Remote sensing estimation method for sunny atmosphere downlink long-wave radiation corrected by proximity effect under cloud influence

Abstract

The invention relates to the technical field of quantitative remote sensing, and discloses a remote sensing estimation method for sunny atmosphere downlink long-wave radiation with proximity effect correction under the influence of cloud, and the method comprises the following steps: S1, obtaining an atmosphere top radiance sample and a corresponding atmosphere downlink long-wave radiation sample under a sunny condition; s2, establishing a mathematical relationship between atmospheric top radiance and atmospheric downlink long-wave radiation; s3, acquiring atmospheric top radiance data of the satellite image, and calculating atmospheric downlink long-wave radiation under a clear sky condition; s4, calculating downlink long-wave radiation under the cloudless condition and the cloudless condition, and a difference value of the downlink long-wave radiation under the cloudless condition and the cloudless condition, and defining as a cloud radiation source item; s5, analyzing the propagation process of radiation from the bottom of the cloud layer to the adjacent clear sky, and normalizing to generate an atmospheric point spread function; s6, performing convolution calculation on the cloud radiation source item and the atmospheric point spread function to obtain an influence value of cloud radiation on an adjacent clear sky; and S7, correcting the atmospheric downlink long wave radiation to obtain an estimated value with higher precision.

Description

Technical Field

[0001] This invention relates to the field of quantitative remote sensing technology, specifically to a remote sensing estimation method for clear-sky atmospheric downward longwave radiation with proximity effect correction under cloud influence. Background Technology

[0002] Downward longwave radiation (DLR) is a crucial component of the Earth's surface energy budget, representing the heat energy transferred from the atmosphere to the surface. DLR plays a vital role in land-atmosphere interactions, energy balance estimation, and climate change monitoring. Clouds cover most of the globe and significantly enhance DLR in the longwave band through absorption and re-emission processes, becoming one of the main sources of error in remote sensing DLR estimation. Although numerous studies have focused on the DLR enhancement effect directly beneath clouds, the radiative influence of clouds extends beyond their projected area; adjacent clear-sky areas may also be affected by the non-local radiation of clouds. Therefore, neglecting the "proximity effect" of clouds can lead to errors in DLR estimation for clear-sky areas.

[0003] Currently, DLR estimation methods mainly rely on three approaches: ground-based radiation observations, numerical simulation products, and remote sensing estimation methods. Ground-based observation data offers high accuracy and temporal resolution, but due to the sparse distribution of stations, it is difficult to achieve large-scale spatial modeling of DLR. While numerical simulation products have advantages in terms of temporal and spatial coverage, their spatial resolution is low, and they may introduce significant errors under strong non-uniform cloud fields. Remote sensing methods mainly include physical models, empirical models, hybrid models, and data-driven methods. Remote sensing data can provide large-scale, long-term DLR estimations, but existing methods are generally based on a binary classification of clouds and clear skies, neglecting the non-local radiative impact of clouds on adjacent clear-sky areas. Under clear-sky conditions, the mathematical relationship between radiance at the top of the atmosphere and downward longwave radiation can be estimated using satellite remote sensing data, and has become an important tool for remote sensing DLR estimation. By establishing a mathematical model between radiance and downward longwave radiation, DLR under clear-sky conditions can be effectively predicted. However, existing remote sensing DLR estimation methods typically only consider the situation under clear-sky conditions, ignoring radiative disturbances caused by clouds. This method has limitations in practical applications, especially under cloudy conditions, where the radiative impact of clouds on nearby clear-sky pixels is severely underestimated. Clouds significantly affect surface radiation flux by enhancing downward longwave radiation, making DLR estimation more complex. Within a radius of several kilometers around clouds, clear-sky pixels are affected by cloud radiation; this effect is known as the proximity effect. Ignoring the proximity effect will result in significant biases in remote sensing estimation of DLR, especially in areas with strong cloud fields. Therefore, accurately quantifying the radiative impact of clouds on nearby clear-sky pixels is crucial to improving the accuracy of DLR estimation.

