A method and device for estimating hourly near-surface air temperature by satellite remote sensing

Abstract

The present invention discloses a method and device for estimating hourly near-surface air temperature by satellite remote sensing. The method includes: establishing a first equation between the near-surface air temperature and the surface temperature according to the surface energy balance equation; determining the algorithm parameters required for estimating the near-surface air temperature according to the first equation, and obtaining the algorithm parameters; establishing a spatio-temporal relationship model between the algorithm parameters and the near-surface air temperature based on a geospatial relationship model; and estimating the hourly near-surface air temperature by using the spatio-temporal relationship model according to the algorithm parameters and meteorological station data. The method in the present invention can accurately estimate the hourly near-surface air temperature by satellite remote sensing.

Description

Technical Field

[0001] The present invention relates to the field of satellite communication, and particularly to a method and device for estimating hourly near-surface air temperature by satellite remote sensing. Background Art

[0002] The near-surface air temperature refers to the atmospheric temperature observed near the surface (usually at a height of 2 m above the ground). As an important indicator describing the interaction between the surface and the atmospheric environment, the near-surface air temperature controls most of the biological and physical processes in the natural system and is an important driving parameter for various surface process models, such as surface evapotranspiration models, hydrological models, and soil-vegetation-moisture system dynamic models. Accurately obtaining the spatio-temporal distribution of the near-surface air temperature has very important scientific significance and practical value for research on agricultural disaster prevention and mitigation, land surface processes, and global change. Hourly near-surface air temperature is mostly obtained using the air temperature data measured by meteorological stations. Meteorological stations can provide relatively accurate and time-continuous near-surface air temperature data, but they cannot describe the spatial continuity of the near-surface air temperature. The discrete point recording method of station air temperature data restricts the development of regional-scale climate environment research to a certain extent. The spatial interpolation method based on meteorological station air temperature data can make up for this defect to a certain extent. However, due to the lack of observation stations in remote and complex areas, under complex topographic and landscape conditions, the spatial range that the air temperature of a single station can represent is very limited, resulting in the interpolation accuracy being difficult to meet the requirements of regional-scale research. With the development of remote sensing technology, satellite remote sensing can obtain large-scale spatially continuous observations, and there is great application potential in estimating near-surface air temperature using satellite remote sensing data.

[0003] Currently, satellite remote sensing research on near-surface air temperature mainly focuses on polar-orbiting satellites. Polar-orbiting satellites have relatively high spatial resolution but low temporal resolution. Usually, a polar-orbiting satellite observes the same location only twice a day, and this observation frequency far from meets the requirement of obtaining hourly near-surface air temperature. Affected by complex atmospheric radiation and small signal-to-noise ratio, there are some difficulties in estimating near-surface air temperature by satellite remote sensing. The current main research methods are as follows. One is the empirical statistical method, which solves by establishing the correlation between surface temperature, vegetation index, atmospheric temperature profile, etc. and near-surface air temperature. This method is affected by different underlying surfaces and atmospheric structures, as well as the number and representativeness of samples. This method simplifies the relationship between near-surface air temperature and other factors, does not consider spatial non-stationarity, has low accuracy, poor portability, and cannot be extended to the regional scale. The other is the machine learning method based on neural networks, which simulates neuron features through a large number of training samples. Its calculation is complex but the problem-solving process is not clear, and the mathematical relationship between different parameters cannot be obtained. Moreover, if the parameters or machine learning algorithms are not selected properly, phenomena such as long training time or overfitting are likely to occur.

[0004] Therefore, there is an urgent need to propose a remote sensing estimation method for hourly near-surface air temperature that is applicable to geostationary satellites, has clear physical principles and reliable accuracy, and has high availability of algorithm parameters and is convenient for calculation. Summary of the Invention

[0005] The present invention provides a method for remotely sensing and estimating the hourly near-surface air temperature by satellite to solve the problem of inaccurate estimation of the near-surface air temperature in the prior art.

[0006] The method for remotely sensing and estimating the hourly near-surface air temperature by satellite in the present invention includes:

[0007] Establish a first equation between the near-surface air temperature and the surface temperature according to the surface energy balance equation;

[0008] Determine the algorithm parameters required for estimating the near-surface air temperature according to the first equation, and obtain the algorithm parameters;

[0009] Establish a spatio-temporal relationship model between the algorithm parameters and the near-surface air temperature based on the geospatial relationship model;

[0010] Estimate the hourly near-surface air temperature by using the spatio-temporal relationship model according to the algorithm parameters and the meteorological station data.

