CN102661811B - Remote sensing earth surface temperature up-scaling method and system - Google Patents

Abstract

The invention discloses a remote sensing surface temperature upscaling method and system, and relates to the field of satellite remote sensing technology. The invention calculates a normalized differential vegetation index sensitive to the surface temperature by making full use of the first scale remote sensing image, and establishes a normalized The non-linear model between the differential vegetation index and the surface temperature uses the difference in the estimated surface temperature of the normalized differential vegetation index under the two scales to realize the error correction of the upscaled surface temperature, which reduces the occurrence of surface temperature upscaling. error.

Description

Remote sensing surface temperature upscaling method and system

technical field

The invention relates to the technical field of satellite remote sensing, in particular to a method and system for remote sensing surface temperature upscaling.

Background technique

Surface temperature is an important indicator of land surface energy balance and climate change. It is a key parameter of surface physical processes at regional and global scales. It is widely used in many research fields such as meteorology, hydrology, geology, ecology, urban environment, and disaster monitoring. need. At present, in the process of carrying out global change research, how to use the first-scale (that is, with higher resolution) surface temperature to obtain the second-scale (that is, with lower resolution) surface temperature has become the key. In the existing surface temperature upscaling methods, the average value of the first-scale surface temperature is used to replace the second-scale surface temperature after upscaling. This method does not take into account that the surface temperature is a physical quantity that has nothing to do with the unit area. Calculating the average will inevitably lead to large errors.

Contents of the invention

(1) Technical problems to be solved

The technical problem to be solved by the invention is: how to reduce the error generated when the surface temperature is scaled up.

(2) Technical solution

In order to solve the above-mentioned technical problems, the present invention provides a method for upscaling of remote sensing surface temperature, said method comprising the following steps:

S1: Obtain the first-scale remote sensing image to be scaled up;

S2: Calculate the first normalized difference vegetation index NDVI _high corresponding to each pixel in the first-scale remote sensing image, and aggregate the pixels of the first-scale remote sensing image according to the required upscaling multiple to obtain the upscaling The second normalized difference vegetation index NDVI _low corresponding to each pixel in the scaled second-scale remote sensing image;

S3: Establish a nonlinear model between normalized difference vegetation index and estimated surface temperature;

S4: Substitute the first normalized difference vegetation index NDVI _high corresponding to each pixel in the first-scale remote sensing image into the nonlinear model to obtain the corresponding The first surface estimated temperature of , and the second normalized difference vegetation index NDVI _low corresponding to each pixel in the second-scale remote sensing image is substituted into the nonlinear model to obtain the second-scale remote sensing image The second surface estimated temperature corresponding to each pixel in ;

S5: Using the first estimated surface temperature corresponding to each pixel in the first-scale remote sensing image, the second estimated surface temperature corresponding to each pixel in the second-scale remote sensing image, and the first-scale remote sensing The surface temperature corresponding to each pixel in the image is calculated by using error correction to calculate the surface temperature corresponding to each pixel in the second-scale remote sensing image.

Preferably, step S2 specifically includes the following steps:

S21: Calculate the first normalized difference vegetation index NDVI _high corresponding to each pixel in the first-scale remote sensing image;

S22: Suppose the scale of the first-scale remote sensing image is G1, then the scale G2 of the larger-scale remote sensing image is shown in the following formula,

G2=N*G1

Wherein, N is the multiple that needs to be scaled up;

S23: Each pixel on the remote sensing image with a larger size corresponds to N*N pixels on the remote sensing image with the first scale, and first normalizes the pixels corresponding to the N*N pixels The pixel average value of the differential vegetation index NDVI _high is used as the second normalized differential vegetation index NDVI _low corresponding to the pixel on the larger-sized remote sensing image.

Preferably, step S3 specifically includes the following steps:

S31: Construct the nonlinear model, the expression of the nonlinear model is as follows,

$\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; NDVI</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; [</span>{\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\frac{{\mathrm{<span\; class="notranslate">\; NDVI</span>}}_{\mathrm{<span\; class="notranslate">\; max</span>}}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; NDVI</span>}}{{\mathrm{<span\; class="notranslate">\; NDVI</span>}}_{\mathrm{<span\; class="notranslate">\; max</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; NDVI</span>}}_{\mathrm{<span\; class="notranslate">\; min</span>}}}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 0.625</span>}}<span\; class="notranslate">\; ]</span>$