[0004] Existing radiative transfer models can accurately simulate downlink longwave radiation under various combinations of atmospheric and cloud parameters and are widely used in the development and validation of DLR products. The atmospheric point spread function (PSF) method, as a means of describing the spatial diffusion characteristics of radiation, has been applied in the visible and near-infrared bands. However, in the longwave radiation band, existing remote sensing DLR estimation methods fail to consider the spatial diffusion characteristics of cloud radiation, making it difficult to accurately characterize the radiative changes of neighboring clear-sky pixels under cloudy conditions. Therefore, a new method is urgently needed to correct the DLR estimation of clear-sky pixels under cloudy conditions, accurately considering the influence of clouds on neighboring clear-sky pixels, thereby improving the accuracy of DLR estimation. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a remote sensing estimation method for down-atmosphere longwave radiation under clear sky conditions with cloud cover and proximity effect correction. This method has advantages such as high compatibility of radiation correction and high accuracy of remote sensing estimation, and solves the problem that traditional remote sensing estimation methods for down-atmosphere longwave radiation do not fully consider the cloud proximity effect bias.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a remote sensing estimation method for clear-sky atmospheric downwind longwave radiation with proximity effect correction under cloud cover, comprising the following steps: S1. Obtain the top atmospheric radiance and corresponding downward longwave radiation samples under clear sky conditions; The data samples can be obtained in two ways: First, by utilizing atmospheric profile data from a standard atmospheric profile database for different climate types, atmospheric height, air pressure, temperature, and water vapor content, and combining this with the channel response function of satellite sensors, simulations are performed using a radiative transfer model to obtain samples of the top atmospheric radiance and corresponding downward longwave radiation under clear-sky conditions. The standard atmospheric profile database can be a built-in profile library of TIGR, SeeBor, or MODTRAN. During the simulation, by setting different altitudes, observation zenith angles, and surface types, and combining this with the channel response function of satellite sensors, radiative transfer models (MODTRAN or LibRadtran) are used to calculate samples of the top atmospheric radiance and downward longwave radiation under multiple scenarios.

[0007] S2. Establish the mathematical relationship between the radiance at the top of the atmosphere and the downward longwave radiation of the atmosphere; Based on the sample data obtained in step S1, a mathematical relationship between atmospheric top radiance and atmospheric downward longwave radiation is established. This mathematical relationship can be realized through regression analysis. The radiance and atmospheric downward longwave radiation samples under different altitudes and observation zenith angles are fitted, the regression coefficients are solved, and a lookup table is established for subsequent estimation of atmospheric downward longwave radiation under satellite observation conditions.

[0008] S3. Acquire satellite imagery data and calculate atmospheric downwave radiation under clear-sky conditions; In this step, based on the mathematical relationship established in step 2, the atmospheric top radiance data of the acquired satellite image is used, along with the observation zenith angle and altitude of the satellite image, to obtain the downward longwave radiation of the atmosphere under clear sky conditions.

[0009] S4. Calculate the cloud radiation source term; To quantitatively describe the contribution of cloud radiation to the downlink longwave radiation in adjacent clear-sky areas, this invention proposes a method for calculating cloud radiation in step S4. This method calculates the difference between downlink longwave radiation under cloudless and cloudy conditions to obtain the change in downlink longwave radiation caused by clouds, which is defined as the cloud radiation source term.

[0010] S41. Select suitable atmospheric profile data. set up The target pixel is selected, and based on the atmospheric type of the region to which the pixel belongs, the corresponding atmospheric profile data is extracted from a standard atmospheric profile database or obtained from real-time profile data products. The atmospheric profile data includes information on atmospheric height, air pressure, temperature, and water vapor content. Simultaneously, the spatial location and cloud attributes of neighboring cloud pixels are extracted using satellite imagery, including cloud top height, cloud optical thickness, and cloud facies, and the observed zenith angle and surface elevation information are acquired concurrently.

[0011] S42. Obtain atmospheric downward longwave radiation under cloudless conditions, i.e., atmospheric downward longwave radiation under clouds.

[0012] Based on the cloud top height, cloud optical thickness, and cloud phase obtained in step S41, the cloud base height is calculated. Using this height as the cutoff point, the atmospheric height, air pressure, temperature, and water vapor profiles from the cloud base to the ground surface are extracted. The obtained profile data, altitude, and observed zenith angle are input into the radiative transfer model (such as MODTRAN or LibRadtran) to calculate the downward longwave radiation of the atmosphere under cloudless conditions, i.e., the downward longwave radiation of the atmosphere below the clouds.