[0011] Preferably, establishing a first equation between the near-surface air temperature and the surface temperature according to the surface energy balance equation includes:

[0012] Establish the relationship between the surface energy balance equation and the surface incoming radiation energy and the surface outgoing radiation energy according to the following formula:

[0013]

[0014] where the net radiation energy R n is the net solar radiation energy obtained by the surface, H is the sensible heat flux from the underlying surface to the atmosphere, G is the soil heat flux, LE is the latent heat flux from the underlying surface to the atmosphere, is the downward short-wave radiation, α is the surface albedo, ε a is the effective emissivity of the cloudless atmosphere, T a is the near-surface air temperature, ε s is the surface emissivity, T s is the surface temperature, and σ is the Stefan-Boltzmann constant;

[0015] Establish the first equation based on the parameterization formulas of the sensible heat flux, the latent heat flux, and the soil heat flux:

[0016]

[0017] where γ is the psychrometric constant, γ ais the aerodynamic impedance, ξ is the first coefficient related to the surface vegetation coverage and leaf area index, Δ is the slope of the saturated water vapor pressure with respect to temperature, ρ is the air density, C p is the specific heat capacity of air at constant pressure, e 0 (T a ) is the saturated water vapor pressure of the underlying surface at temperature T a and e a is the water vapor pressure of the air near the surface.

[0018] Preferably, the algorithm parameters include: surface temperature, annual composite normalized difference vegetation index, 0 - 5 c cm soil volumetric water content, and specific humidity near the surface.

[0019] Preferably, obtaining the algorithm parameters includes:

[0020] Obtaining hourly surface temperature data of geostationary satellites, and parsing the hourly surface temperature data to read the satellite observation time, longitude and latitude, cloud - covered pixels, and surface temperature;

[0021] Obtaining the monthly composite vegetation index of the previous year, parsing the vegetation index to read the satellite observation month, longitude and latitude, monthly vegetation index, and calculating the annual composite normalized difference vegetation index;

[0022] Obtaining soil moisture data of the Land Data Assimilation System CLDAS, and parsing the soil moisture data to read the time, longitude and latitude, 0 - 5 cm soil volumetric water content;

[0023] Obtaining the specific humidity data near the surface of the Land Data Assimilation System CLDAS, and parsing the specific humidity data near the surface to read the time, longitude and latitude, and specific humidity near the surface;

[0024] Resampling the annual composite normalized difference vegetation index, 0 - 5 cm soil volumetric water content, and specific humidity near the surface to keep the same spatial resolution as the surface temperature.

[0025] Preferably, establishing the spatio - temporal relationship model between the algorithm parameters and the near - surface air temperature based on the geospatial relationship model includes:

[0026] Performing local multiple linear regression to solve through the geographically weighted regression method to establish the spatio - temporal relationship model between the algorithm parameters and the near - surface air temperature.

[0027] Preferably, establishing the spatio - temporal relationship model between the algorithm parameters and the near - surface air temperature according to the following formula:

[0028]

[0029] where Y j is (u j , vj , t j ) The near-surface air temperature at the point, u is the longitude, v is the latitude, t is the time, X ij is the i-th algorithm parameter, ε j is the random error, β0 and βi are the first and second fitting coefficients respectively;

[0030] The fitting coefficients are solved by the following four formulas:

[0031] β(u j , v j , t j ) = (X T w(u j , v j , t j )X) -1 X T w(u j , v j , t j )Y

[0032]

[0033]

[0034]

[0035] Among them, β(u j , v j , t j ) is the fitting coefficient, w(u j , v j , t j ) is the weight matrix, d jk is the distance between the estimation point j and the surrounding sample point k, and b is the adaptive bandwidth.

[0036] Preferably, estimating the hourly near-surface air temperature according to the algorithm parameters and meteorological station data by using the spatio-temporal relationship model includes:

[0037] Obtaining meteorological station data; the meteorological station data includes: meteorological station number, longitude and latitude, altitude, observation time, and thermometer screen temperature;

[0038] Performing spatio-temporal matching on the surface temperature and the meteorological station data;

[0039] Identifying the surface temperature and the meteorological station data after spatio-temporal matching, and estimating the hourly near-surface air temperature by using the spatio-temporal relationship model.

[0040] Preferably, the method further includes:

[0041] Perform accuracy verification on the estimated hourly near-surface air temperature.

