CN115618174B - Soil humidity inversion method based on pixel scale surface roughness spectrum parameters - Google Patents

Soil humidity inversion method based on pixel scale surface roughness spectrum parameters Download PDF

Info

Publication number
CN115618174B
CN115618174B CN202211304708.7A CN202211304708A CN115618174B CN 115618174 B CN115618174 B CN 115618174B CN 202211304708 A CN202211304708 A CN 202211304708A CN 115618174 B CN115618174 B CN 115618174B
Authority
CN
China
Prior art keywords
soil
roughness spectrum
inversion
pixel
coefficient
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
Application number
CN202211304708.7A
Other languages
Chinese (zh)
Other versions
CN115618174A (en
Inventor
汪左
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Anhui Normal University
Original Assignee
Anhui Normal University
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Anhui Normal University filed Critical Anhui Normal University
Priority to CN202211304708.7A priority Critical patent/CN115618174B/en
Publication of CN115618174A publication Critical patent/CN115618174A/en
Application granted granted Critical
Publication of CN115618174B publication Critical patent/CN115618174B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/10—Complex mathematical operations
    • G06F17/15—Correlation function computation including computation of convolution operations
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
    • Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
    • Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Mathematical Physics (AREA)
  • Computational Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Mathematical Analysis (AREA)
  • Mathematical Optimization (AREA)
  • Pure & Applied Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • Algebra (AREA)
  • Databases & Information Systems (AREA)
  • Software Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Computing Systems (AREA)
  • Radar Systems Or Details Thereof (AREA)

Abstract

The invention discloses a soil humidity inversion method based on pixel scale surface roughness spectrum parameters, which comprises the following steps: construction of a vv-polarized backscatter component based on a roughness spectral parameter p
Figure DDA0003905318060000011
With soil moisture M v Combined roughness Z s A relation model I of (2); construction of vh polarized backscatter component based on pixel incidence angle θ
Figure DDA0003905318060000012
With soil moisture M v Combined roughness Z s A relation model II of (2); construction of a back-scattered component based on vv polarization
Figure DDA0003905318060000013
vh polarized backscatter component
Figure DDA0003905318060000014
And an inversion model of the roughness spectrum parameter p of the pixel incidence angle theta; scattering component of soil surface observed by radar
Figure DDA0003905318060000015
And inputting a relation model I, a relation model II and an inversion model of the roughness spectrum parameter p according to the corresponding pixel incidence angle theta, and inverting the soil humidity of the corresponding pixel position. Inversion of soil humidity is performed based on the roughness spectrum parameter p of the pixel scale, the influence of the space heterogeneity and the incidence angle of the roughness spectrum parameter p is fully considered, and further the accuracy and the adaptability of the soil humidity inversion model are improved.