[0013] S43. Obtain downlink longwave radiation under cloud conditions.

[0014] By inputting the cloud base height, observed zenith angle, altitude, cloud attribute parameters, cloud type, and atmospheric profile data obtained in step S42, along with the data obtained in step S41, into the radiative transfer model, the cloud-to-ground longwave radiation is obtained.

[0015] The cloud radiation source term can be obtained by subtracting the cloud-below longwave radiation calculated in step S42 from the cloud-below longwave radiation calculated in step S44 and step S43.

[0016] S5. Using the same atmospheric profile data from step S4, input the data into the radiative transfer model to obtain the atmospheric transmittance of each layer, and calculate the normalized atmospheric point spread function. ; In this step, firstly, based on the atmospheric profile data from step S41, the transmittance of each atmospheric layer is obtained using a radiative transfer model. Then, combining the cloud base height and atmospheric transmittance obtained in step S42, the normalized atmospheric point spread function is further calculated. Normalized atmospheric point spread function The propagation process of cloud radiation from the cloud base to adjacent clear-sky pixels was quantified, and its normalization formula is as follows: In the formula, For cloud image elements, For wavelength, The scattering phase function, The scattering angle is... This represents the atmospheric transmittance from a cloud pixel to a scattering point. The atmospheric optical thickness from the cloud pixel to the scattering point. , For the launch angle, The atmospheric transmittance from the scattering point to the ground. The atmospheric optical thickness from the scattering point to the ground. , To observe the zenith angle, The distance from the cloud pixel to the scattering point. Which layer of atmosphere is it? The number of atmospheric layers. This represents the number of pixels within the effective radius of the proximity effect, centered on the cloud pixel.

[0017] S6. Combine the cloud radiation source term obtained in step S4 with the normalized atmospheric point diffusion function obtained in step S5. Convolution calculations are performed to obtain the spatial distribution of cloud radiation source terms and the normalized atmospheric point spread function. Spatial attenuation characteristics, and target pixels The expression for the influence value of radiation received from nearby clouds is: In the formula, For target pixel The influence value of radiation received from nearby clouds, The set of neighboring cloud pixels surrounding the target pixel. to For wavelength range, The cloud radiation source term obtained in step S4, The wavelength obtained in step S5 is Normalized weight kernel, representing cloud pixels Within the neighborhood centered, for offsets of The weights of the target pixels, where, for , for .

[0018] S7. Obtain the target pixel by performing step S3. The atmospheric downdraft longwave radiation value and the cloud radiation value obtained in step S6 for the target pixel The influence values ​​are added together to obtain the corrected estimate of the downward longwave radiation under cloud cover in clear skies. This effectively improves the accuracy of the estimation of downward longwave radiation under cloud cover and reduces the estimation bias in traditional methods. The calculation formula is as follows: In the formula, This is the corrected image of longwave radiation descending through clear skies under cloud cover. For target pixel The value of the influence of radiation received from nearby clouds.

[0019] Compared with existing technologies, this invention provides a remote sensing estimation method for clear-sky atmospheric downwind longwave radiation with proximity effect correction under cloud cover, which has the following advantages: 1. This invention, by introducing the convolution of cloud radiation source terms and atmospheric point spread functions, can effectively correct the proximity effect of clear-sky pixels under cloudy conditions, improve the accuracy of downlink longwave radiation estimation under cloudy conditions, break through the limitation of traditional methods that rely solely on binary classification of clear sky and cloudy, and make up for the defect of ignoring the radiation influence of clouds on adjacent clear-sky areas.