[0042] An apparatus for estimating hourly near-surface air temperature by satellite remote sensing in the present invention, the apparatus includes:

[0043] A first calculation unit, configured to establish a first equation between the near-surface air temperature and the surface temperature according to the surface energy balance equation;

[0044] A first confirmation unit, configured to determine algorithm parameters required for estimating the near-surface air temperature according to the first equation, and obtain the algorithm parameters;

[0045] A model establishment unit, configured to establish a spatio-temporal relationship model between the algorithm parameters and the near-surface air temperature based on a geospatial relationship model;

[0046] A second calculation unit, configured to estimate the hourly near-surface air temperature by using the spatio-temporal relationship model according to the algorithm parameters and meteorological station data.

[0047] A computer-readable storage medium in the present invention, the computer-readable storage medium stores one or more programs, and the one or more programs can be executed by one or more processors to implement the steps of any of the above hourly near-surface air temperature satellite remote sensing estimation methods.

[0048] The method for estimating hourly near-surface air temperature by satellite remote sensing in the present invention uses the method of estimating near-surface air temperature from the surface temperature of a geostationary satellite, makes full use of the observation advantages of the high spatio-temporal resolution of the geostationary satellite, combines other easily obtainable auxiliary algorithm parameters, and accurately estimates the hourly near-surface air temperature by satellite remote sensing. Description of the Drawings

[0049] Figure 1 is a flow chart of the method for estimating hourly near-surface air temperature by satellite remote sensing in an embodiment of the present invention;

[0050] Figure 2 is a structural diagram of the apparatus for estimating hourly near-surface air temperature by satellite remote sensing in an embodiment of the present invention;

[0051] Figure 3 is a comparison chart of the near-surface air temperature estimated based on the present invention and the measured data of a meteorological station. Detailed Description of the Invention

[0052] The present invention will be further described in detail below with reference to the drawings and embodiments. It can be understood that the specific embodiments described herein are only used to explain the present invention, rather than limiting the present invention. In addition, it should be noted that, for the sake of description, only parts related to the present invention are shown in the drawings, rather than all the structures.

[0053] An embodiment of the present invention provides a method for estimating hourly near-surface air temperature by satellite remote sensing, as follows Figure 1 shown, the method includes:

[0054] Step 100, establish a first equation between the near-surface air temperature and the surface temperature according to the surface energy balance equation. Specifically, the surface energy balance equation is a mathematical expression used to measure the relationship between surface energy and meteorological elements that directly or indirectly affect the energy budget balance. The surface energy balance equation is the fundamental expression of the surface long-wave radiation, solar radiation, and surface heat budget balance. In this step, through the surface energy balance equation, based on the law of conservation and conversion of energy, that is, it means that the energy received by the surface is converted into other forms of motion in different ways to keep the energy balanced, a first equation between the near-surface air temperature and the surface temperature can be established, providing a clear physical principle for estimating the near-surface air temperature by satellite remote sensing.

[0055] Step 200, determine the algorithm parameters required for estimating the near-surface air temperature according to the first equation, and obtain the algorithm parameters. Specifically, when the first equation is determined through Step 100, in order to obtain the solution of the first equation, it is necessary to determine the required algorithm parameters according to this first equation for the solution process. Specifically, the algorithm parameters in the embodiments of the present invention are physical variables related to the estimation of the near-surface air temperature. In an embodiment of the present invention, the algorithm parameters may include satellite data inversion parameters and the land surface data assimilation system CLDAS parameters. Among them, some algorithm parameters are obtained through hourly data of geostationary satellites.

[0056] Step 300, establish a spatio-temporal relationship model between the algorithm parameters and the near-surface air temperature based on the geospatial relationship model. Specifically, during the establishment of the spatio-temporal relationship model, since there is a non-stable relationship in the global space between the algorithm parameters and the near-surface air temperature, the geographically weighted regression method can be used to perform local multiple linear regression to solve, substitute the distance weight matrix, and use the least squares method to perform local area linear fitting.

[0057] Step 400, estimate the hourly near-surface air temperature according to the algorithm parameters and meteorological station data by using the spatio-temporal relationship model. Specifically, match the algorithm parameters and meteorological station data spatio-temporally, and use the spatio-temporal relationship model of the near-surface air temperature to model and calculate the hourly near-surface air temperature.