Description

Soil humidity inversion method based on pixel scale surface roughness spectrum parameters
Technical Field
The invention belongs to the technical field of remote sensing, and particularly relates to a soil humidity inversion method based on pixel scale surface roughness spectrum parameters.
Background
Construction of a vv, vh polarized backscatter component
Figure BDA0003905318040000011
Incidence angle θ with pixel, soil humidity M v Combined roughness Z s The combined roughness Z is eliminated by a simultaneous manner as in the relation models of formulas (1) and (2) s Thereby obtaining the soil humidity M v 。
Figure BDA0003905318040000012
Figure BDA0003905318040000013
In the method, in the process of the invention,
Figure BDA0003905318040000014
and->
Figure BDA0003905318040000015
The scattering components of the soil surface under the v and vh polarization are respectively; a is that vv 、B vv 、C vv 、D vv 、A vh 、B vh 、C vh 、D vh Are each related to the incident angle θ.
The influence of the p value of the roughness spectrum parameter on the backscattering coefficient is huge, the influence belongs to a strong sensitive parameter, the p value has spatial heterogeneity, and an existing soil humidity inversion model generally takes a unified p value in a regional scale range, and the influence of the spatial heterogeneity and the incidence angle of the roughness spectrum parameter p is not considered, so that the precision and the adaptability of the soil humidity inversion model are reduced, and the soil humidity high-precision inversion of a refined scale cannot be realized.
Disclosure of Invention
The invention provides a soil humidity inversion method based on pixel scale surface roughness spectrum parameters, and aims to improve the problems.
The invention discloses a soil humidity inversion method based on pixel scale surface roughness spectrum parameters, which specifically comprises the following steps:
s1, constructing a vv polarization backward scattering component based on a roughness spectrum parameter p
Figure BDA0003905318040000021
With soil moisture M v Combined roughness Z s A relation model I of (2);
s2, constructing vh polarized backward scattering component based on pixel incidence angle theta
Figure BDA0003905318040000022
With soil moisture M v Combined roughness Z s A relation model II of (2);
s3, constructing a back scattering component based on v polarization
Figure BDA0003905318040000023
vh polarized backscatter component->
Figure BDA0003905318040000024
And an inversion model of the roughness spectrum parameter p of the pixel incidence angle theta;
s4, scattering components of the soil surface observed by the radar
Figure BDA0003905318040000025
And inputting a relation model I, a relation model II and an inversion model of a roughness spectrum parameter p according to the corresponding pixel incidence angle theta, and inverting the soil humidity of bare soil and vegetation covered earth surface.
Further, the relation model I is specifically as follows:
Figure BDA0003905318040000026
wherein A is vv (θ,p)、B vv (θ, p) and C vv (theta, p) is a coefficient related to the roughness spectrum parameter p, the pixel incidence angle theta;
or is;
Figure BDA0003905318040000027
wherein A is vv (p)、B vv (p) and C vv (p) is a coefficient related to the roughness spectrum parameter p.
Further, A vv (θ,p)、B vv (θ, p) and C vv The expression of (θ, p) is specifically as follows:
A vv (θ,p)＝r 1 ·sin 3 θ·p+s 1 ·sin 2 θ+t 1 ·sinθ+u 1
B vv (θ,p)＝r 2 ·ln(p)+s 2 ·sin 2 θ+t 2 ·sinθ+u 2
C vv (θ,p)＝r 3 ·p 3 ·sin 2 θ+s 3 ·p 2 ·sinθ+t 3 ·p+u 3
wherein r is 1 、s 1 、t 1 、u 1 、r 2 、s 2 、t 2 、u 2 、r 3 、s 3 、t 3 、u 3 Each of which is a coefficient of various types.
Further, coefficient A vv (p)、B vv (p) and C vv The determination method of (p) is specifically as follows:
at different incident angles theta of pixels, respectively, the coefficient A vv (p)、B vv (p) and C vv (p) performing nonlinear regression on the relation between the roughness spectrum parameter p to obtain a coefficient A under different pixel incidence angles theta vv (p)、B vv (p) and C vv And (p) and a roughness spectrum parameter p.
Further, coefficient A vv (p)、B vv (p) and C vv The relation between (p) and the roughness spectrum parameter p is specifically as follows:
A vv (p)＝a 1 ·p 3 +b 1 ·p 2 +c 1 ·p+d 1
B vv (p)＝a 2 ·p 3 +b 2 ·p 2 +c 2 ·p+d 2
C vv (p)＝a 3 ·p 4 +b 3 ·p 3 +c 3 ·p 2 +d 3 ·p+e 3
wherein a is 1 、b 1 、c 1 、d 1 、a 2 、b 2 、c 2 、d 2 、a 3 、b 3 、c 3 、d 3 、e 3 The coefficients of the fitting formulas are respectively used, the incident angles theta of the pixels are different, and the values of the coefficients are different.
