Spatial neighborhood constraint-based surface temperature and angle collaborative normalization method

Abstract

The invention discloses a land surface temperature and angle collaborative normalization method based on spatial neighborhood constraint, and the method comprises the steps: taking a multi-temporal multi-pixel multi-angle land surface temperature observation value as an input, and coupling an angle effect nuclear drive model with a time angle collaborative normalization model of a land surface temperature daily variation curve model of a time effect; a least square mathematical method is adopted, parameters of a time angle collaborative normalization model are determined, space neighborhood information is added, finally, the unknown number of the surface temperature is reduced, the ill-conditioned problem is solved, and collaborative normalization of a surface temperature product is achieved; in addition, time-angle collaborative normalization of the surface temperature product can be realized under a small amount of observation.

Description

Technical Field

[0001] The invention relates to a method for co-normalizing surface temperature and angle, and in particular to a method for co-normalizing surface temperature and angle based on spatial neighborhood constraints. Background Art

[0002] Land surface temperature is an important physical quantity that characterizes the interaction and energy exchange between the ground and the atmosphere. It is widely used in the fields of regional climate change and ecological environment monitoring. The acquisition of high-precision land surface temperature depends on thermal infrared remote sensing technology. However, existing land surface temperature remote sensing products have temporal and directional uncertainties. On the one hand, due to the inherent observation scanning characteristics of polar-orbiting satellites, the local observation time differences between different polar-orbiting satellite data and on the east-west scanning line of the same satellite data are large, which leads to a lack of consistency in different land surface temperature products; on the other hand, due to the complex and diverse surface types and their three-dimensional structural characteristics, there may be significant differences in land surface temperature products obtained from different observation angles. These uncertainties seriously limit the widespread application of land surface temperature products.

[0003] Existing time normalization methods mainly use diurnal temperature variation models or empirical statistical models to describe the time variation characteristics of thermal radiation; angle normalization is achieved through physical or parameterized models of emissivity directionality or component ratio changes in the field of view. However, these methods can usually only handle time or angle factors separately, ignoring the synergistic effect of the two, resulting in limited accuracy of normalized results. Summary of the invention

[0004] In order to solve the shortcomings of the above-mentioned technology, the present invention provides a method for co-normalization of surface temperature and angle based on spatial neighborhood constraints.

[0005] In order to solve the above technical problems, the technical solution adopted by the present invention is: a surface temperature and angle collaborative normalization method based on spatial neighborhood constraints, the surface temperature and angle collaborative normalization method is:

[0006] Using multi-phase, multi-pixel and multi-angle surface temperature observations as input, the time-angle co-normalization model of the angle effect kernel driving model and the time effect surface temperature daily variation curve model is coupled. The least squares mathematical method is used to determine the parameters of the time-angle co-normalization model, and the spatial neighborhood information is added to finally achieve the co-normalization of surface temperature and time angle to produce the time-angle co-normalized surface temperature product.

[0007] As preferably, the method specifically includes the following steps:

[0008] Step 1: Extraction of multi-phase, multi-angle, and multi-pixel data: According to the spatial and temporal range of the area to be acquired, extract the surface temperature data and its corresponding solar zenith angle, observation zenith angle, solar azimuth angle, and observation azimuth angle data; extract the imaging time and convert the imaging time into decimal time;

[0009] Step 2: Parametric expression of the surface temperature time-angle synergistic model: including the parametric expression of angle effect, the parametric expression of time effect and the parametric expression of time-angle synergy;

[0010] Step 3: Introduce neighborhood information to calibrate the parameters of the surface temperature time-angle collaborative model;

[0011] Step 4: After solving the parameters, the normalized surface temperature product at any time and any observation angle can be obtained in each pixel.

[0012] Preferably, in step 2, the parameterized expression of the angle effect is the surface temperature measured for the sun-sensor geometry, expressed as a linear combination of kernels, as shown in formulas (1)-(3):

[0013]

[0014] K geo (θ v )=1-cos(θ v ) (2)

[0015]

[0016] Among them, θ s ,θ v Respectively represent the solar zenith angle and the sensor observation zenith angle, Indicates the difference between the solar azimuth and the observation azimuth; K geo (θ v ) is the geometric optics kernel, which represents the influence of three-dimensional structure; K solar is the solar core, which indicates the influence of the sun-sensor observation geometry. At night, due to the absence of solar radiation, K solar =0; f geo 、f solar Respectively represent the weight of each core; T N is the surface temperature observed at nadir.