Among them, T(NDVI) is the estimated surface temperature, a ₀ and a ₁ are constants respectively, NDVI is the normalized difference vegetation index, NDVI _max is the maximum value of the normalized difference vegetation index, and NDVI _min is the normalized difference vegetation index the minimum value of the index;

S32: Substituting the first normalized difference vegetation index NDVI _high corresponding to each pixel in the first-scale remote sensing image and the surface temperature corresponding to each pixel in the first-scale remote sensing image into the nonlinear model , to calculate a ₀ and a ₁ of the nonlinear model;

S33: Substituting the calculated a ₀ and a ₁ into the nonlinear model to establish a nonlinear model between the normalized difference vegetation index and the estimated surface temperature.

Preferably, step S5 specifically includes the following steps:

所述较大尺寸遥感图像上的每个像元分别对应所述第一尺度遥感图像上的N*N个像元，获得所述N*N个像元对应的第一地表估计温度的平均值

S51: Let the second estimated surface temperature corresponding to each pixel in the second-scale remote sensing image be

Each pixel on the larger-scale remote sensing image corresponds to N*N pixels on the first-scale remote sensing image, and the average value of the first surface estimated temperature corresponding to the N*N pixels is obtained

和所述第二地表估计温度

之间的差值ΔT_low；S52: Calculate the average value of the first surface estimated temperature corresponding to the N*N pixels

and the second estimated surface temperature

The difference between ΔT _low ;

通过下列公式计算所述第二尺度遥感图像中每个像元对应的地表温度T_low，S53: Obtain the average value of the surface temperature corresponding to the N*N pixels

The surface temperature T _low corresponding to each pixel in the second-scale remote sensing image is calculated by the following formula,

${\mathrm{<span\; class="notranslate">\; T</span>}}_{\mathrm{<span\; class="notranslate">\; low</span>}}<span\; class="notranslate">\; =</span>{\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; T</span>}}}_{\mathrm{<span\; class="notranslate">\; high</span>}}^{\mathrm{<span\; class="notranslate">\; N</span>}<span\; class="notranslate">\; *</span>\mathrm{<span\; class="notranslate">\; N</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; \&Delta;T</span>}}_{\mathrm{<span\; class="notranslate">\; low</span>}}$

in,

The invention also discloses a remote sensing surface temperature upscaling system, said system comprising:

An acquisition module, configured to acquire a first-scale remote sensing image to be scaled up;

The vegetation index calculation module is used to calculate the first normalized difference vegetation index NDVI _high corresponding to each pixel in the first-scale remote sensing image, and calculate the pixel of the first-scale remote sensing image according to the multiple of upscaling required. Aggregation to obtain the second normalized difference vegetation index NDVI _low corresponding to each pixel in the upscaled second-scale remote sensing image;

A model building module for building a nonlinear model between the normalized difference vegetation index and the estimated surface temperature;

An estimated temperature calculation module, for substituting the first normalized differential vegetation index NDVI _high corresponding to each pixel in the first-scale remote sensing image into the nonlinear model to obtain the first-scale remote sensing image The first surface estimated temperature corresponding to each pixel, and the second normalized difference vegetation index NDVI _low corresponding to each pixel in the second-scale remote sensing image are respectively substituted into the nonlinear model to obtain the The second surface estimated temperature corresponding to each pixel in the second-scale remote sensing image;

An error correction module, configured to use the first estimated surface temperature corresponding to each pixel in the first-scale remote sensing image, the second estimated surface temperature corresponding to each pixel in the second-scale remote sensing image, and the The surface temperature corresponding to each pixel in the first-scale remote sensing image is calculated by using an error correction method to calculate the surface temperature corresponding to each pixel in the second-scale remote sensing image.

(3) Beneficial effects

The present invention calculates the normalized difference vegetation index sensitive to the surface temperature by making full use of the first-scale remote sensing image, and establishes a nonlinear model between the normalized difference vegetation index and the surface temperature, and utilizes two scales to normalize The estimated surface temperature difference of the differential vegetation index realizes the error correction of the upscaled surface temperature and reduces the error generated when the surface temperature is upscaled.

Description of drawings

Fig. 1 is a flowchart of a method for upscaling remote sensing land surface temperature according to an embodiment of the present invention.

Detailed ways

The specific implementation manners of the present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. The following examples are used to illustrate the present invention, but are not intended to limit the scope of the present invention.