Further, the roughness spectrum parameter p is inverted based on the inversion model of the roughness spectrum parameter p, and the inversion model of the roughness spectrum parameter p is specifically as follows:
Figure BDA0003905318040000031
wherein a is 4 、b 4 、c 4 、d 4 Is the fitting coefficient.
Further, the method for constructing the inversion model of the roughness spectrum parameter p specifically comprises the following steps:
(1) Acquiring actual measurement values of soil humidity of a plurality of ground sample points through satellite-ground synchronous observation, and a backscattering coefficient and a pixel incidence angle theta of the radar satellite synchronous observation;
(2) The scattering component of the soil surface obtained by synchronous observation
Figure BDA0003905318040000032
And->
Figure BDA0003905318040000033
Inputting soil humidity inversion models under different p values by using each pixel incidence angle theta to obtain soil humidity values under different p values;
(3) And matching the soil humidity values which are inverted under different p values with the actual measured values of the soil humidity of the corresponding pixels, and finding out the p value corresponding to the inversion value closest to the actual measured value, namely the optimal p value, or taking 2 p values which are most matched with the actual measured value for linear interpolation to obtain the optimal p value.
(4) For the optimal p value
Figure BDA0003905318040000034
Performing regression analysis on the sine value sin theta of the pixel incidence angle to obtain an inversion model of the roughness spectrum parameter p;
further, the expression of the relational model II is specifically as follows:
Figure BDA0003905318040000041
wherein C is vh (θ) is a coefficient related to the incident angle θ of the pixel, A vh 、B vh Fitting coefficients are respectively used. Further, C vh The expression of (θ) is specifically as follows:
C vh (θ)＝a 5 ·cos 2 θ+b 5 ·cosθ+c 5
wherein a is 5 、b 5 、c 5 Is the fitting coefficient.
The soil humidity inversion model consists of an inversion model of a relation model I, a relation model II and a roughness spectrum parameter p.
Further, if the earth surface is bare soil, the backscattering coefficient observed by the radar is the soil surface scattering component
Figure BDA0003905318040000042
And
Figure BDA0003905318040000043
if the ground surface is covered by vegetation, the backscattering coefficient observed by radar is the backscattering coefficient of the canopy>
Figure BDA0003905318040000044
And->
Figure BDA0003905318040000045
Carrying out surface body scattering separation to obtain soil surface scattering component->
Figure BDA0003905318040000046
And->
Figure BDA0003905318040000047
According to the invention, the soil humidity inversion is performed based on the roughness spectrum parameter p of the pixel scale, the influence of the space heterogeneity and the incidence angle of the roughness spectrum parameter p is fully considered, the precision and the adaptability of the soil humidity inversion model are further improved, and the soil humidity inversion requirement of the fine scale is met.
Drawings
FIG. 1 shows the vV polarized backscattering coefficient provided by an embodiment of the present invention
Figure BDA0003905318040000048
Relation to the roughness spectral parameter p, wherein (a) is the vv polarized backscattering coefficient at different roughness spectral parameter p values +.>
Figure BDA0003905318040000049
A variation curve according to the incident angle theta of the pixel; (b) vV polarized backscattering coefficient for different roughness spectral parameter p values +.>
Figure BDA00039053180400000410
A change curve according to the soil humidity;
FIG. 2 shows the vh polarized backscattering coefficients at different incident angles θ of the pixels according to the embodiment of the present invention
Figure BDA00039053180400000411
And combined roughness Z s Is a relationship of (2); wherein (a) is vh polarized backscattering coefficient +.>
Figure BDA00039053180400000412
And combined roughness Z s Is a relationship of (2); (b) Vh polarized backscattering coefficient +.>
Figure BDA00039053180400000413
And combined roughness Z s Is a relationship of (2); (c) Vh polarized backscattering coefficient +.>
Figure BDA00039053180400000414
And combined roughness Z s Is a relationship of (2); (d) Vh polarized backscattering coefficient +.>
Figure BDA00039053180400000415
And combined roughness Z s Is a relationship of (2);
FIG. 3 shows the vh polarized backscattering coefficients at different incident angles θ of the pixels according to the embodiment of the present invention
Figure BDA0003905318040000051
With soil moisture M v Is a relationship of (2);