[0058] The method for estimating the hourly near-surface air temperature by satellite remote sensing according to the embodiments of the present invention uses the method of estimating the near-surface air temperature by the geostationary satellite surface temperature, makes full use of the observation advantages of the high spatio-temporal resolution of the geostationary satellite, combines other easily obtained auxiliary algorithm parameters, and estimates the hourly near-surface air temperature by satellite remote sensing to solve the problem of inaccurate estimation of the near-surface air temperature in the prior art.

[0059] The hourly near-surface air temperature satellite remote sensing estimation method described in the specific embodiments of the present invention. Preferably, the first equation established between the near-surface air temperature and the surface temperature according to the surface energy balance equation includes:

[0060] Establish the relationship between the surface energy balance equation and the incoming surface radiation energy and the outgoing surface radiation energy according to the following formula:

[0061]

[0062] Among them, the net radiation energy R n is the net solar radiation energy obtained by the surface, which is the heat source dominating the meteorological environment. In the absence of other heat exchange methods, it determines the warming or cooling of the surface. H is the sensible heat flux from the underlying surface to the atmosphere, G is the soil heat flux, and LE is the latent heat flux from the underlying surface to the atmosphere. is the downward shortwave radiation (W·m -2 ), α is the surface albedo, ε a is the effective emissivity of the cloudless atmosphere, T a is the near-surface air temperature (K), ε s is the surface emissivity, T s is the surface temperature (K), σ is the Stefan-Boltzmann constant (5.56×10 -8 W·m -2 K -4 ). Specifically, according to the law of conservation and conversion of energy, the energy received by the surface is converted into other forms of motion in different ways to keep the energy balanced. Ignoring the heat storage from the surface to the vegetation canopy and part of the energy used for vegetation photosynthesis, the net radiation energy obtained by the surface includes the turbulent flux and the soil heat flux. At the same time, the net radiation energy obtained by the surface can also be expressed as the difference between the incoming surface radiation energy and the outgoing surface radiation energy. Therefore, an equation is established based on the above heat balance. And the net radiation energy obtained by the surface is equal to the sum of the sensible heat flux from the underlying surface to the atmosphere, the soil heat flux, and the latent heat flux from the underlying surface to the atmosphere, that is, the latent heat flux and the sensible heat flux from the underlying surface to the atmosphere are equal to the turbulent flux. In the above formula, the near-surface air temperature usually uses a reference height of 2m.

[0063] Furthermore, establish the first equation based on the parameterization formulas of the sensible heat flux, latent heat flux, and soil heat flux:

[0064]

[0065] Among them, γ is the psychrometric constant, γ a is the aerodynamic impedance, ξ is the first coefficient related to the surface vegetation coverage and leaf area index, Δ is the slope of the saturated water vapor pressure with respect to temperature, ρ is the air density, C p is the specific heat at constant pressure of air, e0 (T a ) is the temperature T a The saturated water vapor pressure of the underlying surface is e a The water vapor pressure of the air near the surface. Specifically, the sensible heat flux, latent heat flux, and soil heat flux are parameterized respectively. Through the derivation of the parameterization formula, the first equation between the near-surface air temperature and the surface temperature can be established according to the surface energy balance equation. The parameterization formulas of the sensible heat flux, latent heat flux, and soil heat flux are prior arts and will not be elaborated in this embodiment.

[0066] Specifically, for the hourly near-surface air temperature satellite remote sensing estimation method described in the specific embodiment of the present invention, preferably, the algorithm parameters include: surface temperature, annual composite normalized difference vegetation index, 0-5 cm soil volumetric water content, and near-surface specific humidity. Specifically, in an embodiment of the present invention, the satellite data inversion parameters include surface temperature and annual composite normalized difference vegetation index, and the land data assimilation system CLDAS parameters include 0-5 cm soil volumetric water content and near-surface specific humidity. The above constitute the algorithm parameters required in a specific embodiment of the present invention. Other algorithm parameters may vary according to the requirements of different embodiments and are not limited to the present invention. Preferably, the near-surface specific humidity usually selects the specific humidity data with a reference height of 2 m.

[0067] For the hourly near-surface air temperature satellite remote sensing estimation method described in the specific embodiment of the present invention, preferably, obtaining the algorithm parameters includes:

[0068] Obtaining the hourly surface temperature data of the geostationary satellite, and parsing the hourly surface temperature data to read the satellite observation time, longitude and latitude, cloud-covered pixels, and surface temperature. Specifically, the surface temperature is also called the surface temperature of clear sky pixels. Because only clear sky pixels have valid surface temperature values, and for cloud-covered pixels, since the satellite optical sensor cannot penetrate clouds, surface temperature inversion is not performed, so cloud-covered pixels have no valid surface temperature values and generally use fill values. In the embodiment of the present invention, the cloud-covered pixels and the surface temperature of clear sky pixels are obtained respectively, in order to only perform spatio-temporal matching of the surface temperature of clear sky pixels with the meteorological station data when establishing the geospatial relationship model for estimating the near-surface air temperature, and when the model estimates the near-surface air temperature, it only outputs the near-surface air temperature within the clear sky range, without estimating the near-surface air temperature of cloud-covered pixels.

