A method for simulating brightness temperature in snow and ice environments

Abstract

This invention discloses a brightness temperature simulation method for observing ice and snow environments, belonging to the field of image processing technology. The method includes: Step 1, constructing a physical model of ice and snow layers using the SMRT model, and setting the electromagnetic theory model and ice and snow microstructure of the SMRT model; Step 2, calculating the surface emissivity ε corresponding to the uplink radiation; Step 3, acquiring mixed pixels, correcting the surface emissivity of the mixed pixels to obtain the emissivity of the mixed pixels; Step 4, using the RTTOV model combined with ERA5 atmospheric profile data and near-surface parameters to generate the simulated brightness temperature for Arctic ice and snow environments. This invention's brightness temperature simulation method for observing ice and snow environments utilizes the SMRT model to simulate the transmission process of microwave radiation in multiple layers of snow, sea ice, and other media, allowing for a better understanding of the physical properties of snow and its impact on microwave signals. By effectively coupling the SMRT and RTTOV models, more accurate model support can be provided for remote sensing inversion of Arctic ice and snow environment parameters.

Description

Technical Field

[0001] This invention belongs to the field of image processing technology, specifically, it relates to a method for simulating brightness and temperature in Arctic ice and snow environments based on a coupled SMRT and RTTOV model. Background Technology

[0002] Satellite-observed brightness temperature data is a direct product of remote sensing observations, accurately reflecting the radiation characteristics of the Earth's surface and atmosphere in the microwave band. However, the brightness temperature signal received by satellite sensors is a comprehensive result of the interaction between surface physical parameters (such as temperature, salinity, density, and microstructure) and the atmospheric environment, often making it difficult to directly deduce the specific contributions of each factor. Microwave radiative transfer models can quantitatively describe the influence of the internal structure of the snow and ice layer and the characteristics of the underlying surface on microwave radiation signals, thereby establishing a physical correlation between satellite-observed brightness temperature and surface physical parameters. Brightness temperature simulation based on microwave radiative transfer models not only helps to understand the radiative transfer mechanism of microwaves in snow and ice media, but also provides theoretical support for the interpretation of satellite remote sensing data, the development of parameter inversion algorithms, and the assimilation of multi-source observation data, making it an important means of assessing the impact of climate change on the snow and ice environment.

[0003] Based on existing radiative transfer models, some scholars have conducted comparative analyses of different models. Due to frequency or underlying surface limitations, some models cannot be effectively applied to brightness temperature simulations on Arctic ice. For example, HUT, DMRT-ML, and WALOMIS cannot simulate ice and snow environments. While MEMLS and RTTOV can simulate ice and snow environments, the former requires coupling with other sea ice models and cannot simulate the L-band, while the latter can only set simple ice and snow classifications, lacking the characterization of microscopic ice and snow processes and the accuracy of surface simulation.

[0004] With the introduction of the new generation of snow microwave radiative transfer model (SMRT), scholars have conducted in-depth research and application on it in recent years. The main problems at present are: (1) Since there are generally mixed pixels of "sea ice snow + seawater" in satellite observation pixels, and the SMRT model assumes that the underlying surface is pure sea ice, the simulated brightness temperature is higher than the actual situation, and it cannot reflect the influence of seawater on brightness temperature simulation and emissivity calculation. Existing studies mostly introduce sea ice concentration data to correct the simulated brightness temperature, but the sea ice concentration product is limited by spatial resolution and inversion algorithm, and it is easy to misjudge some inter-ice channels as pure ice areas. (2) Some inter-ice channels have re-icing phenomenon, so the conventional correction method based on sea ice concentration is no longer applicable to inter-ice channels. (3) A single SMRT model often ignores atmospheric interference.

[0005] Based on this, the main technical problem addressed by this invention is how to propose a method that combines surface radiation simulation with atmospheric transport processes to simulate satellite-observed brightness temperature in the C-K band of the Arctic ice and snow environment, so that the simulated data can better match the actual transport process of brightness temperature in the Arctic ice and snow environment. This is important for understanding the radiation transport mechanism of microwaves in ice and snow media, and can also provide theoretical support for the interpretation of satellite remote sensing data, the development of parameter inversion algorithms, and the assimilation of multi-source observation data. Summary of the Invention

[0006] To address the technical problems of brightness temperature simulation in snow and ice environments, which cannot resolve the specific contributions of surface and atmospheric factors, and the limitations of existing radiative transfer models in terms of simulation band and accuracy in characterizing microscopic processes of snow and ice and simulating the surface, this invention proposes a brightness temperature simulation method for snow and ice environments, which can solve the above problems.