FIG. 4 is a flow chart of modeling an inversion model of a roughness spectrum parameter p value provided by an embodiment of the invention;
FIG. 5 shows the accuracy of the p-value inversion model provided by the embodiment of the invention;
fig. 6 is a graph of soil moisture inversion accuracy verification results provided by the embodiment of the invention.
Detailed Description
The following detailed description of the embodiments of the invention, given by way of example only, is presented in the accompanying drawings to aid in a more complete, accurate, and thorough understanding of the inventive concepts and aspects of the invention by those skilled in the art.
The influence of the p-value of the roughness spectral parameter on the backscattering coefficient is great, belonging to the strongly sensitive parameter (as can be seen from fig. 1). As can be seen from fig. 1 (a), in the case where the soil humidity, the root mean square height and the correlation length are the same, the backscattering coefficient corresponding to the same incident angle gradually decreases as the p value increases, and the larger the incident angle is, the larger the decreasing amplitude is, and when p increases from 1.2 to 3, the difference in backscattering coefficient at the incident angle of 33 ° is 13.12dB, and the difference at the incident angle of 44 ° can reach 18.64dB. As can be seen from fig. 1 (b), when the incident angle, the root mean square height and the correlation length are the same, if p is different, the soil humidity corresponding to the same backscatter coefficient is greatly different.
Accordingly, in view of the above problems, the present invention proposes: (1) A soil humidity inversion model taking the roughness spectrum parameter p into consideration; (2) Aiming at the problem that the regional scale unified p value ignores the influence of the spatial heterogeneity and the incidence angle, a roughness spectrum parameter p inversion model of a pixel scale is established; (3) And (3) combining the step (1) and the step (2) to completely construct a soil humidity inversion model based on the pixel scale surface roughness spectrum parameters.
(1) Soil humidity inversion model considering roughness spectrum parameter p
Analysis of an Oh model simulation database shows that the vh polarized backscattering coefficient under the condition of different pixel incidence angles theta
Figure BDA0003905318040000061
With soil moisture M v Combined roughness Z of earth's surface s Only the curves of the relationship are represented by the translation up and down (see fig. 2 and 3), only the intercept of the curves is changed. Thus construct a +.>
Figure BDA0003905318040000062
And M is as follows v And Z s As shown in equation 3:
Figure BDA0003905318040000063
in the formula, the coefficient A vh 、B vh Independent of the angle of incidence θ, coefficient C vh As a function of the angle of incidence θ. Obtained by least squares fitting:
A vh ＝3.040，B vh ＝3.255，C vh (θ)＝-7.2358cos 2 θ+23.633cosθ-20.08 (4)
since the value of the roughness spectrum parameter p has a great influence on the backscattering coefficient, the roughness spectrum parameter p belongs to a strong sensitivity parameter (as can be seen from fig. 1). At the position of
Figure BDA0003905318040000064
And M is as follows v And Z s The relation of the parameters p is added with the parameters p of the roughness spectrum to construct a parameter p of the roughness spectrum
Figure BDA0003905318040000065
And M is as follows v And Z s See formula (5):
Figure BDA0003905318040000066
wherein A is vv 、B vv 、C vv Is related to the angle of incidence θ, which can be obtained from the radar image, and the roughness spectrum parameter p. Nonlinear regression of the 3 coefficients in equation (5) using data of different incidence angles, different roughness spectral parameters p in the backscatter simulation databaseThe calculation formula of the coefficients can be obtained:
A vv (θ,p)＝0.044sin 3 θ·p-0.080sin 2 θ+0.031sinθ+2.579 (6)
R 2 ＝0.996，RMSE＝0.0004，MRE＝0.014％
B vv (θ,p)＝4.722ln(p)-7.643sin 2 θ+9.243sinθ-0.511 (7)
R 2 ＝0.995，RMSE＝0.0930，MRE＝1.486％
C vv (θ,p)＝0.698p 3 ·sin 2 θ-8.085p 2 ·sinθ+14.079p-6.445 (8)
R 2 ＝0.987，RMSE＝0.6925，MRE＝26.311％
fitting accuracy R of the coefficient fitting formulas 2 Is very high, above 0.98, but still has certain errors, especially C vv The Root Mean Square Error (RMSE) was 0.6925 and the average relative error (MRE) reached 26.311%.
