CN109883957B - MODIS image-based apparent reflectivity model construction method, system and calibration method - Google Patents
MODIS image-based apparent reflectivity model construction method, system and calibration method Download PDFInfo
- Publication number
- CN109883957B CN109883957B CN201811573615.8A CN201811573615A CN109883957B CN 109883957 B CN109883957 B CN 109883957B CN 201811573615 A CN201811573615 A CN 201811573615A CN 109883957 B CN109883957 B CN 109883957B
- Authority
- CN
- China
- Prior art keywords
- image
- apparent
- calibration
- modis
- apparent reflectivity
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Landscapes
- Image Processing (AREA)
Abstract
The invention relates to an MODIS image-based apparent reflectivity model construction method, a system and a calibration method, wherein an effective calibration field image is selected to obtain corresponding image apparent reflectivity and related angle information, and a kernel-driven model is further utilized to obtain an apparent reflectivity correction coefficient so as to establish an apparent reflectivity model for obtaining the apparent reflectivity, thereby realizing the cross calibration of satellite image pairs of different sensors on the basis of the apparent reflectivity model, the method can be suitable for the on-orbit calibration of a multispectral satellite, particularly the cross calibration between large-angle image pairs, the method has the advantages of high calibration precision, capability of being used for historical data recalibration and the like, the method does not need to develop a satellite-to-ground synchronous experiment, cancels the assumption that the field must have lambertian property, can be used for the cross calibration of small-angle images, and can be used for the cross calibration of image pairs under large angles, the frequency and the precision of the calibration are greatly improved.
Description
Technical Field
The invention relates to an MODIS image-based apparent reflectivity model construction method, an MODIS image-based apparent reflectivity model construction system and an MODIS image-based apparent reflectivity model calibration method, and belongs to the technical field of calibration and calibration.
Background
High-precision and high-frequency on-orbit radiometric calibration is the premise and basis for realizing inversion of quantitative remote sensing products. In the on-orbit radiation calibration, a ground or star specific target is used as a reference radiation source during the operation of a satellite, and the conversion relation between the radiation brightness and the image gray value is established to obtain the calibration coefficient of the sensor. The on-orbit radiation calibration method is further classified into an on-satellite calibrator calibration method, a sun calibration method, a moon calibration method, a field calibration method, a scene calibration method, a cross calibration method, and the like according to the difference of reference radiation sources. The on-satellite calibrator calibration method utilizes an on-satellite built-in integrating sphere as a calibration source to realize on-orbit radiation calibration. The sun calibration method and the moon calibration method respectively select the sun and the moon as reference radiation sources, and the on-orbit radiation calibration is realized by adjusting the satellite attitude and acquiring sun and moon images by using the on-satellite device. The three methods all require that a corresponding calibration device is configured on the satellite, the satellite attitude needs to be adjusted during calibration, the control requirement on the satellite load is high, the frequency of the calibration is limited, and most of the current domestic terrestrial satellites do not have the three calibration capabilities. The field calibration method relies on ground satellite-ground synchronous experimental data, and utilizes the actual measurement of ground surface and atmospheric parameters of a calibration field to realize on-orbit radiation calibration. The calibration precision and frequency of the method are limited by the times of satellite-ground synchronous experiments and the weather conditions of transit time, a large amount of manpower and material resources are consumed for each calibration, and the expandable calibration times are limited. The scene calibration method uses a specific natural scene target as a reference radiation source and uses the model radiance and the gray value of an image to realize the on-orbit radiation calibration. According to different natural scenes, the method is divided into a desert scene method, a cloud scene method, an ocean scene method, an polar scene method and the like. The scene scaling method has the advantages of high scaling frequency, low cost and capability of realizing the scaling of historical image data. However, the calibration accuracy is limited by the model of the calibration scene, the method assumes that the scenes in different time phases in different regions have the same model, and actually, the scene images in different time phases in different regions have larger uncertainty due to the differences of observation geometry, atmospheric conditions and earth surface changes, so that the accuracy of the scene calibration method is lower. The cross calibration method selects a reference satellite as a reference radiation source, and realizes the absolute radiation calibration of the sensor by establishing a conversion relation between the target sensor and the reference sensor image. The method takes the difference between different satellite imaging conditions, spectral resolution and the like into consideration, and the calibration precision of the method is directly related to the selected reference image. In general, image pairs under the same observation angle and the same observation time are selected to have higher calibration accuracy.
The existing 'meteorological satellite solar reflection waveband radiometric calibration method based on cold cloud target' (CN105092055B) discloses a meteorological satellite solar reflection waveband radiometric calibration method based on cold cloud target. The method comprises the steps of extracting a cold cloud target object and normalizing the cold cloud reflectivity; and (3) performing on-orbit state monitoring and daily attenuation model building and the like on radiometric calibration response, and utilizing a radiometric reference satellite to realize the recalibration of the solar reflection wave band of the meteorological satellite. The method belongs to one of cross calibration. The invention realizes the continuity and consistency of observation of the satellite instrument; the method is not influenced by weather conditions, saves time and labor, and can conveniently, quickly and accurately obtain satellite response change in real time. However, the method does not take the influence of the reflectivity of the earth surface direction under a large observation angle (the observation angle is more than 30 degrees) into consideration, and is not suitable for cross calibration of image pairs under the large observation angle.
