground mulching classification results are divided into 6km grid, calculate each grid
Interior classification level landscape index；Same segmentation is carried out to AOD, takes central value of the AOD averages of each segmentation block as grid,
And to the two progress correlation analysis, to differentiate " source " " remittance " landscape of air haze pollution.Selected classification level landscape index
As shown in table 3.Landscape index of all categories and AOD coefficient correlation are as shown in table 4.
Landscape index and meaning selected by table 3
The correlation analysis result of each landscape index of the classification level of table 4 and AOD1)
* represents significantly correlated in 0.01 level
Table 4 as can be seen that PLAND, LPI, FRAC_MN, COHESION of building is proportionate with AOD, and PD with
Its is negatively correlated.Building occupied area is bigger i.e. in some scale, single plaque area is bigger, geometry is more regular, spot
More compact between block, AOD values are higher.It is therefore contemplated that " source " landscape of air haze pollution is building.And for forest land
And shrub, each index are negatively correlated with AOD.I.e. shrub and the forest land occupied area in some scale are bigger, connectivity is better,
Distribution is more uniform, and AOD values are lower.It is therefore contemplated that " remittance " landscape of air haze pollution is forest land and shrub.
And in the environmental effect research of landscape pattern, be counted as " converging " always landscape waters but with AOD without obvious
Correlation.Pass through analysis, it may be possible to caused by following two reasons：1) figure it is seen that the Changjiang river overhead aerosol concentration very
Height, this is due to that ship of the road through Wuhan City is more, and the waste gas that its fossil fuel consumed is released is easy in relative humidity very
High river surface, which reacts, generates aerosol[；2) the reason for Wuhan City's weather conditions, Wuhan is throughout the year high, near in relative humidity
The low state of ground wind speed.Higher relative humidity, which turns into, to be condensed and forms " catalyst " of aerosol；Relatively low near-earth wind speed is not
Beneficial to the diffusion of aerosol precursor.
For arable land, due to its wide variety：Common field and the field [Fig. 5 (a)] for having vinyl house, dry land and paddy field
Arable land [Fig. 5 (c)] of [Fig. 5 (b)], growth period and harvest time etc., difference is very big in the effect polluted to air haze, and more
Kind farmland mixing cross-distribution, this is and the unconspicuous main cause of AOD correlations.
In summary, " source " landscape that present invention selection building pollutes for air haze, shrub and forest land are as air haze
Pollute for " remittance " landscape, the Geographical Weighted Regression for carrying out " source " " remittance " landscape and AOD is analyzed, and reflects " source " " remittance " landscape pattern pair
The influence of air haze pollution.
Correlation between the classification level landscape index of source remittance landscape classification between S5, analysis grid, is selected
The minimum classification level landscape index of correlation, Geographical Weighted Regression analysis is carried out with air haze concentration, obtains the institute of each grid
The Geographical Weighted Regression coefficient of analysis of source remittance landscape classification is stated, so as to obtain the analysis of the Geographical Weighted Regression of ground mulching distribution map
Coefficient is distributed.
Air haze effect quantitative analysis to landscape pattern often use based on statistical model, as correlation analysis,
Linear regression model (LRM), exponential model etc., it is world model, actually air haze pollution often shows part with landscape pattern
Variation characteristic, even same landscape types, the haze order of severity corresponding to its alignment placement difference would also vary from, therefore
The coupled relation of landscape pattern and atmospheric environment can more accurately be reflected using partial model.Geographical Weighted Regression
(geographically weighted regression) is used as a kind of local statistical models, can avoid being drawn by geographical position
The influence of the Space atmosphere risen, therefore the present invention chooses Geographical Weighted Regression Model, to dependent variable AOD and each scape of independent variable
See index and carry out regression analysis.
Geographical Weighted Regression Model is the extension to normal linear regression model, and the geographical position of data is embedded into recurrence
In parameter, i.e.,：
In formula, (ui,vi) be ith sample point coordinate (such as longitude and latitude), β0(μi,νi) returned often for ith sample point
Number, βk(μi,νi) for k-th of regression coefficient on ith sample point, it is the function in geographical position, εiFor stochastic error, xik
For the qualitative effect factor, i.e. independent variable.