[0020] 2. This invention improves the accuracy of down-flying longwave radiation estimation under cloudy conditions by accurately estimating atmospheric down-flying longwave radiation under clear sky conditions and introducing a proximity effect correction mechanism. It overcomes the limitations of traditional methods that rely solely on a binary classification of clear sky and cloudy conditions, and avoids neglecting the influence of clouds on radiation in adjacent clear sky areas. It has important application prospects in fields such as atmospheric radiation budget and climate change monitoring, and also provides new ideas for the application of remote sensing technology under complex meteorological conditions. It has the advantages of high radiation correction compatibility and high remote sensing estimation accuracy. Attached Figure Description

[0021] Figure 1 This is a diagram illustrating the steps of the method of the present invention; Figure 2 This is a schematic diagram of the atmospheric downward longwave radiation flux of the present invention. Detailed Implementation

[0022] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0023] Example Please see Figure 1 and Figure 2 This invention provides a remote sensing estimation method for clear-sky atmospheric downward longwave radiation with proximity effect correction under cloud cover, comprising the following steps: S1. Simulate the radiance at the top of the atmosphere and the corresponding downward longwave radiation samples under clear sky conditions using a radiative transfer model; First, atmospheric profile data from the standard atmospheric profile database TIGR is acquired. This data includes atmospheric parameters such as atmospheric height, air pressure, temperature, water vapor content, and ozone content. Then, atmospheric profile data under clear-sky conditions is selected, and the input parameters for the radiative transfer model are set. The altitude is set to 0 km to 5.9 km with an interval of 0.5 km, the observation zenith angle is set to 0° to 70° with an interval of 10°, and the land surface type is set to desert, farmland, forest, fresh snow, grassland, seawater, soil, city, and wetland. The satellite sensor channel response function is also set. These parameters are input into the radiative transfer model, and the top atmospheric radiance and the corresponding downward longwave radiation are output, which are the top atmospheric radiance sample and the downward longwave radiation sample. In this embodiment, the radiative transfer model is set to MODTRAN, and the satellite sensor channel response functions are the response functions of channels 28, 29, 31, 33, 34, and 36 of MODIS. In practical applications, other models with infrared band downlink longwave radiation calculation capabilities can be selected according to the usage scenario.

[0024] S2. In this step, based on the radiance samples of the top of the atmosphere and the down-atmosphere longwave radiation samples obtained in step S1 under clear sky conditions, a mathematical relationship between the radiance of the top of the atmosphere and the down-atmosphere longwave radiation is established by looking up a table and using a regression model. In this embodiment, based on the samples obtained in S1, an improvement is made to the parameterization method for downdraft longwave radiation in clear-sky atmosphere. The least squares method is used to fit the atmospheric top radiance and downdraft longwave radiation under different altitudes and observation zenith angles, the regression coefficients are solved, and a lookup table is established. The specific formula is as follows: In the formula, For the target pixel, It is the regression coefficient. It is a MODIS channel The measured radiance at the top of the atmosphere, ; The following lookup table experimental data is shown in Table 1: Table 1. Lookup table of regression coefficients for clear-sky atmospheric downward longwave radiation at different observation zenith angles under DEM=3km condition. S3. In this step, based on the mathematical relationship established in step S2, the atmospheric top radiance data in the satellite image is obtained, and the atmospheric downward longwave radiation under clear sky conditions is calculated. In this embodiment, MODIS satellite imagery data MOD021KM, MOD03, and MOD35 are used, which can acquire accurate radiance data over a large area to estimate atmospheric downdraft longwave radiation under clear-sky conditions. It is 179.43 W / m²; The experimental data of atmospheric downward longwave radiation under clear sky conditions are shown in Table 2: Table 2. Estimation of down-atmospheric longwave radiation under clear sky based on MODIS thermal infrared radiance (target pixel 38.05°N, 100.46°E) S4. In this step, atmospheric profile data is obtained based on the spatial location of neighboring cloud pixels and atmospheric conditions of the target pixel. The atmospheric profile may be derived from MODIS data or other applicable remote sensing atmospheric products, and includes key parameters such as atmospheric height, air pressure, temperature, water vapor content, and ozone content.

[0025] In this embodiment, the target area is located at 38.05°N, 100.46°E, with an altitude of approximately 3.03 km, belonging to a typical mid-latitude plateau region. According to the atmospheric type classification of the MODTRAN model, this region corresponds to the mid-latitude summer atmospheric type; therefore, the corresponding mid-latitude summer standard atmospheric profile in the MODTRAN radiative transfer model is selected as the input atmospheric data.