In the existing cross calibration research, the earth surface is assumed to be lambertian, the influence of different satellite observation geometries is ignored, the assumption is suitable when the satellite observation angle is small, but when the satellite has a large observation angle (more than 30 degrees), the influence of the field reflectivity directionality is ignored, and a large error is brought. Although the influence of field-direction reflectivity is considered in the radiation scaling coefficient of the CBERS202 satellite CCD camera (remote sensing science 2006, volume 10, stage 5) after the feature improvement of the ground feature BRDF is considered in the conventional method, the method adopts multi-angle reflectivity data measured on the ground, and only the influence of the ground surface reflectivity directionality is considered, but the influence of the atmospheric layer top apparent reflectivity directionality is not considered. In the existing method, in the research on the in-orbit radiometric calibration method based on deep learning (aerospace return and remote sensing, 2017, vol 38, 2 nd period), a time sequence MODIS is used as a reference satellite, a large number of historical satellite images, historical atmospheric data and historical spectral data of a calibration site are utilized, and a calibration site model closest to a real scene is constructed by learning and screening the data. And simulating the apparent reflectivity of the satellite to be calibrated under the observation geometry corresponding to the imaging moment by using the calibration site model, thereby realizing the absolute radiometric calibration of the sensor. Although the method also obtains a better calibration result, the method does not give a detailed process for realizing the screening of effective data, and whether the method is suitable for the cross calibration of a large-angle image pair or not is not effectively verified.
In summary, in the conventional cross calibration method, the difference of directional reflectivity between satellite image pairs at a large observation angle is less considered, or the influence of the directivity of the apparent reflectivity of the top of the atmospheric layer is replaced by the actually measured reflectivity on the ground, so that the conventional cross calibration method is not suitable for cross calibration of the satellite image pairs at the large observation angle. In the actual calibration process, most of the cross-calibration images have large difference in observation angle. Therefore, the apparent reflectivity model capable of eliminating the large observation angle difference is provided, high-precision cross calibration under a large angle is realized, and the method has important research significance and practical application requirements.
Disclosure of Invention
The invention aims to overcome the defects in the prior art and provides an MODIS image-based apparent reflectivity model construction method, which can eliminate large observation angle difference, obtain apparent reflectivity by establishing an apparent reflectivity model, further complete satellite in-orbit calibration, and has the advantages of high calibration precision, capability of being used for re-calibration of historical data and the like.
Another object of the present invention is to provide an apparent reflectivity model construction system based on MODIS images.
Another objective of the present invention is to provide a calibration method and a calibration system based on the MODIS image.
The above purpose of the invention is mainly realized by the following technical scheme:
the method for constructing the apparent reflectivity model based on the MODIS image comprises the following steps:
obtaining a time sequence MODIS satellite image of a calibration site;
extracting image gray values, image scaling coefficients and image observation geometric information corresponding to the space above a calibration site from the MODIS satellite image according to the position of the calibration site;
obtaining the apparent radiance, the apparent reflectivity and the apparent brightness temperature of the time sequence of the calibration site according to the image calibration coefficient;
obtaining envelope brightness temperature according to the apparent brightness temperature, establishing an envelope brightness temperature curve, and selecting an image of which the difference value of the envelope brightness temperature and the image brightness temperature is less than 10K in the envelope brightness temperature curve;
calculating the variation coefficient of the gray value of the field image of the 1 st wave band in the MODIS visible near infrared region, and selecting the image with the variation coefficient larger than 4%;
taking an image which simultaneously satisfies that the difference value between the brightness temperature of the envelope line and the brightness temperature of the image is less than 10K and the coefficient of variation is more than 4% as an effective calibration field image to obtain the apparent reflectivity and angle information of the image corresponding to the effective calibration field image;
obtaining an apparent reflectivity correction coefficient by utilizing a kernel driving model according to the image apparent reflectivity and the angle information corresponding to the effective calibration field image;
and substituting the apparent reflectivity correction coefficient into the nuclear driving model to establish an apparent reflectivity model for obtaining the apparent reflectivity.
In the above method for constructing an apparent reflectance model based on an MODIS image, the requirements of the calibration site are as follows: (1) the area of the field is not less than 5 kilometers multiplied by 5 kilometers; (2) the calibration field is even and flat and has no vegetation cover; (3) the field has more days in sunny days; (4) the field is located in an arid region; and acquiring the time sequence MOD 021 KM satellite images of the calibration site for more than 1 year.
In the method for constructing the apparent reflectivity model based on the MODIS image, the gray value of the image comprises a wave band 1-7 of a visible near-infrared interval of the MODIS and thermal infrared wave bands 31 and 32; the image scaling coefficients comprise reflectivity scaling coefficients and radiance scaling coefficients of all channels; the image observation geometric information comprises a solar zenith angle, a solar azimuth angle, an observation zenith angle and an observation azimuth angle which correspond to the site position.