The present invention carries out the air haze effect analysis of " source " " remittance " landscape using Geographical Weighted Regression Model, with " classification level "
Landscape index is independent variable, and AOD is dependent variable, and the harmony that analysis " source " " remittance " landscape is distributed is reasonable with permutation and combination method
Property to air haze pollution influence.
Very big correlation between landscape index be present, this allows for going out when carrying out Geographical Weighted Regression analysis with AOD
Existing independent variable Combinational redundancy phenomenon, causes model unstable, unreliable.Therefore, the present invention analyzes each landscape under 6km yardsticks and referred to
Correlation (such as table 5) between number, the less landscape index of correlation is taken to carry out Geographical Weighted Regression analysis.
The autocorrelation analytical table of each landscape index on the landscape class hierarchies of table 51)
Table5 Self correlation of landscape index in landscape class
* represents significantly correlated in 0.01 level
Table 5 is shown, in each classification, PD and PLAND, LPI non-correlation, but PLAND and LPI correlations are obvious, and
PLAND meanings are more directly perceived, therefore choose PLAND and PD here and carry out Geographical Weighted Regression analysis to landscape of all categories.
PLAND the and PD exponential distributions of building area are as shown in fig. 6, carrying out the degree of fitting of Geographical Weighted Regression analysis to it and being
0.7521, shown in concrete analysis result such as Fig. 7 (a), 7 (b).
Fig. 7 (a) is shown, is negative generally in main city zone patch density PD regression coefficient, and is in non-main city zone coefficient
Just.Fig. 7 (b) shows PLAND coefficients almost all for just.This is due to occur in main city zone building area in large stretch of patch, although
The patch density of building is small, but its proportion in landscape pattern is very big；And in non-main city zone, building and other classifications
Cross arrangement, the increase of its patch density are mixed, building occupied area has reduced.It can thus be concluded that reduce shared by the floor area of building
Ratio, increasing its patch density plays the role of to reduce AOD, i.e., building type plaque rupture degree is lower, accounts for area percentage
Than bigger, AOD concentration is higher.
Influence of the terrain features such as building to atmosphere pollution has certain difference, i.e. discharge is in non-thread with earth's surface key element
Sexual intercourse, is equally building, has significant difference in industrial area, shopping centre and residential block, will in order to probe into compound earth's surface
Influence of the element to the pollution of air haze, the present invention divide 7 kinds of cities such as industrial area, residential block, shopping centre, waters in 6km grids
Functional areas (such as Fig. 8), difference of the building to atmospheric pollution effect is analyzed in industrial area, shopping centre and 3 kinds of residential block functional areas
It is different.And it is as shown in table 6 to count PD, PLAND and AOD average and standard deviation corresponding to three kinds of functional areas respectively.
PD, PLAND and AOD average and standard deviation corresponding to 6 three kinds of urban function regions of table
Table 6 has confirmed reduction floor area of building proportion, and AOD conclusion can be reduced by increasing its patch density, while
Show, same building that there is the characteristics of significant difference to the pollution of air haze in industrial area, shopping centre and residential block.It is right
For Wuhan City, the real source of air haze pollution is shopping centre and residential block, i.e., from civil plantation.And industrial area
Disposal of pollutants supervision is constantly strengthened causing the major source that industrial area has no longer been the pollution of air haze.
The PLAND and PD in shrub and forest land distribution map carry out Geographical Weighted Regression point respectively as shown in Fig. 9,11, to it
The degree of fitting of analysis is 0.7708 and 0.6997, and concrete analysis result is as shown in Figure 10,12.
Figure 10 (a), 10 (b) display, area but proportion minimum very big in the patch density of main city zone shrub, now
PLAND accounts for leading role, rather than main city zone, and the area proportion of shrub distribution is suitable, and now PD accounts for leading role, generally
PD and PLAND regression coefficient is negative value, it can thus be concluded that, shrub classification has the function that to reduce to AOD concentration, and it is distributed
The uniformity is higher, it is bigger to account for area percentage, and it is more obvious that it reduces the effect of AOD concentration.