[0026] Cloud characteristic parameters of neighboring cloud pixels were extracted using the MOD06 cloud product. The results showed a cloud top height of 3.55 km, a cloud optical thickness of 5.25 km, and a cloud facies type of water cloud. The cloud thickness was estimated to be approximately 0.637 km using the Minnis parametric method, leading to a calculated cloud base height of 2.913 km (cloud top height minus cloud thickness). Further atmospheric data below the cloud base, including atmospheric height, pressure, temperature, and water vapor content, were extracted and combined with an observed zenith angle of 42.12°, serving as input data for downward longwave radiation under clear-sky conditions. Radiative transfer calculations were performed using the MODTRAN radiative transfer model to obtain the results of downward longwave radiation under the cloud conditions, as shown below. Figure 2 As shown.

[0027] Furthermore, based on the attribute information of cloud pixels in the image, the cloud facies type was determined to be water cloud. In the model, the cloud type was set to Cumulus, and the cloud base height was set to 2.913 km, cloud thickness to 0.637 km, cloud optical thickness to 5.25, and the observed zenith angle to 42.12°. The mid-latitude summer atmospheric profile, consistent with the aforementioned model, was input into the MODTRAN radiative transfer model to calculate the down-current longwave radiation results under cloud conditions, as shown below. Figure 2 As shown. Finally, by calculating the difference in downward longwave radiation under cloudy and clear-sky conditions, the cloud radiation source term value is obtained, as shown. Figure 2 As shown.

[0028] S5. In this embodiment, the parameters of each atmospheric layer are obtained from the atmospheric profile data in S4, including altitude, air pressure, temperature, water vapor content, and ozone content. Using a radiative transfer model, the parameters of each atmospheric layer are input into the model. Other settings are consistent with the downward longwave radiation of the atmosphere below the clouds in S4. The transmittance of each atmospheric layer is obtained, and the cloud base height obtained in step S4 is substituted into the atmospheric point spread function. In the atmosphere, the point spread function The propagation process of radiation from clouds to nearby clear-sky pixels was quantified, and its normalization formula is as follows: In the formula, For cloud image elements, For wavelength, The scattering phase function, The scattering angle is... This represents the atmospheric transmittance from a cloud pixel to a scattering point. The atmospheric optical thickness from the cloud pixel to the scattering point. , For the launch angle, The atmospheric transmittance from the scattering point to the ground. The atmospheric optical thickness from the scattering point to the ground. , To observe the zenith angle, The distance from the cloud pixel to the scattering point. Which layer of atmosphere is it? The number of atmospheric layers is determined by the number of atmospheric layers between the cloud base and the ground, as described in step S4. This indicates the number of pixels within the effective radius of the proximity effect, centered on the cloud pixel. S6. Combine the cloud radiation source term obtained in step S4 with the atmospheric point spread function obtained in step S5. Convolution calculations are performed to obtain the spatial distribution of cloud radiation source terms and the atmospheric point spread function. Spatial attenuation characteristics, and target pixels The expression for the influence value of radiation received from nearby clouds is: In the formula, For target pixel The influence value of radiation received from nearby clouds, The set of neighboring cloud pixels surrounding the target pixel. to The wavelength range is from 4μm to 100μm. The cloud radiation source term obtained in step S4, The wavelength obtained in step S5 is Normalized weight kernel, representing cloud pixels Within the neighborhood centered, for offsets of The weights of the target pixels, where, for , for ; The following is the atmospheric point spread function. The experimental data are shown in Table 3: Table 3 shows the results of the 5×5 neighborhood convolution of the cloud radiation source term and the normalized atmospheric point spread function obtained in step S4. S7. In step S3, the estimated value of the downward longwave radiation of the target pixel was obtained as 179.43 W / m². The influence value of the radiation received by the target pixel from the nearby cloud obtained in step S6 is then used. Based on the fact that the horizontal distance between the cloud pixel and the target pixel is approximately 2.77 km, the influence value of the radiation received by the target pixel from the nearby cloud, calculated in step S6, is... Approximately 0.981 W / m², for the target pixel The downward longwave radiation of the atmosphere is corrected, and the formula for calculating the estimated value of the downward longwave radiation of the atmosphere under cloud cover in clear sky after correction is as follows: In the formula, The value is 180.411 W / m², representing the corrected downwind longwave radiation under cloud cover in clear-sky conditions.