In the method for constructing the apparent reflectivity model based on the MODIS image, the apparent radiance calculation formula is as follows:
wherein L isiFor the i-band apparent radiance of MODIS,
and
respectively the gain and the intercept of the apparent radiance scaling factor of the I wave band of the MODIS;
the apparent reflectance calculation formula is as follows:
where ρ isiFor the i-band apparent reflectivity of MODIS,
and
scaling the gain and intercept, θ, of the coefficient for the apparent reflectivity of the MODIS band i, respectivelysIs the corresponding solar zenith angle; DN is the counting value of the remote sensing image;
the apparent brightness temperature calculation formula is as follows:
wherein T is the apparent brightness temperature, h is the Planck constant, K is the Boltzmann constant, c is the speed of light, λ is the wavelength, and L is the apparent radiance.
In the method for constructing the apparent reflectivity model based on the MODIS image, the variation coefficient of the field image gray value of the 1 st wave band in the MODIS visible near infrared interval is calculated, namely the ratio of the field image gray value standard deviation to the mean value is calculated; the angle information comprises an observation zenith angle, an observation azimuth angle, a solar zenith angle and a solar azimuth angle.
In the method for constructing an apparent reflectivity model based on an MODIS image, the kernel-driven model is represented as follows:
wherein the content of the first and second substances,
is the two-way apparent reflectance; k is a radical ofgeoIs a geometric optical nucleus, kvolIs a volume scattering nucleus; thetasIs the solar zenith angle; thetavObserving a zenith angle;
the relative azimuth angle is the relative difference value of the solar azimuth angle and the observation azimuth angle; f. ofiso，fgeo，fvolThe correction coefficients respectively represent the proportion of the isotropic scattering, the geometric optical scattering and the volume scattering.
In the method for constructing the apparent reflectivity model based on the MODIS image, the inversion accuracy of the field model is calculated according to the obtained apparent reflectivity, and the average absolute deviation and the average relative deviation of all clear-day images are obtained, which are specifically as follows:
the calculation formula of the average absolute deviation is as follows:
the average relative deviation is calculated as:
wherein: epsilonaveIs the mean absolute deviation, εrelIs the average relative deviation,
Obtaining the apparent reflectivity of the I wave band of the MODIS according to the established apparent reflectivity model; rhoiThe apparent reflectivity of the I wave band of MODIS.
The system for constructing the apparent reflectivity model based on the MODIS image comprises an image information extraction module, an apparent parameter acquisition module, an effective image screening module, a correction coefficient calculation module and an apparent reflectivity model establishment module, wherein:
an image information extraction module: acquiring a time sequence MODIS satellite image of a calibration site, extracting an image gray value, an image calibration coefficient and image observation geometric information corresponding to the space above the calibration site from the MODIS satellite image according to the position of the calibration site, and transmitting the image gray value, the image calibration coefficient and the image observation geometric information to an apparent parameter acquisition module;
an apparent parameter acquisition module: according to the image scaling coefficient, obtaining the apparent radiance, the apparent reflectivity and the apparent brightness temperature of the time sequence of the scaling field, and sending the apparent radiance, the apparent reflectivity and the apparent brightness temperature to an effective image screening module;
effective image screening module: obtaining envelope brightness temperature according to the apparent brightness temperature, establishing an envelope brightness temperature curve, and selecting an image of which the difference value of the envelope brightness temperature and the image brightness temperature is less than 10K in the envelope brightness temperature curve; calculating the variation coefficient of the gray value of the field image of the 1 st wave band in the MODIS visible near infrared region, and selecting the image with the variation coefficient larger than 4%; selecting an image which simultaneously meets the requirements that the difference value of the envelope brightness temperature and the image brightness temperature is less than 10K and the variation coefficient is more than 4% as an effective calibration field image, obtaining the image apparent reflectivity and angle information corresponding to the effective calibration field image, and sending the image apparent reflectivity and angle information to a correction coefficient calculation module;
a correction coefficient calculation module: obtaining an apparent reflectivity correction coefficient by utilizing a kernel driving model according to the image apparent reflectivity and the angle information corresponding to the effective calibration field image, and sending the apparent reflectivity correction coefficient to an apparent reflectivity model establishing module;
an apparent reflectance model building module: and substituting the apparent reflectivity correction coefficient into the nuclear driving model to establish an apparent reflectivity model for obtaining the apparent reflectivity.
In the above system for constructing an apparent reflectivity model based on an MODIS image, the system further includes an inversion accuracy calculation module, where the inversion accuracy calculation module receives the apparent reflectivity sent by the apparent reflectivity model establishment module, and calculates the inversion accuracy of the field model according to the apparent reflectivity, so as to obtain the average absolute deviation and the average relative deviation of all images on a sunny day, and the method specifically includes:
mean absolute valueThe calculation formula of the deviation is as follows:
the average relative deviation is calculated as:
wherein: epsilonaveIs the mean absolute deviation, εrelIs the average relative deviation,
Obtaining the apparent reflectivity of the I wave band of the MODIS according to the established apparent reflectivity model; rhoiThe apparent reflectivity of the I wave band of MODIS.