[0069] Obtaining the monthly composite vegetation index of the previous year, parsing the vegetation index to read the satellite observation month, longitude and latitude, monthly vegetation index, and calculating the annual composite normalized difference vegetation index. Specifically, generally, the normalized difference vegetation index only has dekadal, 16-day, and monthly composite products, and the annual composite normalized difference vegetation index is calculated by the maximum value composite method on the basis of the monthly composite product.

[0070] Obtain the soil moisture data of the Land Data Assimilation System CLDAS, and parse the soil moisture data to read the time, longitude and latitude, and the soil volume water content of 0 - 5 cm.

[0071] Obtain the near - surface specific humidity data of the Land Data Assimilation System CLDAS, and parse the near - surface specific humidity data to read the time, longitude and latitude, and the near - surface specific humidity.

[0072] Resample the annual - synthesized normalized difference vegetation index, the soil volume water content of 0 - 5 cm, and the near - surface specific humidity to keep the spatial resolution consistent with that of the surface temperature. In a specific embodiment, the bilinear interpolation method is used to resample the annual - synthesized normalized difference vegetation index, the soil volume water content of 0 - 5 cm, and the near - surface specific humidity respectively until the spatial resolutions of the annual - synthesized normalized difference vegetation index, the soil volume water content of 0 - 5 cm, and the near - surface specific humidity are consistent with the surface temperature data.

[0073] For the hourly near - surface air temperature satellite remote sensing estimation method described in the specific embodiment of the present invention, preferably, establishing the spatio - temporal relationship model between the algorithm parameters and the near - surface air temperature based on the geospatial relationship model includes:

[0074] Perform local multiple linear regression to solve through the geographically weighted regression method to establish the spatio - temporal relationship model between the algorithm parameters and the near - surface air temperature. The embodiment of the present invention is based on the surface energy balance equation as the physical basis, considering the globally spatially non - stationary relationship between the algorithm parameters and the near - surface air temperature. Therefore, local multiple linear regression is performed on the spatially relationship model established based on the geographically weighted regression method, obtaining high - precision hourly near - surface air temperature results while being convenient for calculation and implementation.

[0075] For the hourly near - surface air temperature satellite remote sensing estimation method described in the specific embodiment of the present invention, preferably, establish the spatio - temporal relationship model between the algorithm parameters and the near - surface air temperature according to the following formula:

[0076]

[0077] Among them, Y j is the near - surface air temperature at the point of (u j , v j , t j ), u is the longitude, v is the latitude, t is the time, X ij is the i - th algorithm parameter, ε j is the random error, β 0 , β i are the first fitting coefficient and the second fitting coefficient respectively. Specifically, the algorithm parameters include the surface temperature, the annual - synthesized normalized difference vegetation index, the soil volume water content of 0 - 5 cm, and the near - surface specific humidity described in the above - mentioned embodiment. β0， β i The linear fitting coefficient of the local area is obtained by regression calculation of a certain number of sample points around the pixel to be estimated, and the least square method is used to solve the linear fitting coefficient of the local area. Specifically, the fitting coefficient is solved by the following four formulas:

[0078] β(u j , v j , t j )=(X T w(u j , v j , t j )X) -1 X T w(u j , v j , t j )Y

[0079]

[0080]

[0081]

[0082] Among them, β(u j , v j , t j ) is the fitting coefficient, w(u j , v j , t j ) is a weight matrix, which uses the spatial distance between the sample point and the pixel point to be estimated as the weight to represent the influence of the sample point on the coefficient estimation, d jk To estimate the distance between point j and the surrounding sample points k, b is the adaptive bandwidth, and the optimal adaptive bandwidth is determined by the cross-validation method used for local regression analysis.