[0007] To solve the above-mentioned technical problems, the present invention adopts the following technical solution: A method for simulating brightness temperature in snow and ice environments includes: Step 1: Construct a physical model of the ice and snow layers using the SMRT model, and set the electromagnetic theory model and ice and snow microstructure of the SMRT model. The SMRT model is used to output the uplink radiation in the C to K bands; Step 2: Calculate the surface emissivity ε corresponding to the upward radiation: ; in, For the SMRT model in downlink radiation The uplink radiation output at 100K. For the SMRT model in downlink radiation The uplink radiation output when it equals 0K. For the transmittance of multilayer media, The calculation method is as follows: ; in, The effective transmission coefficient of the multilayer medium. Where is the dielectric constant of air. The angle of incidence of the satellite. Here is the dielectric constant of sea ice. The angle of transmission of the ice layer. The calculation method is as follows: ; in, This represents the transmittance coefficient between the air interface and the snow interface. This represents the reflectance between the air interface and the snow interface. This represents the transmission coefficient between the snow interface and the sea ice interface. The reflection coefficient between the snow interface and the sea ice interface. This represents the total propagation attenuation factor for snow and ice layers; Step 3: Obtain the mixed pixels, correct the surface emissivity of the mixed pixels, and obtain the emissivity of the mixed pixels. Calculate the equivalent emission layer temperature for all pixels; Step 4: Obtain ERA5 atmospheric profile data and its near-surface parameters. Input the surface emissivity of non-mixed pixels, the emissivity of mixed pixels, and the equivalent emissivity layer temperature into the underlying surface parameters of the RTTOV model. The RTTOV model combines the ERA5 atmospheric profile data and near-surface parameters to generate the simulated brightness temperature of the Arctic ice and snow environment.

[0008] In some embodiments, the surface physical parameters mentioned in step one include any combination of ice / snow density, ice / snow thickness, ice / snow temperature, ice / snow particle size, ice / snow viscosity, and ice salinity. The ice / snow thickness is derived from satellite data, while the ice / snow density, ice / snow particle size, and ice / snow viscosity are constants.

[0009] In some embodiments, the surface emissivity ε mentioned in step two includes the surface emissivity under vertical polarization. and surface emissivity under horizontal polarization Calculate the surface emissivity under each polarization mode: ; ; in, The transmittance of the multilayer medium under vertical polarization. The transmittance of the multilayer medium under horizontal polarization; ; ; and These are the effective transmission coefficients of the multilayer medium under vertical polarization and horizontal polarization, respectively. ; ; , These represent the transmission coefficients between the air and snow interfaces and between the snow and sea ice interfaces, respectively, under vertical polarization. , These represent the reflection coefficients between the air-snow interface and the snow-ice interface under vertical polarization, respectively. , These represent the transmission coefficients between the air and snow interfaces and between the snow and sea ice interfaces, respectively, under horizontal polarization. , These represent the reflection coefficients between the air interface and the snow interface, and the reflection coefficients between the snow interface and the sea ice interface, respectively, under horizontal polarization.

[0010] In some embodiments, the surface emissivity ε mentioned in step two includes the surface emissivity under vertical polarization. and surface emissivity under horizontal polarization Calculate the surface emissivity under each polarization mode: ; ; in, The transmittance of the multilayer medium under vertical polarization. The transmittance of the multilayer medium under horizontal polarization; ; ; and These are the effective transmission coefficients of the multilayer medium under vertical polarization and horizontal polarization, respectively. ; ; , These represent the transmission coefficients between the air and snow interfaces and between the snow and sea ice interfaces, respectively, under vertical polarization. , These represent the reflection coefficients between the air-snow interface and the snow-ice interface under vertical polarization, respectively. , These represent the transmission coefficients between the air and snow interfaces and between the snow and sea ice interfaces, respectively, under horizontal polarization. , These represent the reflection coefficients between the air interface and the snow interface, and the reflection coefficients between the snow interface and the sea ice interface, respectively, under horizontal polarization.

[0011] In some embodiments, in step two The calculation method is as follows: ; , These are the propagation attenuation factors for snow and ice layers, respectively, and are calculated as follows: ; ; in, , These are the scattering coefficients of the snow layer and the ice layer, respectively. and These are the absorption coefficients for snow and ice layers, respectively. and These represent the thicknesses of the snow layer and the ice layer, respectively.

[0012] In some embodiments, and Calculated using Fresnel's law: ; .

[0013] In some embodiments, step three further includes obtaining the sea ice concentration SIC and determining whether it is a mixed pixel based on the sea ice concentration SIC. When the sea ice concentration SIC of a pixel is not 1, the pixel is a mixed pixel.

[0014] In some embodiments, the method for correcting the emissivity of the mixed pixels includes: ; in, Brightness temperature of pure sea ice snow cover simulated by the SMRT model. Sea surface temperature, Let T be the emissivity of seawater, and T be the emissivity of pure ice and snow.

[0015] In some embodiments, the equivalent emitter layer temperature = .

[0016] In some embodiments, T is calculated as follows: .