Figure 11 shows the lack of uniformity of forest land distribution, Figure 12 (a) displays, except at arable land, other classifications area is returned
It is negative value to return coefficient；Figure 12 (b) is as can be seen that its absolute coefficient of region big forest land PLAND is bigger, i.e., to airborne particulate
The reduction ability of thing is stronger.Aggregate analysis, effect of the forest land to air is almost identical with shrub, that is, increases its distribution area and divide
The cloth uniformity can effectively reduce atmosphere particle concentration.
Embodiments of the invention are described above in conjunction with accompanying drawing, but the invention is not limited in above-mentioned specific
Embodiment, above-mentioned embodiment is only schematical, rather than restricted, one of ordinary skill in the art
Under the enlightenment of the present invention, in the case of present inventive concept and scope of the claimed protection is not departed from, it can also make a lot
Form, these are belonged within the protection of the present invention.
Claims (9)
1. a kind of analysis method of the multi- source Remote Sensing Data data source remittance landscape based on scale effect, it is characterised in that include following step
Suddenly：
S1, satellite-based remote sensing image data are classified to obtain ground mulching distribution map to ground mulching, are defended based on described
The remote sensing aerosol optical depth data of star calculate air haze concentration profile；
S2, under multiple different scale-values, ground mulching distribution map and air haze concentration profile are divided into grid respectively,
Every time during division, ground mulching distribution map and air haze concentration distribution use identical yardstick；
S3, calculate overall landscape heterogeneity index respectively in each grid of ground mulching distribution map, and calculate air respectively
The concentration average of each grid of haze concentration profile, it is respectively that overall landscape heterogeneity index and concentration is equal on each yardstick
Value carries out correlation analysis, selects correlation highest yardstick as optimal scale；
S4, in optimal scale, calculate each grid of ground mulching distribution map in classification level landscape index, then by classification level
Landscape index carries out correlation analysis with air haze concentration, obtains the source remittance landscape classification of air haze concentration；
Correlation between the classification level landscape index of source remittance landscape classification between S5, analysis grid, selects correlation
Property minimum classification level landscape index, carry out Geographical Weighted Regression analysis with air haze concentration, obtain the source of each grid
The Geographical Weighted Regression coefficient of analysis of remittance landscape classification, so as to obtain the Geographical Weighted Regression coefficient of analysis of ground mulching distribution map
Distribution.
2. analysis method according to claim 1, it is characterised in that the classification in step S1 refer to according to building, forest land,
Waters, shrub and arable land are classified.
3. analysis method according to claim 1, it is characterised in that remote sensing image data derives from satellite in step S1
Landsat 8, remote sensing aerosol optical depth data are MODIS aerosol optical depth products MOD04 data.
4. analysis method according to claim 1, it is characterised in that selected in step S3 when carrying out choice of optimal scale
Overall landscape heterogeneity index be patch density, contagion index, Shannon diversity index and Shannon evenness index.
5. analysis method according to claim 1, it is characterised in that the multiple different scale-values of step S2 are 2000
Rice, 3000 meters, 4000 meters, 5000m, 6000 meters, 7000 meters, 8000 meters, 9000 meters.
6. analysis method according to claim 1, it is characterised in that correlation analysis is carried out in the step S5 to carry out
When source remittance landscape classification differentiates, selected classification level landscape index is plaque type area percentage, patch density, point dimension
Number, patch conjugation and maximum plaque index.
7. analysis method according to claim 1, it is characterised in that the correlation minimum selected in the step S5
Classification level landscape index is plaque type area percentage and patch density.
8. analysis method according to claim 7, it is characterised in that the minimum class of correlation is selected in the step S5
Other level landscape index, carrying out Geographical Weighted Regression analysis with air haze concentration is specially：
Geographical Weighted Regression point is carried out on optimal scale with air haze concentration with plaque type area percentage and patch density
Analysis.