[0029] This embodiment calculates the cloud radiation source term and the atmospheric point spread function. Furthermore, by performing convolution operations on the two, the impact of cloud radiation on neighboring clear-sky pixels was accurately quantified. By adding the cloud radiation impact value received by the target pixel to the estimated value of atmospheric down-longwave radiation under clear-sky conditions, the estimation correction of atmospheric down-longwave radiation under cloud influence was realized, which can significantly improve the accuracy of atmospheric down-longwave radiation estimation, especially in the case of clouds, and reduce the estimation bias in traditional remote sensing methods.

[0030] The above formulas are all derived from software simulation using a large amount of data and are selected to be close to the actual values. The coefficients in the formulas are set by those skilled in the art according to the actual situation. The above description is only a preferred embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any equivalent substitutions or changes made by those skilled in the art within the technical scope disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the protection scope of the present invention.

Claims

1. A remote sensing estimation method for clear-sky atmospheric downward longwave radiation with proximity effect correction under cloud influence, characterized in that, Includes the following steps: Step S1: Using atmospheric profile data of different climates, altitudes and water vapor in the standard atmospheric profile database, combined with the satellite sensor channel response function, the atmospheric top radiance sample and the corresponding atmospheric downward longwave radiation sample under clear sky conditions are obtained by simulation calculation through the radiative transfer model. Step S2: Based on the sample data obtained in Step S1, establish a mathematical relationship between the top atmospheric radiance and the downward atmospheric longwave radiation. The mathematical relationship is constructed by analyzing paired sample data under different profiles, altitudes, observation zenith angles, and surface types. It can be implemented using lookup tables and regression models to characterize the quantitative relationship between radiance and downward atmospheric longwave radiation. Step S3: Obtain the atmospheric top radiance data from the satellite imagery and calculate the downward longwave radiation under clear-sky conditions based on the mathematical relationship in Step S2. Step S4: Select the corresponding atmospheric profile data according to the atmospheric type of the satellite image area in Step S3, calculate the downlink longwave radiation under cloudless and cloudy conditions respectively, and use the difference between the two as the cloud radiation source term. Step S5: Based on the spatial location of cloud pixels in the satellite imagery from Step S4, establish a normalized atmospheric point spread function. ; Step S6: Combine the cloud radiation source term obtained in step S4 with the normalized atmospheric point diffusion function obtained in step S5. Convolution calculations are performed to obtain the impact of cloud radiation on nearby clear sky. Step S7: The influence value of cloud radiation on the nearby clear sky obtained in step S6 is used to correct the downward longwave radiation of the atmosphere obtained in step S3, and finally the corrected estimated value of downward longwave radiation of the atmosphere under the influence of clouds is obtained.

2. The remote sensing estimation method for clear-sky atmospheric downward longwave radiation under cloud cover with proximity effect correction as described in claim 1, characterized in that: In step S1, a clear-sky atmospheric profile from the standard atmospheric profile database is selected, and different altitudes, observation zenith angles, and surface types are set. The atmospheric profile, altitude, observation zenith angle, surface type, and satellite sensor channel response function parameters are input into the radiative transfer model for simulation calculation to obtain the atmospheric top radiance and corresponding atmospheric downward longwave radiation samples under different scenarios.

3. The remote sensing estimation method for clear-sky atmospheric downward longwave radiation under cloud cover with proximity effect correction as described in claim 2, characterized in that: In step S2, based on the simulated sample data obtained in step S1, a regression analysis method is used to fit the atmospheric top radiance and atmospheric downward longwave radiation samples under different altitude and observation zenith angle conditions, solve the regression coefficients and establish a lookup table. The regression analysis method is a linear regression algorithm to obtain the quantitative relationship between atmospheric top radiance and atmospheric downward longwave radiation under different altitude and observation zenith angle combinations.

4. The remote sensing estimation method for clear-sky atmospheric downward longwave radiation under cloud cover with proximity effect correction as described in claim 3, characterized in that: In step S3, the satellite imagery data comes from medium resolution imaging spectrometer images and geostationary meteorological satellite infrared imagers. The atmospheric top radiance data of the satellite remote sensing images are obtained, and the atmospheric down-current longwave radiation under clear sky conditions is calculated based on the mathematical relationship established in step S3. The calculation process includes: inputting the satellite radiance data and the corresponding observation zenith angle and altitude into the mathematical relationship model to obtain clear sky atmospheric down-current longwave radiation that matches the satellite observation conditions.