[0017] Compared with existing technologies, the advantages and positive effects of this invention are as follows: The brightness temperature simulation method for ice and snow environment observation of this invention first uses the SMRT model to construct a physical model of ice and snow layers. This model can simulate the transmission process of microwave radiation in multiple layers of snow, sea ice, and other media. By using different electromagnetic theories and the microstructure representation of snow, it simulates the microwave scattering behavior in the snow layer, improving the accuracy of characterizing the microscopic processes of ice and snow and simulating the surface. This allows for a better understanding of the physical characteristics of snow and its impact on microwave signals. The model outputs uplink radiation in the high-frequency band (C to K band) as a basic radiation model. To address the problems of the SMRT model, which is mainly based on terrestrial snow cover and does not fully consider the influence of atmospheric effects on simulated brightness temperature in the higher frequency band of 6.9–19 GHz, thus lacking accuracy in simulating brightness temperature in Arctic ice and snow environments, this invention calculates the surface emissivity of the uplink radiation output by the SMRT model corresponding to vertical polarization. The surface emissivity of mixed pixels containing seawater and ice / snow is then corrected, ensuring that the corrected emissivity reflects the influence of seawater on brightness temperature simulation and emissivity calculation. To address the issue that the data simulated by the SMRT model cannot reflect atmospheric interference, the effective coupling of the SMRT and RTTOV models was used to obtain high-frequency data on brightness temperature of the ice and snow environment, which can accurately characterize the micro-processes of ice and snow and the effects on the Earth's surface. This provides more accurate model support for the remote sensing inversion of Arctic ice and snow environment parameters.

[0018] Other features and advantages of the present invention will become clearer after reading the detailed description of the embodiments of the present invention in conjunction with the accompanying drawings. Attached Figure Description

[0019] Figure 1 This is a flowchart of an embodiment of the brightness temperature simulation method for ice and snow environment observation proposed in this invention; Figure 2 This is a schematic diagram of a multi-layered medium consisting of "air-snow-sea ice" in one embodiment of the brightness temperature simulation method for ice and snow environment observation proposed in this invention. Figure 3 This is a simulated SMRT map of the OIB region from 2013 to 2019, from one embodiment of the brightness temperature simulation method for ice and snow environment observation proposed in this invention. Figure 3 middle: Figure 3 (a) is a simulation diagram of SMRT in the L-band without considering snow insulation; Figure 3 (b) is a simulation diagram of SMRT considering snow insulation in the L-band; Figure 3 (c) is a simulation diagram of SMRT in the C-band without considering snow insulation; Figure 3 (d) is a SMRT simulation diagram for C-band considering snow insulation; Figure 3 (e) is a simulation diagram of SMRT in the X-band without considering snow insulation; Figure 3 (f) is a SMRT simulation diagram considering snow insulation in the X-band; Figure 3 (g) is a simulation diagram of SMRT in the K-band without considering snow insulation; Figure 3 (h) is a SMRT simulation diagram of the K-band considering snow insulation; Figure 4 This is an example of the brightness temperature simulation method for ice and snow environment observation proposed in this invention. A comparative chart analyzing the influence of the interglacial channel ratio on the simulated brightness temperature error was presented using Willmes interglacial channel products. Figure 4 middle: Figure 4 (a) is a scatter plot of the brightness temperature error of C-band seasonal ice simulation versus the proportion of inter-ice channels; Figure 4 (b) is a scatter plot of the brightness temperature error of multi-year ice simulation in C-band versus the proportion of interglacial channels; Figure 4 (c) is a scatter plot of the brightness temperature error of X-band seasonal ice simulation versus the proportion of inter-ice channels; Figure 4 (d) is a scatter plot of the brightness temperature error of multi-year ice simulation in the X-band versus the proportion of interglacial channels; Figure 4 (e) is a scatter plot of the brightness temperature error of K-band seasonal ice simulation versus the proportion of inter-ice channels; Figure 4 (f) is a scatter plot of the brightness temperature error of K-band multi-year ice simulation versus the proportion of interglacial channels; Figure 5 This is a comparison image of the SMRT-simulated brightness temperature before and after correction by inter-ice waterway in one embodiment of the brightness temperature simulation method for ice and snow environment proposed in this invention. Figure 5 middle: Figure 5 (a) is a comparison of the C-band seasonal ice SMRT simulated brightness temperature before and after correction by inter-ice waterway; Figure 5 (b) is a comparison of the C-band multi-year ice SMRT simulated brightness temperature before and after correction by interglacial channels; Figure 5 (c) is a comparison of the brightness temperature of X-band seasonal ice SMRT simulation before and after correction by inter-ice waterway; Figure 5 (d) is a comparison of the brightness temperature of X-band multi-year ice SMRT simulation before and after correction by interglacial waterway; Figure 5 (e) is a comparison of the brightness temperature of X-band seasonal ice SMRT simulation before and after correction by inter-ice waterway; Figure 5 (f) is a comparison of the brightness temperature of multi-year ice SMRT simulation before and after correction by interglacial channels in the X-band; Figure 6 This is a comparison chart of the SMRT+RTTOV brightness temperature simulation results in the OIB region with satellite brightness temperature in one embodiment of the brightness temperature simulation method for ice and snow environment observation proposed in this invention. Figure 6 middle: Figure 6 (a) is a comparison between the simulated brightness temperature of SMRT+RTTOV in the C-band OIB region and the satellite brightness temperature; Figure 6 (b) is a comparison of the SMRT+RTTOV brightness temperature simulation results and the satellite brightness temperature in the X-band OIB region; Figure 6 (c) is a comparison of the simulated brightness temperature of SMRT+RTTOV in the K-band OIB region with the satellite brightness temperature; Figure 7 This is an embodiment of the brightness temperature simulation method for ice and snow environment observation proposed in this invention, showing a comparison between the brightness temperature simulated by the coupling model of this method and the satellite brightness temperature. Figure 7 middle: Figure 7 (a) is a comparison of the brightness temperature simulated by the coupling model using this method in the C-band with the satellite brightness temperature; Figure 7 (b) is a comparison of the brightness temperature simulated by the coupling model using this method in the X-band with the satellite brightness temperature; Figure 7 (c) is a comparison of the brightness temperature simulated by the coupling model using this method in the K-band with the satellite brightness temperature. Detailed Implementation