[0083] The satellite remote sensing estimation method for hourly near-surface temperature described in the specific embodiment of the present invention preferably estimates the hourly near-surface temperature using the spatiotemporal relationship model based on the algorithm parameters and the meteorological station data, including:

[0084] Acquire meteorological station data; the meteorological station data includes: meteorological station number, longitude and latitude, altitude, observation time, and louvered box temperature. Specifically, conduct comprehensive quality inspection on the louvered box data observed by the meteorological station, including climatological limit value inspection, climate extreme value inspection, internal consistency inspection, spatial consistency inspection, and temporal consistency inspection, remove data with abnormal quality, and retain data that meets the comprehensive quality inspection standards as meteorological station data, including: meteorological station number, longitude and latitude, altitude, observation time, and louvered box temperature.

[0085] Match the land surface temperature with the meteorological station data in both time and space. Specifically, the main process of matching the clear-sky pixel land surface temperature with the meteorological station data includes: selecting the same observation time for both in terms of time, and selecting the average of the clear-sky land surface temperature of the 3*3 pixels surrounding the pixel where the meteorological station data is located for matching in terms of space.

[0086] Identify the land surface temperature and the meteorological station data after spatio-temporal matching, and use the spatio-temporal relationship model to estimate the hourly near-surface air temperature. Specifically, randomly select 70% of the land surface temperature and the meteorological station data after spatio-temporal matching as sample data and identify them, and use the spatio-temporal relationship model for estimating the near-surface air temperature to model and calculate the hourly near-surface air temperature.

[0087] In the specific embodiment of the present invention, for the method for estimating the hourly near-surface air temperature by satellite remote sensing, preferably, the method further includes:

[0088] Verify the accuracy of the estimated hourly near-surface air temperature. In a specific embodiment, first, match the estimation result of the hourly near-surface air temperature with the meteorological station data in both time and space. Select the same observation time for both in terms of time, and select the average of the estimation results of the near-surface air temperature of the 3*3 pixels surrounding the pixel where the meteorological station data is located for matching in terms of space. Based on the previously identified land surface temperature and meteorological station data, select 30% of the data that was not used as sample data to verify the accuracy of the estimation result of the near-surface air temperature.

[0089] In a preferred embodiment, the accuracy verification index is represented by the root mean square error RMSE:

[0090]

[0091] Where, represents the satellite remote sensing estimation result of the hourly near-surface air temperature of the i-th pixel, that is, the hourly near-surface air temperature obtained through the embodiment of the present invention. represents the air temperature in the Stevenson screen of the i-th pixel.

[0092] The embodiment of the present invention also provides a device for estimating the hourly near-surface air temperature by satellite remote sensing, as Figure 2 shown, the device includes:

[0093] The first calculation unit 201 is used to establish the first equation between the near-surface air temperature and the land surface temperature according to the surface energy balance equation;

[0094] The first confirmation unit 202 is used to determine the algorithm parameters required for estimating the near-surface air temperature according to the first equation, and obtain the algorithm parameters;

[0095] A model establishment unit 203, configured to establish a spatio-temporal relationship model between the algorithm parameters and the near-surface air temperature based on a geospatial relationship model;

[0096] A second calculation unit 204, configured to estimate the hourly near-surface air temperature by using the spatio-temporal relationship model according to the algorithm parameters and meteorological station data.

[0097] In the hourly near-surface air temperature satellite remote sensing estimation method in the specific embodiment of the present invention, an equation with clear physical meaning between the near-surface air temperature and the surface temperature is established according to the surface energy balance equation. Satellite data inversion parameters (including surface temperature, annual composite normalized difference vegetation index) and land data assimilation system CLDAS parameters (including 0-5 cm soil volumetric water content, near-surface specific humidity) are selected as algorithm parameters for near-surface air temperature estimation through the equation. Hourly measured data of the air temperature in the Stevenson screen of more than 2,400 national meteorological stations in the Chinese region during the high-temperature period in summer from July 20th to August 20th, 2021 are collected. Synchronously, FY-4A surface temperature product data, FY-3D normalized difference vegetation index, 0-5 cm soil volumetric water content of CLDAS, and near-surface specific humidity data with a reference height of 2 m are collected, and each parameter is preprocessed. A spatio-temporal relationship model for near-surface air temperature estimation is established by using the geographically weighted regression method. On the basis of comprehensively quality-checking the Stevenson screen data observed at the meteorological stations, the surface temperature data and the station data are spatio-temporally matched, and 70% of the matched data is selected as sample data, and the hourly near-surface air temperature is modeled and calculated by using the near-surface air temperature estimation model. The measured air temperature data of the meteorological stations is used to verify the accuracy of the near-surface air temperature result calculated by the model, and the root mean square error is calculated as the accuracy verification index. Taking the air temperature estimation at 14:00 on August 1st, 2021 as an example, the results show that the satellite remote sensing estimation accuracy of the hourly near-surface air temperature provided by the present invention is 1.76 °C, as Figure 3 shown.