In order to further improve the fitting precision of the improved v polarization relation model coefficient, the coefficient A is fixed by each degree based on the formula (5) because the incident angle theta can be obtained through image preprocessing and belongs to known parameters vv 、B vv 、C vv Non-linear regression is carried out again on the relation between the roughness spectrum parameter p to obtain a new relation model (formula 9) of the vv polarization backscattering coefficient considering the roughness spectrum parameter p, the soil humidity and the combined roughness and a new A vv 、B vv 、C vv Fitting equations (10), (11), (12)) and A at different angles of incidence vv 、B vv 、C vv Coefficient lookup tables (table 1, table 2, table 3).
Figure BDA0003905318040000071
A vv (p)＝a 1 ·p 3 +b 1 ·p 2 +c 1 ·p+d 1 (10)
B vv (p)＝a 2 ·p 3 +b 2 ·p 2 +c 2 ·p+d 2 (11)
C vv (p)＝a 3 ·p 4 +b 3 ·p 3 +c 3 ·p 2 +d 3 ·p+e 3 (12)
Wherein a is 1 、b 1 、c 1 、d 1 、a 2 、b 2 、c 2 、d 2 、a 3 、b 3 、c 3 、d 3 、e 3 The coefficients of the fitting formulas can be found from tables 1, 2 and 3.
As can be seen from tables 1, 2 and 3, the coefficient A of the formula (9) at different incident angles vv The fitting accuracy of (2) is closer to that of equation (5). For coefficient B vv And C vv The accuracy of equation (9) is substantially higher than equation (5). Therefore, the invention selects the formula (9) as the relation model under the v polarization, and can find the corresponding coefficient from the coefficient lookup tables (table 1, table 2 and table 3) according to the specific incidence angle of the radar image in practical application.
To sum up, the above equations (9) and (3) are combined, and the coefficients are fitted to the equations (equation (10), equation (11), equation (12), equation (4)) and the coefficient lookup table (table 1, table 2, table 3), so as to create a soil moisture inversion model considering the roughness spectrum parameter p. But in this model, however, in the present model,
Figure BDA0003905318040000072
and->
Figure BDA0003905318040000073
To know the observed value, the soil humidity M v For the final unknown parameters to be inverted, the surface combined roughness Z s And the roughness spectrum parameter p is an unknown parameter. Two simultaneous equations 3 unknown parameters, the soil moisture M can not be solved v The combination roughness Z can be eliminated by further providing the roughness spectrum parameter p s Further inversion of soil humidity M v 。
TABLE 1 different entriesUnder the firing angle A vv Calculating each coefficient of formula and fitting precision
Figure BDA0003905318040000081
TABLE 2B at different angles of incidence vv Calculating each coefficient of formula and fitting precision
Figure BDA0003905318040000091
TABLE 3C at different angles of incidence vv Calculating each coefficient of formula and fitting precision
Figure BDA0003905318040000101
(2) Inversion model of roughness spectrum parameter p of one pixel scale
Aiming at the problem that the regional scale unified p value ignores the influence of the spatial heterogeneity and the incidence angle, a roughness spectrum parameter p inversion model of a pixel scale is established. I.e. for each picture element in the investigation region, the value of the roughness spectrum parameter p which is most matched with the surface condition of the picture element is found. The specific establishment method is shown in fig. 4, and is described as follows:
(1) And carrying out satellite-ground synchronous observation to obtain actual measurement values of soil humidity of a plurality of ground sample points and a backscattering coefficient and an incident angle of the radar satellite synchronous observation. The greater the number of sample points, the better the invention uses 646 sample pels for modeling here.
(2) The scattering component of the soil surface obtained by synchronous observation
Figure BDA0003905318040000111
And->
Figure BDA0003905318040000112
(if the earth surface is bare soil, the backscattering coefficient observed by the radar is directly the soil surface scattering component +.>
Figure BDA0003905318040000113
And->
Figure BDA0003905318040000114
If the ground surface is covered by vegetation, the backscattering coefficient observed by radar is the backscattering coefficient of the canopy>
Figure BDA0003905318040000115
And->
Figure BDA0003905318040000116
(/>
Figure BDA0003905318040000117
Representing the back scattering coefficient of the v polarized canopy, < ->
Figure BDA0003905318040000118
Representing the vh polarized canopy backscattering coefficient), and carrying out surface body scattering separation on the obtained product by using a water cloud model to obtain a soil surface scattering component +.>
Figure BDA0003905318040000119
And
Figure BDA00039053180400001110