[0020] The specific embodiments of the present invention will be described in further detail below with reference to the accompanying drawings.

[0021] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, 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, 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.

[0022] Example 1, see Figure 1 As shown in the figure, this embodiment proposes a brightness temperature simulation method for ice and snow environment observation, including: Step 1: Use the SMRT model to construct a physical model of the ice and snow layers, and set the electromagnetic theory model and ice and snow microstructure of the SMRT model. The SMRT model is used to output the uplink radiation in the C to K bands.

[0023] The electromagnetic theory model is the core of the SMRT model, responsible for calculating the electromagnetic properties of microwaves, such as scattering and absorption, within the snow and ice layer. The SMRT model integrates multiple theories; this embodiment selects the Dense Media Radiative Transfer (DMRT) theory, which is specifically designed for high-density media (such as wet snow and deep snow), considering the "coherent scattering" effect between particles, and can more accurately describe the scattering process in dense snow layers. The snow and ice microstructure model defines the spatial arrangement of ice crystals at the millimeter to micrometer scale, and is key to determining how electromagnetic waves are scattered and absorbed. The SMRT model supports various microstructure models; this embodiment selects the Sticky Hard Spheres (SHS) model, which uses "viscosity" to describe the aggregation and adhesion state between ice crystals, more closely resembling real snow.

[0024] Upward radiation is equivalent to brightness temperature data. The basic settings of the SMRT model directly affect the final brightness temperature simulation results, with temperature and salinity settings having a significant impact on the accuracy of the simulation. Thanks to the high albedo and low thermal conductivity of snow, the snow layer has excellent insulating properties. Even a thin layer of snow on seasonal ice can effectively block heat exchange between the sea ice surface and the atmosphere, thus significantly affecting the ice temperature. Furthermore, the brine in the medium strongly absorbs microwave radiation, significantly altering its radiation characteristics. The effective dielectric constant of sea ice is often dominated by the volume of its internal brine; therefore, the vertical distribution of sea ice salinity directly affects its complex dielectric constant, indirectly leading to changes in brightness temperature. Moreover, simplifying the vertical distribution of sea ice salinity (setting it to a constant value) when simulating sea ice microwave brightness temperature is a major cause of uncertainty in the model simulation results.

[0025] like Figure 3 The image shows the SMRT simulation results for the OIB (Operation IceBridge) region from 2013 to 2019. The first column corresponds to the average polarization brightness temperature of the L-band (1.4 GHz), and the vertical polarization of the C-band (6.9 GHz), X-band (10 GHz), and K-band (19 GHz), with the SMRT model considering the snow insulation effect. The second column shows the results of the same band SMRT model without considering the snow insulation effect. Blue represents seasonal ice, and red represents multi-year ice. Figure 3 It can be seen that the scatter plots of each band are more concentrated after adding the snow insulation effect, and the correlation between each band is significantly improved. In addition, the L band is less affected by the atmosphere, so no further processing is performed on the L band when coupling the RTTOV model.

[0026] Therefore, to fully consider the impact of snow's insulating effect on ice temperature and the vertical distribution of sea ice salinity, this embodiment uses the SMRT model to construct a physical model of the ice and snow layers. This includes setting the surface physical parameters of the SMRT model, including any combination of parameters such as ice / snow density, ice / snow thickness, ice / snow temperature, ice / snow particle size, viscosity, and ice salinity. To fully consider the insulating effect of snow and the vertical distribution of sea ice salinity, this method extends the simulation bands to the C to K bands, improving simulation accuracy.