A calibration method based on MODIS images adopts the apparent reflectivity model construction method to obtain apparent reflectivity, and obtains a radiometric calibration coefficient of a satellite sensor according to the apparent reflectivity and an image count value to finish calibration.
A calibration system based on an MODIS image comprises the apparent reflectivity model building system and a radiometric calibration coefficient obtaining module, wherein the apparent reflectivity model building system is used for obtaining apparent reflectivity, and the radiometric calibration coefficient obtaining module is used for obtaining a radiometric calibration coefficient of a satellite sensor according to the apparent reflectivity and an image count value to finish calibration.
Compared with the prior art, the invention has the following beneficial effects:
(1) the invention provides a calibration field direction apparent reflectivity model construction method based on a long-time sequence MODIS image.
(2) The invention provides a method for recognizing the influence of thick cloud, thin cloud, snow and the like on the apparent reflectivity by utilizing satellite image data, so that the effective apparent reflectivity of a calibration site is screened and determined, and a directional model constructed on the basis has a better correction effect compared with the existing ground actual measurement spectrum.
(3) The inversion accuracy of the field model can be calculated according to the apparent reflectivity, the average absolute deviation and the average relative deviation of all images in sunny days can be obtained, the application range of various models and the uncertainty of corresponding directional models can be determined, and the application range is wide.
(4) The method can be suitable for on-orbit calibration of the multispectral satellite, particularly cross calibration between large-angle image pairs, and has the advantages of high calibration precision, capability of being used for historical data recalibration and the like.
(5) The method does not need to develop a satellite-ground synchronous experiment, cancels the assumption that the field must have Lambert property, can be used for cross calibration of small-angle images and also can be used for cross calibration of image pairs under large angles, and greatly improves the frequency and the precision of calibration.
Drawings
FIG. 1 is a flow chart of an apparent reflectivity model construction method based on MODIS images;
FIG. 2 is a time series plot of apparent reflectance for a field according to the present invention;
FIG. 3 is a graph of apparent equivalent envelope brightness temperature curves for a field time series according to the present invention;
FIG. 4 is a graph of the effective apparent reflectivity extraction (band 1) for a field according to the present invention;
FIG. 5 is a structural diagram of an apparent reflectivity model construction system based on MODIS images according to the present invention.
Detailed Description
The invention is described in further detail below with reference to the following figures and specific examples:
as shown in fig. 1, the method for constructing an apparent reflectivity model based on an MODIS image of the present invention includes the following steps:
step one, obtaining time sequence MODIS image data of a calibration site. The requirements for the calibration site are as follows: 1) the area of the field is not less than 5 kilometers multiplied by 5 kilometers; 2) the calibration field is even and flat and has no vegetation cover; 3) the site has more sunny days (the number of sunny days in one year is more than 200 days); 4) the field is located in arid area and has small rainfall. And acquiring the time sequence MOD 021 KM satellite images of the calibration site for more than 1 year.
And step two, extracting image gray values, scaling coefficients and observation geometric information corresponding to the space above the scaling site from the MOD 021 km image according to the position of the scaling site. The image gray values comprise wave bands 1-7 of an MODIS visible near-infrared interval and thermal infrared wave bands 31 and 32, the field size is 5 kilometers multiplied by 5 kilometers, namely, the gray values of 5 multiplied by 5 pixel sizes of MOD 021 km images are obtained according to the longitude and latitude of the center of the field. The scaling coefficients comprise reflectivity scaling coefficients and radiance scaling coefficients of all channels; the observation geometry comprises a solar zenith angle, a solar azimuth angle, an observation zenith angle and an observation azimuth angle which correspond to the site position.
And step three, respectively obtaining the apparent radiance, the apparent reflectivity and the apparent brightness temperature of the time sequence of the calibration field according to the calibration coefficient of the MODIS image.
Wherein, the apparent radiance calculation formula
Wherein L isiIs the i-band apparent radiance of MODIS,
and
respectively the gain and the intercept of the apparent radiance scaling factor of the I wave band of the MODIS;
the apparent reflectance calculation formula is as follows:
where ρ isiIs the i-band apparent reflectivity of the MODIS,
and
respectively, the gain and intercept, theta, of the scaling factor of the apparent reflectivity of the i-th band of MODISsIs the corresponding solar zenith angle; DN is the counting value of the remote sensing image;
the apparent brightness temperature calculation formula is as follows
Wherein, T is the apparent bright temperature, h is planck constant, h is 6.626e-34, K is boltzmann constant, K is 1.3806e-23, c is the speed of light, and c is 299792.458e3.λ is the wavelength. Selecting the 31 st waveband of the MODIS to carry out brightness temperature inversion, wherein the corresponding central wavelength is 11.006 um; l is the apparent radiance.
Step four, obtaining the envelope brightness temperature according to the apparent brightness temperature, and establishing an envelope brightness temperature curve TCRFig. 3 is a graph showing the apparent equivalent envelope brightness temperature of a field time series according to the present invention, in which the horizontal axis represents time and the vertical axis represents brightness temperature, and the uppermost curve is the envelope. And selecting an image with the difference value of the brightness temperature of the envelope curve and the brightness temperature of the image being less than 10K in the envelope curve, wherein the image is considered as a clear-day image, and otherwise, the image is a cloud image.