) And inputting the soil humidity inversion model taking the roughness spectrum parameter p into consideration under the assumption of different p values by using the incidence angle theta of each pixel, and obtaining the soil humidity inversion values under the assumption of different p values.
(3) And matching the inversion values of the soil humidity under the assumption of the different p values with the actual measurement values of the soil humidity of the corresponding pixels, and finding out the p value corresponding to the inversion value closest to the actual measurement value, namely the optimal p value. In the actual matching process, since the assumed p values are discrete, the 2 p values that are the best match with the measured values can be taken for linear interpolation, so as to obtain a more accurate optimal p value. In this way, the optimal p-values for all sample pixels (646) are obtained.
(4) For the optimal p value
Figure BDA00039053180400001111
And the sine value sin theta of the incident angle of the pixel, the result (Table 4) shows that the p value is equal to +.>
Figure BDA00039053180400001112
And sin theta are all obviously related on the level of 0.01, and the related coefficients are respectively 0.137, 0.665 and-0.539. Thus, using these 646 pel samples, the optimal p-value is regressed with +.>
Figure BDA00039053180400001113
And sin theta, an inversion model of a roughness spectrum parameter p of a pixel scale is established, and the expression of the inversion model is shown in a formula (13). After the inversion model is established, ground observation data are not needed, and inversion of pixel scale p values is realized.
Table 4 correlation coefficient between p value and soil surface scattering component, incidence angle sine value
Figure BDA0003905318040000121
Note that: * Represent significant correlation at the 0.01 level.
The roughness spectrum parameter p inversion model is specifically as follows:
Figure BDA0003905318040000122
R 2 ＝0.875，RMSE＝0.0828，MAE＝0.0638
the scatter diagram of the p value and the optimal p value of the model inversion is shown in fig. 5, and it can be seen that the p value inversion model has high fitting precision and R 2 0.875, 0.0828 RMSE, 0.0638 MAE
(3) Soil humidity inversion model based on pixel scale surface roughness spectrum parameters
The soil humidity inversion model based on pixel-scale earth surface roughness spectrum parameters is completely constructed by combining the simultaneous formulas (9), (3) and (13) and combining the coefficient fitting formulas (formula (10), formula (11), formula (12) and formula (4)) and the coefficient lookup tables (tables 1, 2 and 3).
Scattering component of soil surface observed by radar
Figure BDA0003905318040000123
And->
Figure BDA0003905318040000124
(if the earth surface is bare soil, the backscattering coefficient observed by the radar is directly the soil surface scattering component +.>
Figure BDA0003905318040000125
And->
Figure BDA0003905318040000126
If the ground surface is covered by vegetation, the backscattering coefficient observed by radar is the canopy backscattering coefficient ++>
Figure BDA0003905318040000127
And->
Figure BDA0003905318040000128
(/>
Figure BDA0003905318040000129
Representing the back scattering coefficient of the v polarized canopy, < ->
Figure BDA00039053180400001210
Representing the vh polarized canopy backscattering coefficient), and carrying out surface body scattering separation on the obtained product by using a water cloud model to obtain a soil surface scattering component +.>
Figure BDA00039053180400001211
And->
Figure BDA00039053180400001212
The surface body scattering separation is divided into conventional operation, which is not repeated, and the incident angle theta of the pixel is input into the modelEliminating combined roughness Z s Thereby inverting to obtain soil humidity M v 。
And for the constructed soil humidity inversion model based on the pixel scale surface roughness spectrum parameters, performing inversion result accuracy verification by using soil humidity actual measurement value samples (213 sample pixels in total) except modeling, wherein the verification result is shown in fig. 6. It can be seen that the scattered points of the inversion value and the measured value are distributed on two sides of a 1:1 line, RMSE is 4.03%, MAE is 3.18%, and the model inversion accuracy is high.
While the present invention has been described by way of example, it should be apparent that the practice of the invention is not limited by the foregoing, but rather is intended to cover various insubstantial modifications of the method concepts and teachings of the invention, either as applied to other applications without modification, or as applied directly to other applications, without departing from the scope of the invention.