[0027] Meanwhile, when using the SMRT model to simulate brightness temperature, parameters that have little impact on the simulated brightness temperature, such as the radius, viscosity, and density of ice and snow, are set as constants. For specific parameters, refer to the FYI, MYI, and snow parameter settings of Soriot et al. For other data, snow thickness is based on snow thickness data from the University of Bremen, ice thickness is based on SMOS-CryoSat ice thickness data, and surface temperature is based on L4 reprocessed data from CMS.

[0028] The key to this solution is how to effectively couple the SMRT and RTTOV models.

[0029] Step two, model coupling, involves effectively coupling the SMRT and RTTOV models to address two core issues: first, accurately calculating the surface emissivity and corresponding emissivity layer temperature; and second, correcting the brightness temperature simulation error of mixed pixels of "sea ice / snow cover + seawater". Specifically, this includes: Calculate the surface emissivity ε corresponding to the upward radiation: .

[0030] in, For the SMRT model in downlink radiation The uplink radiation output at 100K. For the SMRT model in downlink radiation The uplink radiation output when it equals 0K. For the transmittance of multilayer media, The calculation method is as follows: .

[0031] Considering the penetrating power of low-frequency radiation—that is, the lower the frequency, the greater the depth it can penetrate through ice and snow—assuming only surface emission would lead to low calculation accuracy. This embodiment's emissivity calculation method introduces transmittance. It is used to calculate the surface emissivity corresponding to upward radiation, and includes the radiation influence at the bottom of the ice and snow layer in the calculation to improve accuracy.

[0032] Furthermore, the method for calculating the surface emissivity corresponding to the upward radiation is based on the SMRT model, by changing the downward radiation... (Setting constant values ​​of 100K and 0K), observe the upward radiation simulated by the SMRT model. The changes in these values ​​were used to deduce the surface emissivity simulated by the SMRT model. .

[0033] in, The effective transmission coefficient of the multilayer medium. The dielectric constant of air, which is 1 by default. The angle of incidence of the satellite. Here is the dielectric constant of sea ice, and the dielectric constant of snow. dielectric constant of sea ice All can be calculated using the internal functions of the SMRT model. The angle of transmission of the ice layer. The calculation method is as follows: .

[0034] in, This represents the transmittance coefficient between the air interface and the snow interface. This represents the reflectance between the air interface and the snow interface. This represents the transmission coefficient between the snow interface and the sea ice interface. The reflection coefficient between the snow interface and the sea ice interface. This represents the total propagation attenuation factor for snow and ice layers.

[0035] In some embodiments, the dielectric constant of the snow accumulation is... dielectric constant of sea ice They can be calculated separately using the internal functions of the SMRT model. It is calculated using Fresnel's law.

[0036] For multi-layered media consisting of "air-snow-sea ice" such as Figure 2 As shown, a propagation attenuation factor needs to be introduced. Where ks and ka are the scattering and absorption coefficients of each layer, respectively, which can be calculated by the SMRT model, and d is the thickness of each layer. Let be the transmission angle, and the total propagation factor be . .

[0037] exist Figure 2 middle, The angle of incidence of the satellite is the known data. The angle of transmission of the snow layer. , , d represents the dielectric constants of air, snow, and sea ice, respectively, which are also known data. snow d represents the snow layer thickness. ice This represents the thickness of the ice layer.

[0038] From the above, we can calculate that: , .

[0039] Satellite observations commonly contain mixed pixels of "sea ice and snow + seawater," while the SMRT model assumes a pure sea ice surface, failing to reflect the impact of seawater on emissivity. This method constructs a coupled calculation framework between the emissivity of mixed pixels and the temperature of the emissive layer to achieve quantitative correction for the influence of seawater.

[0040] Step 3: Obtain the mixed pixels, correct the surface emissivity of the mixed pixels, and obtain the emissivity of the mixed pixels. Calculate the equivalent emission layer temperature for all pixels.

[0041] In this embodiment, the Willmes interglacial channel product was used to analyze the impact of the interglacial channel ratio on simulation error in areas with sea ice concentration greater than 99%. The specific scatter plot relationship is as follows: Figure 4 As shown, for areas with a density >99%, the scatter plot of brightness temperature error and interglacial channel ratio in the SMRT model simulation is shown. It can be seen that, except for the 19GHz multi-year ice, the simulation error and the interglacial channel ratio have a certain correlation, further proving the influence of the existence of interglacial channels on brightness temperature. Brightness temperature correction is urgently needed for this part of the region.