And step five, calculating the variation coefficient of the gray value of the field image in the 1 st wave band of the MODIS visible near infrared interval, namely the ratio of the standard deviation of the gray value of the field image to the mean value, selecting the image with the variation coefficient larger than 4%, if the variation coefficient is larger than 4%, determining that cloud coverage exists, and otherwise, determining that the image is a clear-day image.
And step six, taking the image which simultaneously meets the requirements that the difference value of the brightness temperature of the envelope line and the brightness temperature of the image is less than 10K and the variation coefficient is more than 4% as an effective calibration field image, namely a fine day calibration field image, and obtaining the image apparent reflectivity and angle information corresponding to the effective calibration field image, wherein the angle information comprises information such as an observation zenith angle, an observation azimuth angle, a solar zenith angle, a solar azimuth angle and the like.
And seventhly, obtaining an apparent reflectivity correction coefficient by utilizing the kernel driving model according to the image apparent reflectivity and the angle information corresponding to the effective calibration field image.
The nuclear-driven model is represented as follows:
wherein the content of the first and second substances,
is the two-way apparent reflectance; k is a radical ofgeoIs a geometric optical nucleus, kvolIs a volume scattering nucleus; thetasIs the solar zenith angle; thetavObserving a zenith angle;
the relative azimuth angle is the relative difference value of the solar azimuth angle and the observation azimuth angle; (ii) a f. ofiso，fgeo，fvolThe correction coefficient respectively represents the proportion of the isotropic scattering, the geometric optical scattering and the volume scattering.
And step eight, substituting the obtained apparent reflectivity correction coefficient into the nuclear driving model to establish an apparent reflectivity model for obtaining the apparent reflectivity.
And step nine, obtaining a radiation calibration coefficient of the satellite sensor by utilizing the apparent reflectivity obtained in the step eight and combining the DN count value of the image, and finishing calibration.
The inversion accuracy of the field model can be calculated according to the obtained apparent reflectivity, and the average absolute deviation and average relative deviation of all images on sunny days are obtained, which are as follows:
mean absolute deviationThe calculation formula of (2) is as follows:
the average relative deviation is calculated as:
wherein: epsilonaveIs the mean absolute deviation, εrelIs the average relative deviation,
Obtaining the apparent reflectivity of the I wave band of the MODIS according to the established apparent reflectivity model; rhoiThe apparent reflectivity of the I wave band of MODIS.
Fig. 2 is a time-series graph of apparent reflectivity of a field according to the present invention, and fig. 2 is a graph of apparent reflectivity without effective image screening.
Fig. 4 shows an effective apparent reflectivity extraction (1 st waveband) diagram of a field according to the present invention, and fig. 4 shows an apparent emissivity diagram after effective image screening, where dark dots are discrete dots of Band1 in fig. 2, light dots are discrete dots after effective image screening, and the discrete dots are more concentrated after screening.
The invention also provides an MODIS image-based apparent reflectivity model building system, as shown in FIG. 5, the MODIS image-based apparent reflectivity model building system structure composition diagram of the invention is shown, and the diagram-based model component system comprises an image information extraction module, an apparent parameter acquisition module, an effective image screening module, a correction coefficient calculation module and an apparent reflectivity model building module, wherein:
an image information extraction module: acquiring a time sequence MODIS satellite image of a calibration site, extracting an image gray value, an image calibration coefficient and image observation geometric information corresponding to the space above the calibration site from the MODIS satellite image according to the position of the calibration site, and transmitting the image gray value, the image calibration coefficient and the image observation geometric information to an apparent parameter acquisition module;
and the apparent parameter acquisition module is used for acquiring the apparent radiance, the apparent reflectivity and the apparent brightness temperature of the time sequence of the calibration site according to the image calibration coefficient and sending the apparent radiance, the apparent reflectivity and the apparent brightness temperature to the effective image screening module.
The effective image screening module is used for obtaining envelope brightness temperature according to the apparent brightness temperature, establishing an envelope brightness temperature curve, and selecting an image of which the difference value of the envelope brightness temperature and the image brightness temperature is less than 10K from the envelope brightness temperature curve; calculating the variation coefficient of the gray value of the field image of the 1 st wave band in the MODIS visible near infrared region, and selecting the image with the variation coefficient larger than 4%; and selecting an image which simultaneously meets the requirements that the difference value of the brightness temperature of the envelope line and the brightness temperature of the image is less than 10K and the variation coefficient is more than 4% as an effective calibration field image, obtaining the image apparent reflectivity and angle information corresponding to the effective calibration field image, and sending the image apparent reflectivity and angle information to the correction coefficient calculation module.
And the correction coefficient calculation module obtains an apparent reflectivity correction coefficient by utilizing a kernel driving model according to the image apparent reflectivity and the angle information corresponding to the effective calibration field image and sends the apparent reflectivity correction coefficient to the apparent reflectivity model establishment module.