Fig. 1 is the flow chart of the remote sensing surface temperature upscaling method according to an embodiment of the present invention; With reference to Fig. 1, the method of this embodiment comprises the following steps:

S1: Obtain the first-scale remote sensing image to be scaled up;

S2: Calculate the first normalized difference vegetation index NDVI _high corresponding to each pixel in the first-scale remote sensing image, and aggregate the pixels of the first-scale remote sensing image according to the required upscaling multiple to obtain the upscaling The second normalized difference vegetation index NDVI _low corresponding to each pixel in the scaled second-scale remote sensing image;

S3: Establish a nonlinear model between normalized difference vegetation index and estimated surface temperature;

S4: Substitute the first normalized difference vegetation index NDVI _high corresponding to each pixel in the first-scale remote sensing image into the nonlinear model to obtain the corresponding The first surface estimated temperature of , and the second normalized difference vegetation index NDVI _low corresponding to each pixel in the second-scale remote sensing image is substituted into the nonlinear model to obtain the second-scale remote sensing image The second surface estimated temperature corresponding to each pixel in ;

S5: Using the first estimated surface temperature corresponding to each pixel in the first-scale remote sensing image, the second estimated surface temperature corresponding to each pixel in the second-scale remote sensing image, and the first-scale remote sensing The surface temperature corresponding to each pixel in the image is calculated by using error correction to calculate the surface temperature corresponding to each pixel in the second-scale remote sensing image.

Preferably, step S2 specifically includes the following steps:

S21: Calculate the first normalized difference vegetation index NDVI _high corresponding to each pixel in the first-scale remote sensing image;

S22: Suppose the scale of the first-scale remote sensing image is G1, then the scale G2 of the larger-scale remote sensing image is shown in the following formula,

G2=N*G1

Wherein, N is a multiple of the required upscaling; it means that any pixel at the G2 scale contains N*N pixels at the G1 scale. The NDVI _low of the pixel set is obtained by setting the N*N size window, and the window is sequentially moved on the NDVI _high data at the G1 scale, and the moving step is set to a window size, and finally the corresponding NDVI _high at the G1 scale in each window is The average value of the pixel is used as the NDVI _low pixel value at the G2 scale.

S23: Each pixel on the remote sensing image with a larger size corresponds to N*N pixels on the remote sensing image with the first scale, and first normalizes the pixels corresponding to the N*N pixels The pixel average value of the differential vegetation index NDVI _high is used as the second normalized differential vegetation index NDVI _low corresponding to the pixel on the larger-sized remote sensing image.

Preferably, step S3 specifically includes the following steps:

S31: Construct the nonlinear model, the expression of the nonlinear model is as follows,

$\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; NDVI</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; [</span>{\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\frac{{\mathrm{<span\; class="notranslate">\; NDVI</span>}}_{\mathrm{<span\; class="notranslate">\; max</span>}}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; NDVI</span>}}{{\mathrm{<span\; class="notranslate">\; NDVI</span>}}_{\mathrm{<span\; class="notranslate">\; max</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; NDVI</span>}}_{\mathrm{<span\; class="notranslate">\; min</span>}}}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 0.625</span>}}<span\; class="notranslate">\; ]</span>$

Among them, T(NDVI) is the estimated surface temperature, a ₀ and a ₁ are constants respectively, NDVI is the normalized difference vegetation index, NDVI _max is the maximum value of the normalized difference vegetation index, and NDVI _min is the normalized difference vegetation index the minimum value of the index;

S32: Substituting the first normalized difference vegetation index NDVI _high corresponding to each pixel in the first-scale remote sensing image and the surface temperature corresponding to each pixel in the first-scale remote sensing image into the nonlinear model , to calculate a ₀ and a ₁ of the nonlinear model;

S33: Substituting the calculated a ₀ and a ₁ into the nonlinear model to establish a nonlinear model between the normalized difference vegetation index and the estimated surface temperature.