[0098] The embodiment of the present invention further provides a computer-readable storage medium, where the computer-readable storage medium stores one or more programs, and the one or more programs can be executed by one or more processors to implement the steps of the hourly near-surface air temperature satellite remote sensing estimation method in any one of the above specific embodiments.

[0099] An hourly near-surface air temperature estimation method provided by an embodiment of the present invention, which is applicable to geostationary satellites, has clear physical principles and reliable accuracy, high availability of algorithm parameters and is easy to calculate. The method provides a method for estimating near-surface air temperature using the surface temperature of geostationary satellites, makes full use of the observation advantages of geostationary satellites with high spatio-temporal resolution, combines other easily available auxiliary parameters, and estimates the hourly near-surface air temperature by satellite remote sensing. At the same time, based on the surface energy balance equation as the physical basis, considering the globally spatially non-stable relationship between algorithm parameters and near-surface air temperature, a local multiple linear regression solution is carried out based on the spatial relationship model of geographically weighted regression, and high-precision hourly near-surface air temperature results are obtained while facilitating calculation and implementation.

[0100] It should be understood that in various embodiments of this article, the magnitudes of the serial numbers of the above processes do not mean the order of execution. The order of execution of each process should be determined by its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of this article.

[0101] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system, or a computer program product. Therefore, the present application can adopt the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0102] The present application is described with reference to the flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or block in the flowchart and / or block diagram can be implemented by computer program instructions, and the combination of processes and / or blocks in the flowchart and / or block diagram can also be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate a device for implementing the functions specified in Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.

[0103] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory generate a manufactured product including an instruction device, and the instruction device implements the functions specified in Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.

[0104] These computer program instructions can also be loaded onto a computer or other programmable data processing apparatus, causing a series of operational steps to be performed on the computer or other programmable apparatus to generate a computer-implemented process, so that the instructions executed on the computer or other programmable apparatus provide steps for implementing the functions specified in one process or a plurality of processes and / or boxes Figure 1 one process or a plurality of processes and / or boxes Figure 1 or steps for implementing the functions specified in one box or a plurality of boxes.

[0105] The foregoing description of the specific exemplary embodiments of the present invention has been presented for purposes of illustration and description. It is not intended to limit the invention to the precise forms disclosed, and obviously, many modifications and variations are possible in light of the above teaching. The purpose of selecting and describing the exemplary embodiments is to explain the specific principles of the invention and its practical application so as to enable those skilled in the art to implement and utilize the various different exemplary embodiments of the present invention as well as various different selections and modifications. The scope of the present invention is intended to be defined by the claims and their equivalents.

Claims

1. A method for estimating hourly near-surface air temperature by satellite remote sensing, characterized in that, the method comprises: establishing a first equation between the near-surface air temperature and the surface temperature according to the surface energy balance equation; determining the algorithm parameters required for estimating the near-surface air temperature according to the first equation, and obtaining the algorithm parameters; establishing a spatio-temporal relationship model between the algorithm parameters and the near-surface air temperature based on a geospatial relationship model; estimating the hourly near-surface air temperature by using the spatio-temporal relationship model according to the algorithm parameters and meteorological station data.

2. The method for estimating hourly near-surface air temperature by satellite remote sensing according to claim 1, characterized in that, establishing a first equation between the near-surface air temperature and the surface temperature according to the surface energy balance equation includes: establishing the relationship between the surface energy balance equation and the incoming surface radiation energy and the outgoing surface radiation energy according to the following formula: Among them, the net radiation energy R n is the net solar radiation energy obtained by the earth's surface, H is the sensible heat flux from the underlying surface to the atmosphere, G is the soil heat flux, and LE is the latent heat flux from the underlying surface to the atmosphere. is the downward shortwave radiation, α is the surface albedo, and ε a is the effective emissivity of the atmosphere without clouds, T a is the near-surface air temperature, ε s is the surface emissivity, T s is the surface temperature, and σ is the Stefan-Boltzmann constant. establishing the first equation based on the parameterization formulas of the sensible heat flux, the latent heat flux and the soil heat flux; where γ is the wet-bulb constant, γ a is the aerodynamic impedance, ξ is the first coefficient related to the surface vegetation coverage and leaf area index, Δ is the slope of the saturation vapor pressure with respect to temperature, ρ is the air density, C p is the specific heat at constant pressure of air, e 0 (T a ) is the saturation vapor pressure of the underlying surface at temperature T a , and e a is the water vapor pressure of the air near the surface.