And the apparent reflectivity model establishing module substitutes the apparent reflectivity correction coefficient into the nuclear driving model to establish an apparent reflectivity model for acquiring the apparent reflectivity.
The functions of the modules are described in the description of the apparent reflectivity model construction method, and are not described in detail here.
The invention realizes the screening and determination of the effective apparent reflectivity of the calibration field, and the directional model constructed on the basis has better correction effect compared with the existing ground actual measurement spectrum. The method does not need to develop a satellite-ground synchronous experiment, cancels the assumption that the field must have Lambert property, can be used for cross calibration of small-angle images and also can be used for cross calibration of image pairs under large angles, and greatly improves the frequency and the precision of calibration.
The above description is only for the best mode of the present invention, but the scope of the present invention is not limited thereto, and any changes or substitutions that can be easily conceived by those skilled in the art within the technical scope of the present invention are included in the scope of the present invention.
Those skilled in the art will appreciate that the invention may be practiced without these specific details.
Claims (11)
1. The method for constructing the apparent reflectivity model based on the MODIS image is characterized by comprising the following steps of: the method comprises the following steps:
obtaining a time sequence MODIS satellite image of a calibration site;
extracting image gray values, image scaling coefficients and image observation geometric information corresponding to the space above a calibration site from the MODIS satellite image according to the position of the calibration site;
obtaining the apparent radiance, the apparent reflectivity and the apparent brightness temperature of the time sequence of the calibration site according to the image calibration coefficient;
obtaining envelope brightness temperature according to the apparent brightness temperature, establishing an envelope brightness temperature curve, and selecting an image of which the difference value of the envelope brightness temperature and the image brightness temperature is less than 10K in the envelope brightness temperature curve;
calculating the variation coefficient of the gray value of the field image of the 1 st wave band in the MODIS visible near infrared region, and selecting the image with the variation coefficient larger than 4%;
taking an image which simultaneously satisfies that the difference value between the brightness temperature of the envelope line and the brightness temperature of the image is less than 10K and the coefficient of variation is more than 4% as an effective calibration field image to obtain the apparent reflectivity and angle information of the image corresponding to the effective calibration field image;
obtaining an apparent reflectivity correction coefficient by utilizing a kernel driving model according to the image apparent reflectivity and the angle information corresponding to the effective calibration field image;
and substituting the apparent reflectivity correction coefficient into the nuclear driving model to establish an apparent reflectivity model for obtaining the apparent reflectivity.
2. The method of constructing an apparent reflectivity model based on an MODIS image according to claim 1, wherein: the requirements of the calibration site are as follows: (1) the area of the field is not less than 5 kilometers multiplied by 5 kilometers; (2) the calibration field is even and flat and has no vegetation cover; (3) the field has more days in sunny days; (4) the field is located in an arid region; and acquiring the time sequence MOD 021 KM satellite images of the calibration site for more than 1 year.
3. The method of constructing an apparent reflectivity model based on an MODIS image according to claim 1, wherein: the image gray values comprise wave bands 1-7 of an MODIS visible near infrared interval and thermal infrared wave bands 31 and 32; the image scaling coefficients comprise reflectivity scaling coefficients and radiance scaling coefficients of all channels; the image observation geometric information comprises a solar zenith angle, a solar azimuth angle, an observation zenith angle and an observation azimuth angle which correspond to the site position.
4. The method of constructing an apparent reflectivity model based on an MODIS image according to claim 1, wherein: the apparent radiance calculation formula is as follows:
wherein L isiFor the i-band apparent radiance of MODIS,
and
respectively the gain and the intercept of the apparent radiance scaling factor of the I wave band of the MODIS;
the apparent reflectance calculation formula is as follows:
where ρ isiFor the i-band apparent reflectivity of MODIS,
and
scaling the gain and intercept, θ, of the coefficient for the apparent reflectivity of the MODIS band i, respectivelysIs the corresponding solar zenith angle; DN is the counting value of the remote sensing image;
the apparent brightness temperature calculation formula is as follows:
wherein T is the apparent brightness temperature, h is the Planck constant, k is the Boltzmann constant, c is the speed of light, λ is the wavelength, and L is the apparent radiance.
5. The method of constructing an apparent reflectivity model based on an MODIS image according to claim 1, wherein: calculating the variation coefficient of the gray value of the field image in the 1 st wave band of the MODIS visible near infrared interval, namely calculating the ratio of the standard deviation and the mean value of the gray value of the field image; the angle information comprises an observation zenith angle, an observation azimuth angle, a solar zenith angle and a solar azimuth angle.
6. The method of constructing an apparent reflectivity model based on an MODIS image according to claim 1, wherein: the kernel-driven model is represented as follows:
wherein the content of the first and second substances,
is the two-way apparent reflectance; k is a radical ofgeoIs a geometric optical nucleus, kvolIs a volume scattering nucleus; thetasIs the solar zenith angle; thetavObserving a zenith angle;
the relative azimuth angle is the relative difference value of the solar azimuth angle and the observation azimuth angle; f. ofiso，fgeo，fvolThe correction coefficients respectively represent the proportion of the isotropic scattering, the geometric optical scattering and the volume scattering.