[0017] Preferably, in step 2, the parameterized expression of the time effect uses a cosine function based on the heat diffusion equation to model the daytime temperature change, and uses an exponential function based on Newtonian cooling to model the nighttime temperature decay, which is called a DTC model. The functional relationship is expressed as shown in formula (4):

[0018]

[0019] Where t is the time, T0 is the residual temperature before and after sunrise, and T a is the temperature amplitude, ω is the half-period width of the cosine term and can be regarded as a constant; t m is the time when the temperature reaches its maximum value, t s is the starting time of free decay, δT is the temperature difference between T0 and T(t→∞), and k is the decay constant;

[0020] T day (t) is the diurnal variation of surface temperature, T night (t) is the night-time variation of surface temperature.

[0021] As a preferred embodiment, in step 2, for the time angle collaborative parameterization expression, the surface temperature T of the nadir observation N It can be described in time by the DTC model. Substituting formula (4) into formula (1), the time-angle coordination model of surface temperature is expressed as shown in formula (5):

[0022]

[0023] Among them, T N (t) represents the temporal dynamics of the surface temperature observed at the nadir, which is expressed as:

[0024]

[0025] In the time angle coordination model, the unknown parameters are: five DTC model parameters, namely T0, T a ,t m ,t s and δT; and two nuclear driving parameters, namely, f geo and f solar .

[0026] As a preferred embodiment, the specific processing process of step 3 is: setting the shapes of the DTC models to be the same, that is, sharing a set of T0, t m ,t s and δT, but the temperature amplitude T a different;

[0027] The kernel driving parameter f of each pixel geo and f solar different;

[0028] Therefore, the surface temperature time-angle synergy model is further parameterized as:

[0029]

[0030] Among them, i and j are the row and column numbers of the pixel respectively.

[0031] As a preferred method, after introducing the neighborhood information, the initial values ​​of the parameters are determined first, and then the optimal solution of the parameters of the surface temperature time angle collaborative model is iterated through a nonlinear algorithm. The objective function can be written as the sum of the squares of the observed value and the model predicted value, as shown in formula (8):

[0032]

[0033] Among them, T obs,i,j,t is the observed value, T model,i,j,t is the model prediction value of formula (7);

[0034] By minimizing the objective function, iterating the unknown parameters, including T0, T a ,t m ,t s and δT, as well as pixel-related parameters T a,i,j 、f geo,i,j and f solar,i,j , and finally the optimal model parameters are obtained.

[0035] Preferably, in step 4, the production of the time angle coordinated normalized surface temperature product is achieved by formula (9):

[0036]

[0037] The present invention discloses a method for co-normalization of surface temperature and angle based on spatial neighborhood constraints. Through the temporal, spatial and angular information of remote sensing data, the surface temperature of local observation time is normalized to the same time, and the angle effect is quantified at the same time, so that the surface temperature of different observation angles is normalized to a unified angle. This method will effectively improve the accuracy and consistency of surface temperature products, and provide unified standard data support for research on climate change, vegetation monitoring, etc. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 It is a schematic diagram of the overall technical process of the present invention. DETAILED DESCRIPTION

[0039] The present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments.

[0040] The present invention proposes a method based on spatial neighborhood constraints and coordinated normalization of surface temperature and angle, which utilizes multi-phase, multi-pixel and multi-angle surface temperature observations as input, couples the angle effect kernel driving model and the time effect surface temperature daily variation curve model to form a time-angle coordinated normalization model, adds spatial neighborhood information, and adopts the least squares mathematical method to determine the parameters of the time-angle coordinated normalization model, ultimately achieving coordinated normalization of surface temperature and time angle to produce time-angle coordinated normalized surface temperature products.

[0041] The overall technical flow chart of the present invention is as follows: Figure 1 As shown, it is mainly divided into the following four steps:

[0042] 1) Multi-phase, multi-angle and multi-pixel data extraction;

[0043] 2) Parameterized expression of the surface temperature time-angle synergy model;

[0044] 3) Parameter calibration of the surface temperature time-angle synergy model;

[0045] 4) Production of surface temperature and time angle normalized products.