[0042] like Figure 5 As shown, this is a comparison of SMRT simulated brightness temperature before and after correction for interglacial channels in areas with a density >99%. The first column represents seasonal ice (blue, uncorrected; green, corrected), and the second column represents multi-year ice (red, uncorrected; orange, corrected). This scheme constructs effective linear correction equations for each band in step three to further correct the simulated brightness temperature, ultimately obtaining the equivalent emittance temperature including mixed pixels from interglacial channels. Figure 5 The comparison results before and after correction show that the simulated brightness temperature of the band after interglacial channel correction is more consistent with the satellite observation brightness temperature, effectively addressing the problem of the SMRT model simulating an excessively high brightness temperature in such interglacial channel regions.

[0043] Step 4: Obtain ERA5 atmospheric profile data and its near-surface parameters. Input the surface emissivity of non-mixed pixels, the emissivity of mixed pixels, and the equivalent emissivity layer temperature into the underlying surface parameters of the RTTOV model. The RTTOV model combines the ERA5 atmospheric profile data and near-surface parameters to generate the simulated brightness temperature of the Arctic ice and snow environment.

[0044] The simulation results using OIB measured data in this embodiment are as follows: Figure 6 As shown, Figure 6 The results are from the coupled model simulation of the OIB region from 2013 to 2019. Figure 6 (a) ~ Figure 6 (c) shows the simulated brightness temperature results for each band. Blue represents seasonal ice, and red represents multi-year ice. It can be seen that by introducing the influence of the atmosphere on microwave radiation in each band, the accuracy of the simulated brightness temperature is significantly improved, and the root mean square error is significantly reduced compared to SMRT simulation alone. Applying the coupled model to the entire Arctic region also achieved effective satellite-observed brightness temperature simulation. The scatter plot of simulated brightness temperature versus observed brightness temperature is shown below. Figure 7 The image shows a comparison between the brightness temperature simulated by the SMRT+RTTOV coupled model on Arctic ice on February 25, 2015, and the satellite brightness temperature. Figure 7 (a) ~ Figure 7 (c) shows the simulated brightness temperature results for each band. Blue represents seasonal ice and red represents multi-year ice. It can be seen that the simulation results are consistent with the measured data using OIB. The brightness temperature simulated by the coupled model has a high correlation with the brightness temperature observed by the satellite and a low root mean square error.

[0045] In some embodiments, the surface physical parameters in step one include any combination of ice / snow density, ice / snow thickness, ice / snow temperature, ice / snow particle size, viscosity, and ice salinity.

[0046] In some embodiments, the surface emissivity ε mentioned in step two includes the surface emissivity corresponding to vertical polarization. Surface emissivity corresponding to horizontal polarization Calculate the surface emissivity under each polarization mode: ; ; in, The transmittance of the multilayer medium under vertical polarization. The transmittance of the multilayer medium under horizontal polarization.

[0047] .

[0048] .

[0049] and These are the effective transmission coefficients of the multilayer media under vertical and horizontal polarization, respectively.

[0050] .

[0051] .

[0052] , These represent the transmission coefficients between the air interface and the snow interface, and the snow interface and the sea ice interface, respectively, corresponding to vertical polarization. , These represent the reflection coefficients between the air-snow interface and the snow-ice interface, respectively, corresponding to vertical polarization. , These represent the transmission coefficients between the air-snow interface and the snow-ice interface, respectively, corresponding to horizontal polarization. , These represent the reflection coefficients between the air interface and the snow interface, and between the snow interface and the sea ice interface, respectively, corresponding to horizontal polarization.

[0053] In some embodiments, for a multi-layered medium consisting of "air-snow-sea ice", a propagation attenuation factor is introduced to simulate the attenuation of electromagnetic waves propagating in different media.

[0054] In step two The calculation method is as follows: .

[0055] , These are the propagation attenuation factors for snow and ice layers, respectively, and are calculated as follows: .

[0056] .

[0057] in, , These are the scattering coefficients of the snow layer and the ice layer, respectively. and These are the absorption coefficients of the snow layer and the ice layer, respectively. and The figures represent the thicknesses of the snow and ice layers, respectively; these are known data.

[0058] In some embodiments, and Calculated using Fresnel's law: ; .

[0059] In some embodiments, step three further includes obtaining the sea ice concentration SIC and determining whether it is a mixed pixel based on the sea ice concentration SIC. When the sea ice concentration SIC of a pixel is not 1, the pixel is a mixed pixel; otherwise, it is a pure sea ice pixel.

[0060] Since the SMRT model simulates the brightness temperature of pure sea ice and snow cover, it assumes the underlying surface is pure sea ice. However, actual satellite observations commonly contain mixed pixels of "sea ice and snow cover + seawater," meaning that ε represents the calculated emissivity of pure ice and snow cover, which fails to reflect the influence of seawater on emissivity. This method constructs a coupled calculation framework between the emissivity of mixed pixels and the emissive layer temperature to achieve quantitative correction for the influence of seawater.

[0061] In some embodiments, the method for correcting the emissivity of mixed pixels includes: .