Preferably, step S5 specifically includes the following steps:

所述较大尺寸遥感图像上的每个像元分别对应所述第一尺度遥感图像上的N*N个像元，获得所述N*N个像元对应的第一地表估计温度的平均值

S51: Let the second estimated surface temperature corresponding to each pixel in the second-scale remote sensing image be

Each pixel on the larger-scale remote sensing image corresponds to N*N pixels on the first-scale remote sensing image, and the average value of the first surface estimated temperature corresponding to the N*N pixels is obtained

和所述第二地表估计温度

之间的差值ΔT_low；S52: Calculate the average value of the first surface estimated temperature corresponding to the N*N pixels

and the second estimated surface temperature

The difference between ΔT _low ;

通过下列公式计算所述第二尺度遥感图像中每个像元对应的地表温度T_low，S53: Obtain the average value of the surface temperature corresponding to the N*N pixels

The surface temperature T _low corresponding to each pixel in the second-scale remote sensing image is calculated by the following formula,

${\mathrm{<span\; class="notranslate">\; T</span>}}_{\mathrm{<span\; class="notranslate">\; low</span>}}<span\; class="notranslate">\; =</span>{\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; T</span>}}}_{\mathrm{<span\; class="notranslate">\; high</span>}}^{\mathrm{<span\; class="notranslate">\; N</span>}<span\; class="notranslate">\; *</span>\mathrm{<span\; class="notranslate">\; N</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; \&Delta;T</span>}}_{\mathrm{<span\; class="notranslate">\; low</span>}}$

in,

The invention also discloses a remote sensing surface temperature upscaling system, said system comprising:

An acquisition module, configured to acquire a first-scale remote sensing image to be scaled up;

The vegetation index calculation module is used to calculate the first normalized difference vegetation index NDVI _high corresponding to each pixel in the first-scale remote sensing image, and calculate the pixel of the first-scale remote sensing image according to the multiple of upscaling required. Aggregation to obtain the second normalized difference vegetation index NDVI _low corresponding to each pixel in the upscaled second-scale remote sensing image;

A model building module for building a nonlinear model between the normalized difference vegetation index and the estimated surface temperature;

An estimated temperature calculation module, for substituting the first normalized differential vegetation index NDVI _high corresponding to each pixel in the first-scale remote sensing image into the nonlinear model to obtain the first-scale remote sensing image The first surface estimated temperature corresponding to each pixel, and the second normalized difference vegetation index NDVI _low corresponding to each pixel in the second-scale remote sensing image are respectively substituted into the nonlinear model to obtain the The second surface estimated temperature corresponding to each pixel in the second-scale remote sensing image;

An error correction module, configured to use the first estimated surface temperature corresponding to each pixel in the first-scale remote sensing image, the second estimated surface temperature corresponding to each pixel in the second-scale remote sensing image, and the The surface temperature corresponding to each pixel in the first-scale remote sensing image is calculated by using an error correction method to calculate the surface temperature corresponding to each pixel in the second-scale remote sensing image.

The above embodiments are only used to illustrate the present invention, but not to limit the present invention. Those of ordinary skill in the relevant technical field can make various changes and modifications without departing from the spirit and scope of the present invention. Therefore, all Equivalent technical solutions also belong to the category of the present invention, and the scope of patent protection of the present invention should be defined by the claims.

Claims (4)