[0046] The method based on spatial neighborhood constraint and coordinated normalization of surface temperature and angle proposed in the present invention is mainly aimed at MODIS or VIIRS sensor data that can form 4 observations in one day. The specific implementation steps are as follows:

[0047] Step 1: Extraction of multi-phase, multi-angle and multi-pixel data:

[0048] According to the spatial and temporal range of the area to be acquired, the surface temperature data and its corresponding solar zenith angle, observation zenith angle, solar azimuth angle, and observation azimuth angle data are extracted; and the imaging time is extracted and converted into decimal time.

[0049] Taking MODIS as an example, the specific extraction process of multi-phase, multi-angle and multi-pixel data is as follows: according to the temporal and spatial scope of the area to be acquired, download the L2-level surface temperature products MOD11A1 and MYD11A1 (on the same day) of the MODIS medium-resolution imaging spectrometer on Terra or Aqua satellites. Extract the surface temperature data of MODIS at four moments and their corresponding solar zenith angle, observation zenith angle, solar azimuth angle and observation azimuth angle data from the MOD11A1 and MYD11A1 products in HDF format. In addition, read the accurate imaging time from the HDF file and convert it into decimal time. The four observation times are around 1:30 a.m., 10:30 a.m., 13:30 p.m. and 22:30 p.m.

[0050] Step 2: Parameterized expression of the surface temperature time angle synergistic model:

[0051] It includes parametric expression of angle effect, parametric expression of time effect and parametric expression of time-angle coordination.

[0052] in:

[0053] (1) Parametric expression of angle effect

[0054] Under natural conditions, the land surface at the satellite pixel scale is always heterogeneous, consisting of multiple homogeneous components with different temperatures. For a specific sun-sensor geometry, the surface temperature measured can be expressed as a linear combination of kernels, as shown in formulas (1)-(3):

[0055]

[0056] K geo (θ v )=1-cos(θ v ) (2)

[0057]

[0058] Among them, θ s ,θ v Respectively represent the solar zenith angle and the sensor observation zenith angle, Indicates the difference between the solar azimuth and the observation azimuth; K geo (θ v ) is the geometric optics kernel, which represents the influence of three-dimensional structure; K solar is the solar core, which indicates the influence of the sun-sensor observation geometry. At night, due to the absence of solar radiation, K solar =0; f geo 、f solar Respectively represent the weight of each core; T N is the surface temperature observed at nadir.

[0059] There are three free parameters to model the angle effect, including the weight of each kernel (f geo and f solar ) and the surface temperature T N .

[0060] (2) Parametric expression of time effect

[0061] For the daily variation of surface temperature under clear sky conditions, the cosine function based on the heat diffusion equation can be used to model the daytime temperature variation, and the exponential function based on Newtonian cooling can be used to model the nighttime temperature decay, which is called the DTC model. The functional relationship is shown in formula (4):

[0062]

[0063] Where t is the time, T0 is the residual temperature before and after sunrise, and T a is the temperature amplitude, ω is the half-period width of the cosine term and can be regarded as a constant; t m is the time when the temperature reaches its maximum value, t s is the starting time of free decay, δT is the temperature difference between T0 and T(t→∞), and k is the decay constant;

[0064] T day (t) is the diurnal variation of surface temperature, T night (t) is the night-time variation of surface temperature.

[0065] Therefore, the five free parameters (i.e. T0 、T a ,t m ,t s and δT) to describe the diurnal and nocturnal variations in surface temperature.

[0066] (3) Cooperative parametric expression of time angle

[0067] For the nuclear-driven model describing the angle effect, the surface temperature T N is a time-varying quantity, so T N It can be described in time by the DTC model. Substituting equation (4) into equation (1), the time-angle coordination model of surface temperature can be written as:

[0068]

[0069] Among them, T N (t) represents the temporal dynamics of the surface temperature observed at the nadir, which is expressed as:

[0070]

[0071] In the time angle coordination model, there are five DTC model parameters (i.e., T0, T a ,t m ,t s and δT) and two nuclear driving parameters (f geo and f solar )unknown.