7. The method of constructing an apparent reflectivity model based on an MODIS image according to claim 1, wherein: calculating the inversion accuracy of the field model according to the obtained apparent reflectivity to obtain the average absolute deviation and average relative deviation of all images on sunny days, which is specifically as follows:
the calculation formula of the average absolute deviation is as follows:
the average relative deviation is calculated as:
wherein: epsilonaveIs the mean absolute deviation, εrelIs the average relative deviation,
Obtaining the apparent reflectivity of the I wave band of the MODIS according to the established apparent reflectivity model; rhoiThe apparent reflectivity of the I wave band of MODIS.
8. An apparent reflectivity model construction system based on MODIS images is characterized in that: the system comprises an image information extraction module, an apparent parameter acquisition module, an effective image screening module, a correction coefficient calculation module and an apparent reflectivity model establishment module, wherein:
an image information extraction module: acquiring a time sequence MODIS satellite image of a calibration site, extracting an image gray value, an image calibration coefficient and image observation geometric information corresponding to the space above the calibration site from the MODIS satellite image according to the position of the calibration site, and transmitting the image gray value, the image calibration coefficient and the image observation geometric information to an apparent parameter acquisition module;
an apparent parameter acquisition module: according to the image scaling coefficient, obtaining the apparent radiance, the apparent reflectivity and the apparent brightness temperature of the time sequence of the scaling field, and sending the apparent radiance, the apparent reflectivity and the apparent brightness temperature to an effective image screening module;
effective image screening module: obtaining envelope brightness temperature according to the apparent brightness temperature, establishing an envelope brightness temperature curve, and selecting an image of which the difference value of the envelope brightness temperature and the image brightness temperature is less than 10K in the envelope brightness temperature curve; calculating the variation coefficient of the gray value of the field image of the 1 st wave band in the MODIS visible near infrared region, and selecting the image with the variation coefficient larger than 4%; selecting an image which simultaneously meets the requirements that the difference value of the envelope brightness temperature and the image brightness temperature is less than 10K and the variation coefficient is more than 4% as an effective calibration field image, obtaining the image apparent reflectivity and angle information corresponding to the effective calibration field image, and sending the image apparent reflectivity and angle information to a correction coefficient calculation module;
a correction coefficient calculation module: obtaining an apparent reflectivity correction coefficient by utilizing a kernel driving model according to the image apparent reflectivity and the angle information corresponding to the effective calibration field image, and sending the apparent reflectivity correction coefficient to an apparent reflectivity model establishing module;
an apparent reflectance model building module: and substituting the apparent reflectivity correction coefficient into the nuclear driving model to establish an apparent reflectivity model for obtaining the apparent reflectivity.
9. The MODIS image-based apparent reflectivity model building system of claim 8, wherein: the inversion accuracy calculation module receives the apparent reflectivity sent by the apparent reflectivity model building module, calculates the inversion accuracy of the field model according to the apparent reflectivity, and obtains the average absolute deviation and the average relative deviation of all images on a sunny day, and the inversion accuracy calculation module specifically comprises the following steps:
the calculation formula of the average absolute deviation is as follows:
formula for calculating average relative deviationComprises the following steps:
wherein: epsilonaveIs the mean absolute deviation, εrelIs the average relative deviation,
Obtaining the apparent reflectivity of the I wave band of the MODIS according to the established apparent reflectivity model; rhoiThe apparent reflectivity of the I wave band of MODIS.
10. A calibration method based on MODIS image is characterized in that: the apparent reflectivity is obtained by the method for constructing the apparent reflectivity model according to any one of claims 1 to 7, and the radiation calibration coefficient of the satellite sensor is obtained according to the apparent reflectivity and the image count value, so that calibration is completed.