3. The method for estimating hourly near-surface air temperature by satellite remote sensing according to claim 1, characterized in that, the algorithm parameters include: surface temperature, annual composite normalized difference vegetation index, 0-5 cm soil volumetric water content, and near-surface specific humidity.

4. The method for estimating hourly near-surface air temperature by satellite remote sensing according to claim 3, characterized in that, obtaining the algorithm parameters includes: obtaining the hourly surface temperature data of the geostationary satellite, and analyzing the hourly surface temperature data to read the satellite observation time, longitude and latitude, cloud-covered pixels, and surface temperature; obtaining the monthly composite vegetation index of the previous year, and analyzing the vegetation index to read the satellite observation month, longitude and latitude, monthly vegetation index, and calculating the annual composite normalized difference vegetation index; obtaining the soil moisture data of the Land Data Assimilation System CLDAS, and analyzing the soil moisture data to read the time, longitude and latitude, 0-5 cm soil volumetric water content; obtaining the near-surface specific humidity data of the Land Data Assimilation System CLDAS, and analyzing the near-surface specific humidity data to read the time, longitude and latitude, near-surface specific humidity; resampling the annual composite normalized difference vegetation index, 0-5 cm soil volumetric water content, and near-surface specific humidity to keep the same spatial resolution as the surface temperature.

5. The method for estimating hourly near-surface air temperature by satellite remote sensing according to claim 1, characterized in that, establishing a spatio-temporal relationship model between the algorithm parameters and the near-surface air temperature based on a geospatial relationship model includes: performing local multiple linear regression solution by the geographically weighted regression method to establish a spatio-temporal relationship model between the algorithm parameters and the near-surface air temperature.

6. The method for estimating hourly near-surface air temperature by satellite remote sensing according to claim 5, characterized in that, establishing a spatio-temporal relationship model between the algorithm parameters and the near-surface air temperature according to the following formula: Among them, Y j is the near-surface air temperature at the point (u j , v j , t j ), where u is the longitude, v is the latitude, t is the time, X ij is the i-th algorithm parameter, ε j is the random error, and β 0 , β i are the first and second fitting coefficients, respectively; the fitting coefficient is solved by the following formula: β(u j ,v j ,t j ) = (X T w(u j ,v j ,t j )X) -1 X T w(u j ,v j ,t j )Y where β(u j , v j , t j ) is the fitting coefficient, w(u j , v j , t j ) is the weight matrix, d jk is the distance between the estimated point j and the surrounding sample point k, and b is the adaptive bandwidth.

7. The method for estimating hourly near-surface air temperature by satellite remote sensing according to claim 1, characterized in that, estimating the hourly near-surface air temperature by using the spatio-temporal relationship model according to the algorithm parameters and meteorological station data includes: Obtain meteorological station data; the meteorological station data includes: meteorological station number, longitude and latitude, altitude, observation time, and thermometer screen temperature. Perform spatio-temporal matching of the land surface temperature and the meteorological station data. Identify the land surface temperature and the meteorological station data after spatio-temporal matching, and estimate the hourly near-surface air temperature using the spatio-temporal relationship model.

8. The method for satellite remote sensing estimation of hourly near-surface air temperature according to claim 1, wherein, the method further includes: Verify the accuracy of the estimated hourly near-surface air temperature.

9. An apparatus for satellite remote sensing estimation of hourly near-surface air temperature, wherein, the apparatus includes: A first calculation unit for establishing a first equation between the near-surface air temperature and the land surface temperature according to the land surface energy balance equation. A first confirmation unit for determining the algorithm parameters required for estimating the near-surface air temperature according to the first equation and obtaining the algorithm parameters. A model establishment unit for establishing a spatio-temporal relationship model between the algorithm parameters and the near-surface air temperature based on the geospatial relationship model. A second calculation unit for estimating the hourly near-surface air temperature using the spatio-temporal relationship model according to the algorithm parameters and the meteorological station data.

10. A computer-readable storage medium, wherein, the computer-readable storage medium stores one or more programs, and the one or more programs can be executed by one or more processors to implement the steps of the method for satellite remote sensing estimation of hourly near-surface air temperature according to any one of claims 1 to 8.

Images