[0072] Step 3: Introduce neighborhood information to calibrate the parameters of the surface temperature time angle collaborative model:

[0073] Since MODIS has only four observations per day, it is impossible to solve the seven parameters of the time-angle coordination model (i.e., T0, T a ,t m ,t s and δT and f geo and f solar ), so it is necessary to reduce the number of unknowns by introducing neighborhood information.

[0074] Specifically, within the 3×3 window, it is reasonable to assume that the shapes of the DTC models are the same, that is, they share a set of T0, t m ,t sand δT, but the energy balance state of each pixel is different, so the maximum temperature that can be reached is different, that is, the temperature amplitude T a different.

[0075] In addition, the three-dimensional structure of each pixel is different, so the kernel driving parameter (f geo and f solar ). Therefore, the surface temperature time-angle synergy model can be further parameterized as:

[0076]

[0077] Among them, i and j are the row and column numbers of the pixel respectively.

[0078] After the introduction of neighborhood information, there are 4+3×3×3 unknowns, but there are 3×3×4 observations, which overcomes the ill-conditioned solution problem. These unknowns can be solved by nonlinear optimization methods, but nonlinear solution methods are often sensitive to initial values, so it is necessary to determine the initial values ​​of the parameters first.

[0079] For the DTC model, the initial value of T0 is set to the mean surface temperature observed at 1:30 a.m. in step 1; T a The initial value of t is set to the difference between the mean surface temperature observed at 13:30 pm and 1:30 am; m and t s The initial values ​​are set to 12.5 and 17 respectively; δT is set to 0.5. For the kernel-driven model, the kernel weight f of each pixel geo and f solar The initial value of is set to 0.

[0080] After determining the initial values ​​of the parameters, the parameters in the surface temperature time angle coordinated normalization model can be iterated through a nonlinear algorithm to obtain the optimal solution of the model parameters. The objective function can be written as the sum of the squares of the observed value and the model predicted value, as shown in formula (8):

[0081]

[0082] Among them, T obs,i,j,t is the observed value, T model,i,j,t is the model prediction value of formula (7).

[0083] By minimizing the objective function, iterating the unknown parameters, including the shared parameters T0 and T in the window a ,t m ,t s and δT, pixel-related parameter T a,i,j 、f geo,i,j and f solar,i,j , and finally the optimal model parameters are obtained.

[0084] Step 4: Production of surface temperature time angle normalized product:

[0085] After the parameters are solved, the normalized surface temperature product at any time and any observation angle can be obtained in each pixel, realizing the coordinated normalization of time and angle:

[0086]

[0087] It can be seen that the surface temperature and time angle collaborative normalization method based on spatial neighborhood information constraints proposed in the present invention, by coupling the kernel-driven model describing the surface temperature angle effect and the surface temperature daily change curve model describing the time change, collaboratively models the time angle, introduces spatial neighborhood information, reduces the number of unknowns, solves the pathological problem, and realizes the collaborative normalization of the surface temperature product. Using the normalization method provided by the present invention, the time angle collaborative normalization of the surface temperature product can be realized with a small number of observations.

[0088] The above implementation modes are not limitations of the present invention, and the present invention is not limited to the above examples. Changes, modifications, additions or substitutions made by technicians in this technical field within the scope of the technical solution of the present invention also belong to the protection scope of the present invention.

Claims

1. A method for co-normalization of surface temperature and angle based on spatial neighborhood constraints, characterized by: The surface temperature and angle co-normalization method is: Using multi-phase, multi-pixel and multi-angle surface temperature observations as input, the time-angle co-normalization model of the angle effect kernel driving model and the time effect surface temperature daily variation curve model is coupled. The least squares mathematical method is used to determine the parameters of the time-angle co-normalization model, and the spatial neighborhood information is added to finally achieve the co-normalization of surface temperature and time angle to produce the time-angle co-normalized surface temperature product.