Claims (9)

1. The soil humidity inversion method based on the pixel scale surface roughness spectrum parameter is characterized by comprising the following steps of:
s1, constructing a vv polarization backward scattering component based on a roughness spectrum parameter p
Figure FDA0004147921860000011
With soil moisture M v Combined roughness Z s A relation model I of (2);
s2, constructing vh polarized backward scattering component based on pixel incidence angle theta
Figure FDA0004147921860000012
With soil moisture M v Combined roughness Z s A relation model II of (2);
s3, constructing a back scattering component based on v polarization
Figure FDA0004147921860000013
vh polarized backscatter component->
Figure FDA0004147921860000014
And an inversion model of the roughness spectrum parameter p of the pixel incidence angle theta;
s4, scattering components of the soil surface observed by the radar
Figure FDA0004147921860000015
And inputting a relation model I, a relation model II and an inversion model of a roughness spectrum parameter p according to the corresponding pixel incidence angle theta, and inverting the soil humidity of bare soil and vegetation covered earth surface;
the relation model I is specifically as follows:
Figure FDA0004147921860000016
wherein A is vv (θ,p)、B vv (θ, p) and C vv (theta, p) is a coefficient related to the roughness spectrum parameter p, the pixel incidence angle theta;
or is:
Figure FDA0004147921860000017
wherein A is vv (p)、B vv (p) and C vv (p) is a coefficient related to the roughness spectrum parameter p.
2. The soil moisture inversion method based on pixel scale surface roughness spectrum parameters of claim 1, wherein a vv (θ,p)、B vv (θ, p) and C vv The expression of (θ, p) is specifically as follows:
A vv (θ,p)＝r 1 ·sin 3 θ·p+s 1 ·sin 2 θ+t 1 ·sinθ+u 1
B vv (θ,p)＝r 2 ·ln(p)+s 2 ·sin 2 θ+t 2 ·sinθ+u 2
C vv (θ,p)＝r 3 ·p 3 ·sin 2 θ+s 3 ·p 2 ·sinθ+t 3 ·p+u 3
wherein r is 1 、s 1 、t 1 、u 1 、r 2 、s 2 、t 2 、u 2 、r 3 、s 3 、t 3 、u 3 Each of which is a coefficient of various types.
3. The soil moisture inversion method based on pixel scale surface roughness spectrum parameters as claimed in claim 1, wherein the coefficient a vv (p)、B vv (p) and C vv The determination method of (p) is specifically as follows:
at different incident angles theta of pixels, respectively, the coefficient A vv (p)、B vv (p) and C vv (p) performing nonlinear regression on the relation between the roughness spectrum parameter p to obtain a coefficient A under different pixel incidence angles theta vv (p)、B vv (p) and C vv And (p) and a roughness spectrum parameter p.
4. A soil moisture inversion method based on pixel scale surface roughness spectrum parameters as claimed in claim 3 wherein the coefficient a vv (p)、B vv (p) and C vv The relation between (p) and the roughness spectrum parameter p is specifically as follows:
A vv (p)＝a 1 ·p 3 +b 1 ·p 2 +c 1 ·p+d 1
B vv (p)＝a 2 ·p 3 +b 2 ·p 2 +c 2 ·p+d 2
C vv (p)＝a 3 ·p 4 +b 3 ·p 3 +c 3 ·p 2 +d 3 ·p+e 3
wherein a is 1 、b 1 、c 1 、d 1 、a 2 、b 2 、c 2 、d 2 、a 3 、b 3 、c 3 、d 3 、e 3 The coefficients of the fitting formulas are respectively used, the incident angles theta of the pixels are different, and the values of the coefficients are different.
5. The soil moisture inversion method based on pixel scale surface roughness spectrum parameters of claim 1, wherein the roughness spectrum parameters p are inverted based on an inversion model of the roughness spectrum parameters p, and the inversion model of the roughness spectrum parameters p is specifically as follows:
Figure FDA0004147921860000021
wherein a is 4 、b 4 、c 4 、d 4 Is the fitting coefficient.
6. The soil moisture inversion method based on pixel scale surface roughness spectrum parameters of claim 5, wherein the inversion model construction method of the roughness spectrum parameter p is specifically as follows:
(1) Acquiring actual measurement values of soil humidity of a plurality of ground sample points through satellite-ground synchronous observation, and a backscattering coefficient and a pixel incidence angle theta of the radar satellite synchronous observation;
(2) The scattering component of the soil surface obtained by synchronous observation
Figure FDA0004147921860000022
And->
Figure FDA0004147921860000023
Inputting soil humidity inversion models under different p values by using each pixel incidence angle theta to obtain soil humidity values under different p values;
(3) And matching the soil humidity values which are inverted under different p values with the actual measured values of the soil humidity of the corresponding pixels, and finding out the p value corresponding to the inversion value closest to the actual measured value, namely the optimal p value, or taking 2 p values which are most matched with the actual measured value for linear interpolation to obtain the optimal p value.
(4) For the optimal p value
Figure FDA0004147921860000031
Performing regression analysis on the sine value sin theta of the pixel incidence angle to obtain an inversion model of the roughness spectrum parameter p;
7. the soil moisture inversion method based on pixel scale surface roughness spectrum parameters of claim 1, wherein the expression of the relation model ii is specifically as follows:
Figure FDA0004147921860000032
wherein C is vh (θ) is a coefficient related to the incident angle θ of the pixel, A vh 、B vh Fitting coefficients are respectively used.
8. The soil moisture inversion method based on pixel scale surface roughness spectrum parameters of claim 7, wherein C vh The expression of (θ) is specifically as follows:
C vh (θ)＝a 5 ·cos 2 θ+b 5 ·cosθ+c 5
wherein a is 5 、b 5 、c 5 Is the fitting coefficient.
The soil humidity inversion model consists of an inversion model of a relation model I, a relation model II and a roughness spectrum parameter p.
9. The soil moisture inversion method based on pixel scale surface roughness spectrum parameters as claimed in claim 1, wherein if the surface is bare soil, the backscattering coefficient of radar observation is soil surface scattering component
Figure FDA0004147921860000033
And->
Figure FDA0004147921860000034
If the ground surface is covered by vegetation, the backscattering coefficient observed by radar is the backscattering coefficient of a canopy
Figure FDA0004147921860000035
And->
Figure FDA0004147921860000036
Carrying out surface body scattering separation to obtain soil surface scattering component->
Figure FDA0004147921860000037
And->
Figure FDA0004147921860000038
/>
CN202211304708.7A 2022-10-24 2022-10-24 Soil humidity inversion method based on pixel scale surface roughness spectrum parameters Active CN115618174B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202211304708.7A CN115618174B (en) 2022-10-24 2022-10-24 Soil humidity inversion method based on pixel scale surface roughness spectrum parameters