[0062] in, Brightness temperature of pure sea ice snow cover simulated by the SMRT model. This represents the sea surface temperature, set to 271.35K by default. The emissivity of seawater is calculated using the RTTOV model in this embodiment, specifically employing the era 5 atmospheric profile and near-surface parameter data to simulate the emissivity of pure seawater. T represents the emissivity temperature of pure ice and snow. The temperature of the emitter layer is that of pure ice and snow. The equivalent emission layer temperature of the mixed pixel; The value is the brightness temperature of pure sea ice snow cover simulated by the SMRT model; SIC is the sea ice concentration (value range 0-1, using the Arctic sea ice concentration product published by OSI-SAF).

[0063] Due to limitations in spatial resolution and inversion algorithms, sea ice concentration products can easily misclassify some interglacial channels as pure ice areas, and some interglacial channels exhibit refreezing. Therefore, emissivity correction methods are not suitable for calculating equivalent emitter layer temperatures. This method analyzes the impact of interglacial channels on simulation errors in areas with sea ice concentration greater than 99% using Willmes interglacial channel products. Further SMRT simulation brightness temperature correction is performed on bands where simulation errors are highly correlated with interglacial channels, ultimately obtaining the equivalent emitter layer temperature including mixed pixels containing interglacial channels.

[0064] The presence of interglacial channels has a certain impact on the brightness temperature error of the L-band simulation in the SMRT model. However, due to limitations in spatial resolution and inversion algorithms, sea ice concentration products are prone to misclassifying some interglacial channels as pure ice areas. Furthermore, some interglacial channels undergo refreezing, which alters their radiation characteristics, making emissivity and brightness temperature unavailable. Therefore, existing emissivity correction methods cannot further correct the emissivity of mixed pixels and are no longer applicable to interglacial channels. To address this, this study analyzes the impact of the interglacial channel ratio on simulation errors in areas with sea ice concentration greater than 99%. The specific scatter plot relationships are shown below. Figure 4As shown, there is a certain correlation between simulation error and the interglacial channel ratio. Based on the relationship between simulation error and the interglacial channel ratio, this invention constructs effective linear correction equations for each band to further correct the SMRT simulated brightness temperature (e.g., Figure 5 As shown, the 19GHz error of multi-year ice has a low correlation with the proportion of interglacial channels and was not corrected. The equivalent emitter temperature, including the mixed pixels containing interglacial channels, was ultimately obtained. Figure 5 The comparison results before and after correction show that the simulated brightness temperature and the satellite-observed brightness temperature are more consistent in the band after correction by the interglacial channel.

[0065] In some embodiments, the equivalent emitter layer temperature = By incorporating sea ice concentration data, the influence of seawater on emitter temperature in mixed pixels with SIC values ​​not equal to 1 was introduced, providing more accurate surface temperature parameters for subsequent coupling with the RTTOV model.

[0066] SIC stands for Sea Ice Concentration, with a value range of 0-1. In this embodiment, the Arctic Sea Ice Concentration product published by OSI-SAF is used.

[0067] In some embodiments, T is calculated as follows: .

[0068] In some embodiments, the calculation method for each transmission coefficient and reflection coefficient in step two is as follows: .

[0069] .

[0070] .

[0071] .

[0072] in, The angle of transmission of the snow layer. , , are the dielectric constants of air, snow, and sea ice, respectively.

[0073] .

[0074] .

[0075] .

[0076] .

[0077] according to , , , Calculate the effective transmission coefficient of the multilayer medium corresponding to horizontal polarization. ,according to Calculate the transmittance of multilayer media under horizontal polarization ,according to Calculate the surface emissivity corresponding to the upward radiation in horizontal polarization. .

[0078] Step four also includes calculating the surface emissivity under the vertical polarization of the non-mixed pixels. and surface emissivity under horizontal polarization and pixel emissivity after hybrid pixel correction The underlying surface parameters are input into the RTTOV model, and the final brightness temperature simulation is performed by combining the ERA5 atmospheric profile and near-surface parameters.

[0079] This embodiment, based on the SMRT and RTTOV radiative transfer models, studies the brightness temperature simulation algorithm for the Arctic ice and snow environment. This provides a theoretical basis for establishing the relationship between satellite-observed brightness temperature and Arctic ice and snow parameters, and offers more accurate model support for remote sensing inversion of Arctic ice and snow environment parameters. While the SMRT model can effectively simulate the brightness temperature of the Arctic ice and snow environment, it often ignores atmospheric interference. Although RTTOV can consider atmospheric influences, it lacks characterization of microscopic ice and snow processes. By coupling the SMRT model with RTTOV, effective simulation of satellite-observed brightness temperature in the C to Ka bands of the Arctic ice and snow environment can be achieved. Simultaneously, large-scale, long-term series of Arctic ice and snow environment brightness temperature data can be accurately obtained, providing more precise model support for remote sensing inversion of Arctic ice and snow environment parameters.