5. The remote sensing estimation method for clear-sky atmospheric downward longwave radiation under cloud cover with proximity effect correction as described in claim 4, characterized in that: In step S4, the cloud radiation source term calculation process is as follows: S41, Let The target pixel is used to obtain the corresponding atmospheric profile data based on its atmospheric type, including atmospheric height, air pressure, temperature, and water vapor information. Then, through satellite imagery, the cloud attribute information, observation zenith angle, and altitude of the cloud pixels near the target pixel are obtained. Among them, the cloud attribute information includes cloud top height, cloud optical thickness, and cloud phase. S42. Calculate the cloud base height based on the cloud top height and cloud optical thickness extracted in step S41, and extract the atmospheric profile data obtained in step S41 based on the cloud base height. Input the data into the radiative transfer model according to the altitude, observation zenith angle, and satellite sensor channel response function to obtain the atmospheric downward longwave radiation under cloudless conditions, that is, the atmospheric downward longwave radiation from the bottom of the cloud layer to the ground surface. S43. Input the cloud base height obtained in step S42, as well as the altitude, observation zenith angle, atmospheric profile data and cloud type in step S41 into the radiative transfer model to obtain the downlink longwave radiation under cloud conditions, i.e., the downlink longwave radiation under clouds. S44. Calculate the difference in downlink longwave radiation under cloudless and cloudy conditions to obtain the change in downlink longwave radiation caused by the cloud layer, and define it as the cloud radiation source term.

6. The remote sensing estimation method for clear-sky atmospheric downward longwave radiation under cloud cover with proximity effect correction as described in claim 5, characterized in that: In step S5, the transmittance of each atmospheric layer in the atmospheric profile is obtained using a radiative transfer model, and the atmospheric point spread function is further calculated by combining the cloud base height and the calculated atmospheric transmittance. Among them, atmospheric point spread function The normalization formula used to characterize the propagation process of radiation from the bottom of the cloud layer to adjacent clear-sky pixels is: In the formula, For cloud image elements, For wavelength, The scattering phase function, The scattering angle is... This represents the atmospheric transmittance from a cloud pixel to a scattering point. The atmospheric optical thickness from the cloud pixel to the scattering point. , For the launch angle, The atmospheric transmittance from the scattering point to the ground. The atmospheric optical thickness from the scattering point to the ground. , To observe the zenith angle, The distance from the cloud pixel to the scattering point. Which layer of atmosphere is it? The number of atmospheric layers. This represents the number of pixels within the effective radius of the proximity effect, centered on the cloud pixel.

7. The remote sensing estimation method for clear-sky atmospheric downward longwave radiation under cloud cover with proximity effect correction as described in claim 6, characterized in that: Step S6 includes: combining the cloud radiation source term obtained in step S4 with the atmospheric point spread function obtained in step S5. Convolution calculations are performed to obtain the spatial distribution of cloud radiation source terms and the atmospheric point spread function. Spatial attenuation characteristics, and target pixels The expression for the influence value of radiation received from nearby clouds is: In the formula, For target pixel The influence value of radiation received from nearby clouds, The set of neighboring cloud pixels surrounding the target pixel. to For wavelength range, The cloud radiation source term obtained in step S4, The wavelength obtained in step S5 is Normalized weight kernel, representing cloud pixels Within the neighborhood centered, for offsets of The weights of the target pixels, where, for , for .

8. The remote sensing estimation method for clear-sky atmospheric downward longwave radiation under cloud cover with proximity effect correction as described in claim 7, characterized in that: In step S7, the influence value of the target pixel receiving radiation from nearby clouds obtained in step S6 is... For target pixels The downward longwave radiation of the atmosphere is corrected, and the formula for calculating the estimated value of the downward longwave radiation of the atmosphere under cloud cover in clear sky after correction is as follows: In the formula, This is the corrected image of longwave radiation descending through clear skies under cloud cover. For target pixel The value of the influence of radiation received from nearby clouds.

Images