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202211304708.7A CN115618174B (en) 2022-10-24 2022-10-24 Soil humidity inversion method based on pixel scale surface roughness spectrum parameters

Publications (2)

Publication Number Publication Date
CN115618174A CN115618174A (en) 2023-01-17
CN115618174B true CN115618174B (en) 2023-06-13

Family

ID=84864799

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202211304708.7A Active CN115618174B (en) 2022-10-24 2022-10-24 Soil humidity inversion method based on pixel scale surface roughness spectrum parameters

Country Status (1)

Country Link
CN (1) CN115618174B (en)

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106569210B (en) * 2016-10-18 2017-10-31 长安大学 A kind of Soil Moisture Inversion method and apparatus based on remote sensing
CN111610201B (en) * 2020-04-30 2021-08-10 中国科学院空天信息创新研究院 Passive microwave multi-channel synergistic soil moisture inversion method and device
CN114065643A (en) * 2021-11-24 2022-02-18 电子科技大学长三角研究院(湖州) Plant soil water content estimation method and system based on SAR and polarization decomposition

Also Published As

Publication number Publication date
CN115618174A (en) 2023-01-17

Similar Documents

Publication Publication Date Title
Li et al. An evaluation of the use of atmospheric and BRDF correction to standardize Landsat data
CN110987821B (en) Hyperspectral rapid atmospheric correction parameterization method
CN106960174B (en) Height control point extraction and auxiliary positioning method for high resolution image laser radar
CN102279393B (en) Cross radiometric calibration method of hyper-spectral sensor based on multi-spectral sensor
CN111583214A (en) Method for inverting sea surface wind speed by using marine radar image based on RBF neural network
CN108830889B (en) Global geometric constraint-based remote sensing image and reference image matching method
CN114139444B (en) Offshore sea surface temperature inversion method based on machine learning
CN114740180A (en) Soil organic carbon estimation method and device based on multi-source remote sensing data
CN109344550A (en) A kind of forest reserves inversion method and system based on domestic high score satellite remote sensing date
CN110703220B (en) Multi-baseline PolInSAR vegetation parameter inversion method considering time decoherence factors
CN105404873B (en) A kind of winter wheat recognition methods based on the conversion of NDVI time series coordinates
CN111879725B (en) Spectral data correction method based on weight coefficient
CN109269641B (en) Multi-sensor cooperative radiometric calibration method for space-painting first satellite
CN101216555A (en) RPC model parameter extraction method and geometric correction method
CN114219994A (en) Ocean optical satellite radiometric calibration method based on air-sea cooperative observation
Blaes et al. Characterizing bidimensional roughness of agricultural soil surfaces for SAR modeling
CN102096920A (en) Target image-based sub-pixel registering method
CN113076645B (en) Cosmic ray neutron instrument space characteristic simulation method combining remote sensing data
CN115618174B (en) Soil humidity inversion method based on pixel scale surface roughness spectrum parameters
CN105628052B (en) Optical satellite sensor in-orbit geometric calibration method and system based on control straight line
CN109029735B (en) Earth surface emissivity calculation method and device
Zhuo et al. Quantitative retrieving of soil organic matter using field spectrometer and hyperspectral remote sensing
CN105389466A (en) Middle and low resolution remote sensing product true value acquisition method for correcting scaling effect
CN102901697B (en) Porosity detection method for soil
CN105678716A (en) Correction method and apparatus for atmospheric interference phase in ground-based SAR

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