[0080] Of course, the above description is not intended to limit the present invention, and the present invention is not limited to the examples given above. Any changes, modifications, additions or substitutions made by those skilled in the art within the scope of the present invention should also fall within the protection scope of the present invention.

Claims

1. A method for simulating brightness temperature in ice and snow environments, characterized in that, include: Step 1: Construct a physical model of ice and snow using the SMRT model, and set the electromagnetic theory model and ice and snow microstructure of the SMRT model. The SMRT model is used to output the uplink radiation in the C to K bands. Constructing a physical model of ice and snow using the SMRT model includes setting the surface physical parameters of the SMRT model. Step 2: Calculate the surface emissivity ε corresponding to the upward radiation: ； in, For the SMRT model in downlink radiation The uplink radiation output at 100K. For the SMRT model in downlink radiation The uplink radiation output when equal to 0K. For the transmittance of multilayer media, The calculation method is as follows: ； in, The effective transmission coefficient of the multilayer medium. The dielectric constant of air is . The angle of incidence of the satellite. Here is the dielectric constant of sea ice. The angle of transmission of the ice layer. The calculation method is as follows: ； in, This represents the transmittance coefficient between the air interface and the snow interface. This represents the reflectance between the air interface and the snow interface. This represents the transmission coefficient between the snow interface and the sea ice interface. The reflection coefficient between the snow interface and the sea ice interface. This represents the total propagation attenuation factor for snow and ice layers; Step 3: Obtain the mixed pixels, correct the surface emissivity of the mixed pixels, and obtain the emissivity of the mixed pixels. Calculate the equivalent emission layer temperature for all pixels; Step 4: Obtain ERA5 atmospheric profile data and its near-surface parameters. Input the surface emissivity of non-mixed pixels, the emissivity of mixed pixels, and the equivalent emissive layer temperature into the underlying surface parameters of the RTTOV model. The RTTOV model combines the ERA5 atmospheric profile data and near-surface parameters to generate the simulated brightness temperature of the Arctic ice and snow environment. Step 3 also includes obtaining the sea ice concentration SIC and determining whether it is a mixed pixel based on the sea ice concentration SIC. When the sea ice concentration SIC of a pixel is not 1, the pixel is a mixed pixel. The method for correcting the emissivity of the mixed pixels includes: ； in, Brightness temperature of pure sea ice snow cover simulated by the SMRT model. Sea surface temperature, Let T be the emissivity of seawater, and T be the emissivity of pure ice and snow. Equivalent emitter temperature = .

2. The method for simulating brightness temperature in ice and snow environment observation according to claim 1, characterized in that, The surface physical parameters mentioned in step one include any combination of ice / snow density, ice / snow thickness, ice / snow temperature, ice / snow particle size, viscosity, and ice salinity.

3. The method for simulating brightness temperature in ice and snow environment observation according to claim 1, characterized in that, The surface emissivity ε mentioned in step two includes the surface emissivity under vertical polarization. and surface emissivity under horizontal polarization Calculate the surface emissivity under each polarization mode: ； ; in, The transmittance of the multilayer medium under vertical polarization. The transmittance of the multilayer medium under horizontal polarization; ； ; and These are the effective transmission coefficients of the multilayer medium under vertical polarization and horizontal polarization, respectively. ； ； , These represent the transmission coefficients between the air and snow interfaces and between the snow and sea ice interfaces, respectively, under vertical polarization. , These represent the reflection coefficients between the air-snow interface and the snow-ice interface under vertical polarization, respectively. , These represent the transmission coefficients between the air and snow interfaces and between the snow and sea ice interfaces, respectively, under horizontal polarization. , These represent the reflection coefficients between the air interface and the snow interface, and the reflection coefficients between the snow interface and the sea ice interface, respectively, under horizontal polarization.

4. The method for simulating brightness temperature in ice and snow environment observation according to claim 3, characterized in that, The calculation methods for each transmission coefficient and reflection coefficient in step two are as follows: ； ； ； ； in, The angle of transmission of the snow layer. , , , respectively, are the dielectric constants of air, snow, and sea ice; ; ； ; 。 5. The method for simulating brightness temperature in ice and snow environment observation according to claim 2, characterized in that, In step two The calculation method is as follows: ； , These are the propagation attenuation factors for snow and ice layers, respectively, and are calculated as follows: ; ; in, , These are the scattering coefficients of the snow layer and the ice layer, respectively. and These are the absorption coefficients of the snow layer and the ice layer, respectively. and These represent the thicknesses of the snow layer and the ice layer, respectively.

6. The method for simulating brightness temperature in ice and snow environment observation according to claim 5, characterized in that, and Calculated using Fresnel's law: ； 。 7. The method for simulating brightness temperature in ice and snow environment observation according to claim 1, characterized in that, The method for calculating T is as follows: 。

Images