9. a kind of analysis system of the multi- source Remote Sensing Data data source remittance landscape based on scale effect, it is characterised in that include：
Distribution map acquisition module, ground mulching is classified for satellite-based remote sensing image data to obtain ground mulching point
Butut, air haze concentration profile is calculated based on the remote sensing aerosol optical depth data of the satellite；
Mesh generation module, under multiple different scale-values, respectively dividing ground mulching distribution map and air haze concentration
Butut is divided into grid, and when dividing every time, ground mulching distribution map and air haze concentration distribution use identical yardstick；
Optimal scale computing module, refer to for calculating overall landscape heterogeneity respectively in each grid of ground mulching distribution map
Number, and the concentration average of each grid of air haze concentration profile is calculated respectively, respectively by overall landscape on each yardstick
Heterogeneity index carries out correlation analysis with concentration average, selects correlation highest yardstick as optimal scale；
Converge landscape classification acquisition module in source, in optimal scale, calculating the classification in each grid of ground mulching distribution map
Level landscape index, then carries out correlation analysis by classification level landscape index and air haze concentration, obtains the source of air haze concentration
Remittance landscape classification；
Coefficient is distributed generation module, between the classification level landscape index for analyzing the remittance of the source between grid landscape classification
Correlation, selects the minimum classification level landscape index of correlation, carries out Geographical Weighted Regression analysis with air haze concentration, obtains
The Geographical Weighted Regression coefficient of analysis of the source remittance landscape classification of each grid, so as to obtain the geography of ground mulching distribution map
Weighted regression analysis coefficient is distributed.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710899736.0A CN107766799B (en) | 2017-09-28 | 2017-09-28 | Method and system for analyzing source-sink landscape of multi-source remote sensing data based on scale effect |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710899736.0A CN107766799B (en) | 2017-09-28 | 2017-09-28 | Method and system for analyzing source-sink landscape of multi-source remote sensing data based on scale effect |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107766799A true CN107766799A (en) | 2018-03-06 |
| CN107766799B CN107766799B (en) | 2019-12-17 |
Family
ID=61267305
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710899736.0A Active CN107766799B (en) | 2017-09-28 | 2017-09-28 | Method and system for analyzing source-sink landscape of multi-source remote sensing data based on scale effect |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107766799B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111650102A (en) * | 2020-05-26 | 2020-09-11 | 北京中科锐景科技有限公司 | Haze pollution analysis method, device, medium and equipment based on satellite data |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103413293A (en) * | 2013-07-09 | 2013-11-27 | 武汉理工大学 | Method for obtaining urban land use landscape pattern intrinsic observation scale |
| CN104103016A (en) * | 2014-05-13 | 2014-10-15 | 杭州师范大学 | Comprehensive evaluation method for ecosystem health of wetland on the basis of remote sensing technology |
| KR20150106212A (en) * | 2014-03-11 | 2015-09-21 | 서울대학교산학협력단 | Real Time Haze Readout System Using Skyradiometer And Method Thereof |
| CN106021872A (en) * | 2016-05-10 | 2016-10-12 | 浙江大学 | Dynamic filtering modeling downscaling method of environment variable on the basis of low-resolution satellite remote sensing data |
| CN106290090A (en) * | 2015-05-14 | 2017-01-04 | 株式会社日立制作所 | The method and apparatus of assessment particle concentration |
| CN106650015A (en) * | 2016-11-24 | 2017-05-10 | 中国科学院东北地理与农业生态研究所 | Landscape scale deduction method of urban forest leaf area index |
| CN107202750A (en) * | 2017-05-17 | 2017-09-26 | 河北中科遥感信息技术有限公司 | A kind of satellite-ground integrated monitoring quantitative remote sensing method for amalgamation processing of Atmospheric particulates |
-
2017
- 2017-09-28 CN CN201710899736.0A patent/CN107766799B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103413293A (en) * | 2013-07-09 | 2013-11-27 | 武汉理工大学 | Method for obtaining urban land use landscape pattern intrinsic observation scale |