2. The method for collaborative normalization of surface temperature and angle based on spatial neighborhood constraints according to claim 1 is characterized in that: The specific steps include: Step 1: Extraction of multi-phase, multi-angle, and multi-pixel data: According to the spatial and temporal range of the area to be acquired, extract the surface temperature data and its corresponding solar zenith angle, observation zenith angle, solar azimuth angle, and observation azimuth angle data; extract the imaging time and convert the imaging time into decimal time; Step 2: Parametric expression of the surface temperature time-angle synergistic model: including the parametric expression of angle effect, the parametric expression of time effect and the parametric expression of time-angle synergy; Step 3: Introduce neighborhood information to calibrate the parameters of the surface temperature time-angle collaborative model; Step 4: After solving the parameters, the normalized surface temperature product at any time and any observation angle can be obtained in each pixel.

3. The method for collaborative normalization of surface temperature and angle based on spatial neighborhood constraints according to claim 2 is characterized in that: In step 2, the parameterized expression of the angle effect is the surface temperature measured for the sun-sensor geometry, expressed as a linear combination of kernels, as shown in equations (1)-(3): K geo (i v )=1-cos(θ v ) (2) Among them, θ s ,θ v Respectively represent the solar zenith angle and the sensor observation zenith angle, Indicates the difference between the solar azimuth and the observation azimuth; K geo (θ v ) is the geometric optics kernel, which represents the influence of three-dimensional structure; K solar is the solar core, which indicates the influence of the sun-sensor observation geometry. At night, due to the absence of solar radiation, K solar =0; f geo 、f solar Respectively represent the weight of each core; T N is the surface temperature observed at nadir.

4. The method for collaborative normalization of surface temperature and angle based on spatial neighborhood constraints according to claim 3 is characterized in that: In step 2, the parameterized expression of the time effect uses the cosine function based on the heat diffusion equation to model the daytime temperature change, and uses the exponential function based on Newtonian cooling to model the nighttime temperature decay, which is called the DTC model. The functional relationship is shown in formula (4): Where t is the time, T0 is the residual temperature before and after sunrise, and T a is the temperature amplitude, ω is the half-period width of the cosine term and can be regarded as a constant; t m is the time when the temperature reaches its maximum value, t s is the starting time of free decay, δT is the temperature difference between T0 and T(t→∞), and k is the decay constant; T day (t) is the diurnal variation of surface temperature, T night (t) is the night-time variation of surface temperature.

5. The method for collaborative normalization of surface temperature and angle based on spatial neighborhood constraints according to claim 4 is characterized in that: In step 2, for the time angle co-parameterization expression, the surface temperature T N It can be described in time by the DTC model. Substituting formula (4) into formula (1), the time-angle coordination model of surface temperature is expressed as shown in formula (5): Among them, T N (t) represents the temporal dynamics of the surface temperature observed at the nadir, which is expressed as: In the time angle coordination model, the unknown parameters are: five DTC model parameters, namely T0, T a ,t m ,t s and δT; and two nuclear driving parameters, namely, f geo and f solar .

6. The method for collaborative normalization of surface temperature and angle based on spatial neighborhood constraints according to claim 5 is characterized in that: The specific processing process of step 3 is: set the shape of the DTC model to be the same, that is, share a set of T0, t m ,t s and δT, but the temperature amplitude T a different; The kernel driving parameter f of each pixel geo and f solar different; Therefore, the surface temperature time-angle synergy model is further parameterized as: Among them, i and j are the row and column numbers of the pixel respectively.

7. The method for collaborative normalization of surface temperature and angle based on spatial neighborhood constraints according to claim 6 is characterized in that: After introducing the neighborhood information, the initial values ​​of the parameters are determined first, and then the optimal solution of the surface temperature time angle collaborative model parameters is iterated through a nonlinear algorithm. The objective function can be written as the sum of the squares of the observed value and the model predicted value, as shown in formula (8): Among them, T obs,i,j,t is the observed value, T model,i,j,t is the model prediction value of formula (7); By minimizing the objective function, iterating the unknown parameters, including T0, T a ,t m ,t s and δT, as well as pixel-related parameters T a,i,j 、f geo,i,j and f solar,i,j , and finally the optimal model parameters are obtained.

8. The method for collaborative normalization of surface temperature and angle based on spatial neighborhood constraints according to claim 7 is characterized in that: Step 4: The production of the time-angle coordinated normalized surface temperature product is achieved through formula (9):

Images