11. A calibration system based on MODIS image, its characterized in that: the system comprises the apparent reflectivity model building system of claim 8 or 9 and a radiometric calibration coefficient acquisition module, wherein the apparent reflectivity model building system is used for acquiring apparent reflectivity, and the radiometric calibration coefficient acquisition module is used for acquiring a radiometric calibration coefficient of the satellite sensor according to the apparent reflectivity and an image count value, so as to complete calibration.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811573615.8A CN109883957B (en) | 2018-12-21 | 2018-12-21 | MODIS image-based apparent reflectivity model construction method, system and calibration method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811573615.8A CN109883957B (en) | 2018-12-21 | 2018-12-21 | MODIS image-based apparent reflectivity model construction method, system and calibration method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109883957A CN109883957A (en) | 2019-06-14 |
| CN109883957B true CN109883957B (en) | 2021-08-03 |
Family
ID=66925195
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811573615.8A Active CN109883957B (en) | 2018-12-21 | 2018-12-21 | MODIS image-based apparent reflectivity model construction method, system and calibration method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109883957B (en) |
Families Citing this family (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113495060A (en) * | 2020-03-19 | 2021-10-12 | 国家卫星气象中心(国家空间天气监测预警中心) | Method and device for transmitting radiation reference between channels |
| CN111795936B (en) * | 2020-08-03 | 2021-11-12 | 长安大学 | Multispectral remote sensing image atmospheric correction system and method based on lookup table and storage medium |
| CN113029977B (en) * | 2021-03-11 | 2022-03-15 | 武汉大学 | Automatic cross radiometric calibration method for wide-field-angle multispectral sensor |
| CN112967346B (en) * | 2021-03-30 | 2022-04-29 | 国家卫星海洋应用中心 | Radiometric calibration method and device for remote sensing sensor and electronic equipment |
| CN113254499B (en) * | 2021-05-21 | 2023-09-29 | 国家卫星气象中心(国家空间天气监测预警中心) | Climate data set production method based on long-sequence historical data recalibration |
| CN115422703B (en) * | 2022-07-19 | 2023-09-19 | 南京航空航天大学 | Surface thermal infrared emissivity inversion method based on MODIS data and transducer network |
| CN117726566B (en) * | 2024-02-05 | 2024-05-10 | 中国科学院空天信息创新研究院 | Networking satellite optical load radiation consistency correction method and device and electronic equipment |
| CN118089939A (en) * | 2024-04-26 | 2024-05-28 | 南京信息工程大学 | Satellite spectrum radiometer solar channel radiation calibration method based on cloud target |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103018736A (en) * | 2012-12-03 | 2013-04-03 | 北京航空航天大学 | Satellite-borne remote sensor radiation calibration method based on atmospheric parameter remote sensing retrieval |
| CN105844000A (en) * | 2016-03-18 | 2016-08-10 | 江苏铨铨信息科技有限公司 | MCC ocean surface current inversion method |
| CN106840409A (en) * | 2017-01-23 | 2017-06-13 | 北京师范大学 | A kind of forest fire point detection method based on MODIS |
-
2018
- 2018-12-21 CN CN201811573615.8A patent/CN109883957B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103018736A (en) * | 2012-12-03 | 2013-04-03 | 北京航空航天大学 | Satellite-borne remote sensor radiation calibration method based on atmospheric parameter remote sensing retrieval |
| CN105844000A (en) * | 2016-03-18 | 2016-08-10 | 江苏铨铨信息科技有限公司 | MCC ocean surface current inversion method |
| CN106840409A (en) * | 2017-01-23 | 2017-06-13 | 北京师范大学 | A kind of forest fire point detection method based on MODIS |
Non-Patent Citations (1)
| Title |
|---|
| 基于深度学习的在轨辐射定标方法研究;刘李等;《航天返回与遥感》;20170430;第38卷(第2期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109883957A (en) | 2019-06-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109883957B (en) | MODIS image-based apparent reflectivity model construction method, system and calibration method | |
| Li et al. | Temperature-based and radiance-based validation of the collection 6 MYD11 and MYD21 land surface temperature products over barren surfaces in northwestern China | |
| CN102901516B (en) | A kind of multispectral image radiation correction method based on absolute radiometric calibration | |
| US7171328B1 (en) | Method for measuring thermal properties using a long-wavelength infrared thermal image | |
| CN108362267B (en) | Rape yield loss remote sensing quantitative evaluation method under wet damage stress based on satellite data | |
| CN102853916B (en) | Method and system for conducting remote infrared temperature measurement on coal pile surfaces | |
| CN104535979A (en) | Remote sensing inversion method and system for land cloud optical thickness | |
| Liu et al. | A new method for cross-calibration of two satellite sensors | |
| CN113324656B (en) | Unmanned aerial vehicle-mounted infrared remote sensing earth surface heat anomaly detection method and system | |
| CN103438900A (en) | Three-line-array camera image collaborative absolute radiometric calibration and compensation method | |
| CN109269641B (en) | Multi-sensor cooperative radiometric calibration method for space-painting first satellite | |
| CN108120510B (en) | On-orbit absolute radiation calibration method of optical remote sensor based on reflector array | |
| CN114564767A (en) | Under-cloud surface temperature estimation method based on sun-cloud-satellite observation geometry | |
| CN116519557B (en) | Aerosol optical thickness inversion method | |
| CN102901563B (en) | Method and device for determining land surface emissivity of narrow band and broad band simultaneously | |
| CN112464980A (en) | Method for inverting earth surface temperature by fusing thermal infrared and passive microwave remote sensing data | |
| CN115507959A (en) | Infrared radiation characteristic analysis method for target detection | |
| CN114970214A (en) | Aerosol optical thickness inversion method and device | |
| CN110689505B (en) | Scene-based satellite-borne remote sensing instrument self-adaptive correction method and system | |
| CN110009584B (en) | Multispectral remote sensing image atmospheric correction system and method based on reference spectrum matching | |
| CN105183989A (en) | Landsat 8 satellite data surface reflectance inversion method | |
| Lu et al. | An automatic radiometric cross-calibration method for wide-angle medium-resolution multispectral satellite sensor using landsat data | |
| CN110702228B (en) | Edge radiation correction method for aviation hyperspectral image | |
| CN115269549A (en) | Atmospheric water vapor inversion method coupling physics-statistics-deep learning | |
| CN114296061A (en) | Cross calibration method based on multivariate variable detection and different radiation transmission models |
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 |