1. A remote sensing surface temperature upscaling method, characterized in that the method comprises the following steps: S1: Obtain the first-scale remote sensing image to be scaled up; S2: Calculate the first normalized difference vegetation index NDVI _high corresponding to each pixel in the first-scale remote sensing image, and aggregate the pixels of the first-scale remote sensing image according to the required upscaling multiple to obtain the upscaling The second normalized difference vegetation index NDVI _low corresponding to each pixel in the scaled second-scale remote sensing image; S3: Establish a nonlinear model between the normalized difference vegetation index and the estimated surface temperature, which specifically includes the following steps: S31: Construct the nonlinear model, the expression of the nonlinear model is as follows, $\mathrm{<span\; class="notranslate">\; T</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; NDVI</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>{<span\; class="notranslate">\; (</span>\frac{{\mathrm{<span\; class="notranslate">\; NDVI</span>}}_{\mathrm{<span\; class="notranslate">\; max</span>}}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; NDVI</span>}}{{\mathrm{<span\; class="notranslate">\; NDVI</span>}}_{\mathrm{<span\; class="notranslate">\; max</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; NDVI</span>}}_{\mathrm{<span\; class="notranslate">\; min</span>}}}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 0.625</span>}}<span\; class="notranslate">\; ]</span>$ Among them, T(NDVI) is the estimated surface temperature, a ₀ and a ₁ are constants respectively, NDVI is the normalized difference vegetation index, NDVI _max is the maximum value of the normalized difference vegetation index, and NDVI _min is the normalized difference vegetation index the minimum value of the index; S32: Substituting the first normalized difference vegetation index NDVI _high corresponding to each pixel in the first-scale remote sensing image and the surface temperature corresponding to each pixel in the first-scale remote sensing image into the nonlinear model , to calculate a ₀ and a ₁ of the nonlinear model; S33: Substituting the calculated a ₀ and a ₁ into the nonlinear model to establish a nonlinear model between the normalized difference vegetation index and the estimated surface temperature; S4: Substitute the first normalized difference vegetation index NDVI _high corresponding to each pixel in the first-scale remote sensing image into the nonlinear model to obtain the corresponding The first surface estimated temperature of , and the second normalized difference vegetation index NDVI _low corresponding to each pixel in the second-scale remote sensing image is substituted into the nonlinear model to obtain the second-scale remote sensing image The second surface estimated temperature corresponding to each pixel in ; S5: Using the first estimated surface temperature corresponding to each pixel in the first-scale remote sensing image, the second estimated surface temperature corresponding to each pixel in the second-scale remote sensing image, and the first-scale remote sensing The surface temperature corresponding to each pixel in the image is calculated by using error correction to calculate the surface temperature corresponding to each pixel in the second-scale remote sensing image. 2. The method according to claim 1, wherein step S2 specifically comprises the following steps: S21: Calculate the first normalized difference vegetation index NDVI _high corresponding to each pixel in the first-scale remote sensing image; S22: Suppose the scale of the first scale remote sensing image is G1, then the scale G2 of the second scale remote sensing image is shown in the following formula, G2=N*G1 Wherein, N is the multiple of the required upscaling; S23: Each pixel on the second-scale remote sensing image corresponds to N*N pixels on the first-scale remote sensing image, and first normalizes the N*N pixels corresponding to The pixel average value of the differential vegetation index NDVI _high is used as the second normalized differential vegetation index NDVI _low corresponding to the pixel on the second-scale remote sensing image. 3. The method according to claim 1, wherein step S5 specifically comprises the following steps: S51: Let the second estimated surface temperature corresponding to each pixel in the second-scale remote sensing image be

Each pixel on the second-scale remote sensing image corresponds to N*N pixels on the first-scale remote sensing image, and the average value of the first surface estimated temperature corresponding to the N*N pixels is obtained

S52: Calculate the average value of the first surface estimated temperature corresponding to the N*N pixels

and the second estimated surface temperature The difference between ΔT _low ; S53: Obtain the average value of the surface temperature corresponding to the N*N pixels

The surface temperature T _low corresponding to each pixel in the second-scale remote sensing image is calculated by the following formula, ${\mathrm{<span\; class="notranslate">\; T</span>}}_{\mathrm{<span\; class="notranslate">\; low</span>}}<span\; class="notranslate">\; =</span>{\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; T</span>}}}_{\mathrm{<span\; class="notranslate">\; high</span>}}^{\mathrm{<span\; class="notranslate">\; N</span>}<span\; class="notranslate">\; *</span>\mathrm{<span\; class="notranslate">\; N</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; \&Delta;T</span>}}_{\mathrm{<span\; class="notranslate">\; low</span>}}$ in,

4. A remote sensing surface temperature upscaling system, characterized in that the system comprises: An acquisition module, configured to acquire a first-scale remote sensing image to be scaled up; The vegetation index calculation module is used to calculate the first normalized difference vegetation index NDVI _high corresponding to each pixel in the first-scale remote sensing image, and calculate the pixel of the first-scale remote sensing image according to the multiple of upscaling required. Aggregation to obtain the second normalized difference vegetation index NDVI _low corresponding to each pixel in the upscaled second-scale remote sensing image; A model building module for building a nonlinear model between the normalized difference vegetation index and the estimated surface temperature; An estimated temperature calculation module, for substituting the first normalized differential vegetation index NDVI _high corresponding to each pixel in the first-scale remote sensing image into the nonlinear model to obtain the first-scale remote sensing image The first surface estimated temperature corresponding to each pixel, and the second normalized difference vegetation index NDVI _low corresponding to each pixel in the second-scale remote sensing image are respectively substituted into the nonlinear model to obtain the The second surface estimated temperature corresponding to each pixel in the second-scale remote sensing image; An error correction module, configured to use the first estimated surface temperature corresponding to each pixel in the first-scale remote sensing image, the second estimated surface temperature corresponding to each pixel in the second-scale remote sensing image, and the The surface temperature corresponding to each pixel in the first-scale remote sensing image is calculated by using an error correction method to calculate the surface temperature corresponding to each pixel in the second-scale remote sensing image.

Images