| KR20150106212A (en) * | 2014-03-11 | 2015-09-21 | 서울대학교산학협력단 | Real Time Haze Readout System Using Skyradiometer And Method Thereof |
| CN104103016A (en) * | 2014-05-13 | 2014-10-15 | 杭州师范大学 | Comprehensive evaluation method for ecosystem health of wetland on the basis of remote sensing technology |
| CN106290090A (en) * | 2015-05-14 | 2017-01-04 | 株式会社日立制作所 | The method and apparatus of assessment particle concentration |
| CN106021872A (en) * | 2016-05-10 | 2016-10-12 | 浙江大学 | Dynamic filtering modeling downscaling method of environment variable on the basis of low-resolution satellite remote sensing data |
| CN106650015A (en) * | 2016-11-24 | 2017-05-10 | 中国科学院东北地理与农业生态研究所 | Landscape scale deduction method of urban forest leaf area index |
| CN107202750A (en) * | 2017-05-17 | 2017-09-26 | 河北中科遥感信息技术有限公司 | A kind of satellite-ground integrated monitoring quantitative remote sensing method for amalgamation processing of Atmospheric particulates |
Non-Patent Citations (2)
| Title |
|---|
| 张明明: "大气气溶胶光学厚度遥感反演及其环境效应研究", 《中国优秀硕士学位论文全文数据库 工程科技I辑》 * |
| 赵磊: "基于多源遥感数据的区域景观格局尺度效应", 《遥感应用》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111650102A (en) * | 2020-05-26 | 2020-09-11 | 北京中科锐景科技有限公司 | Haze pollution analysis method, device, medium and equipment based on satellite data |
| CN111650102B (en) * | 2020-05-26 | 2023-08-29 | 北京中科锐景科技有限公司 | Haze pollution analysis method, device, medium and equipment based on satellite data |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107766799B (en) | 2019-12-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109001091B (en) | Satellite remote-sensing monitoring method, device and the computer-readable medium of atmosphere pollution | |
| CN103413293B (en) | The acquisition methods of urban land use landscape pattern intrinsic observation scale | |
| Balázs et al. | Simulation of the mean urban heat island using 2D surface parameters: empirical modelling, verification and extension | |
| Millward-Hopkins et al. | Aerodynamic parameters of a UK city derived from morphological data | |
| Bi et al. | A new framework for analysis of the morphological spatial patterns of urban green space to reduce PM2. 5 pollution: A case study in Wuhan, China | |
| CN106557869A (en) | A kind of air pollutant emission inventory space allocation method and device based on POI points | |
| CN106372807A (en) | County region based resource environment carrying capacity evaluation method | |
| Li et al. | Wind tunnel study on the morphological parameterization of building non-uniformity | |
| Zhang et al. | Influence of the spatial distribution pattern of buildings on the distribution of PM2. 5 concentration | |
| Burian et al. | Morphological analyses using 3D building databases: Los Angeles, California | |
| CN106546958B (en) | A kind of radar data assimilation method of optimization | |
| CN107766799A (en) | The analysis method and system of multi- source Remote Sensing Data data source remittance landscape based on scale effect | |
| Heath et al. | Modelling the behaviour of nutrients in the coastal waters of Scotland | |
| CN112784395B (en) | Method for predicting and simulating total phosphorus concentration of river water body | |
| Burian et al. | Morphological analyses using 3D building databases: Oklahoma City, Oklahoma | |
| Burian et al. | High-resolution dataset of urban canopy parameters for Houston, Texas | |
| CN116882741A (en) | Method for dynamically and quantitatively evaluating super-standard flood disasters | |
| CN116757357A (en) | Land ecological condition assessment method coupled with multisource remote sensing information | |
| Fang et al. | Exploring coupling effect between urban heat island effect and PM 2.5 concentrations from the perspective of spatial environment | |
| CN113361114B (en) | Multi-scale non-point source pollutant river entering coefficient measuring and calculating method based on runoff path | |
| Li et al. | Spatial heterogeneity of the effects of human activities on ecosystem health of a coastal tourism city—A case study of Rizhao, China | |
| CN103617452A (en) | Photometry network layout method in large-scale photovoltaic base area | |
| Pan et al. | ESTIMATION OF ECOLOGICAL ASSET VALUES IN SHANGRILA BASED ON REMOTELY SENSED DATA. | |
| Burian et al. | Morphological analyses using 3d building databases: salt lake city, utah | |
| Liu et al. | GIS-based urban land development intensity impact factors analysis |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |