A method and system for evaluating the accuracy of satellite-borne microwave hyperspectral temperature and humidity profile inversion

Abstract

The present invention discloses a satellite-borne microwave hyperspectral temperature and humidity profile inversion accuracy assessment method and system, the method comprising the following steps: obtaining ground-based sounding observation data, screening to obtain clear sky sounding profiles, and performing linear interpolation on the clear sky sounding profiles; constructing a priori error covariance matrix of temperature profiles and humidity profiles; calculating the deviation between the observed brightness temperature and the ideal brightness temperature caused by the center frequency and spectral response function errors, obtaining the observation error, and constructing the instrument observation error covariance matrix based on the instrument error and the observation error; calculating the temperature Jacobian matrix and the humidity Jacobian matrix; constructing the posterior error covariance matrix of the temperature profile and the humidity profile; comparing the posterior error of the temperature profile and the humidity profile with the a priori error covariance matrix, and obtaining the accuracy of the hyperspectral microwave inversion temperature and humidity profiles. The present invention solves the influence of the satellite-borne hyperspectral channel resolution on the inversion accuracy, introduces high-quality references for weather forecasts, and helps to improve the accuracy of weather forecasts.

Description

Technical Field

[0001] The present invention relates to the technical field of microwave remote sensing data inversion accuracy assessment, and in particular to a satellite-borne microwave hyperspectral temperature and humidity profile inversion accuracy assessment method and system. Background Art

[0002] Atmospheric temperature and humidity are two of the most basic meteorological elements. Accurate temperature and humidity profile inversion can effectively improve the accuracy of weather forecasts, which is of great significance to related research such as climate change and environmental change.

[0003] Spaceborne microwave radiometers have always played a very important role in the study of vertical distribution of global atmospheric temperature and humidity. Due to the sensitivity of different channels of microwave radiometers to the atmosphere at different altitudes, they can provide temperature and humidity information from different atmospheric altitudes. At present, the mainstream microwave atmospheric temperature and humidity detection uses microwave radiometers with several or more than a dozen channels, and the channel bandwidth generally ranges from several hundred MHz to several GHz. Therefore, the limited number of channels of microwave radiometers limits the vertical resolution of their observation data.

[0004] With the continuous development of microwave radiation detection technology and the increasing maturity of hyperspectral microwave technology, microwave radiometers can gradually realize spectrum segmentation functions, and the number of channels of microwave radiometers has increased exponentially, even reaching hundreds or thousands of channels, and even achieving channel bandwidths of several MHz or more than ten MHz to obtain fine vertical resolution atmospheric information. The development of hyperspectral microwave radiometers has brought great potential in improving the vertical resolution of temperature and humidity profiles, but no effective solution has been proposed for the application effect and accuracy evaluation of fine hyperspectral microwave radiometer observation results. Summary of the invention

[0005] In order to overcome the defects and shortcomings of the prior art, the present invention provides a method for evaluating the inversion accuracy of a satellite-borne microwave hyperspectral temperature and humidity profile. The present invention evaluates the inversion accuracy of the temperature and humidity profile in real time based on the brightness temperature observation results of a satellite-borne hyperspectral microwave radiometer and the sensitivity related to the channel resolution, solves the influence of the satellite-borne hyperspectral channel resolution on the inversion accuracy, introduces a high-quality reference for weather forecasting, and helps to improve the accuracy of weather forecasts.

[0006] The second object of the present invention is to provide a satellite-borne microwave hyperspectral temperature and humidity profile inversion accuracy assessment system;

[0007] In order to achieve the above object, the present invention adopts the following technical solutions:

[0008] The present invention provides a satellite-borne microwave hyperspectral temperature and humidity profile inversion accuracy assessment method, comprising the following steps:

[0009] Obtain ground-based sounding observation data, filter out clear sky sounding profiles, and perform linear interpolation on the clear sky sounding profiles;

[0010] The prior error covariance matrix of temperature profile and humidity profile is constructed based on the clear sky sounding profile;

[0011] Calculate the ideal brightness temperature corresponding to the ideal center frequency and rectangular spectrum response function, calculate the actual brightness temperature corresponding to the center frequency drift and the actual measured non-ideal channel spectrum response function, compare the difference between the ideal brightness temperature and the actual brightness temperature to obtain the observation error, and construct the instrument observation error covariance matrix based on the instrument error and observation error;

[0012] Calculate the temperature Jacobian matrix and humidity Jacobian matrix based on atmospheric radiation transfer;

[0013] The temperature profile posterior error covariance matrix is ​​constructed based on the temperature Jacobian matrix, the instrument observation error covariance matrix and the temperature profile prior error covariance matrix;

[0014] The humidity profile posterior error covariance matrix is ​​constructed based on the humidity Jacobian matrix, the instrument observation error covariance matrix and the humidity profile prior error covariance matrix;

[0015] The a posteriori errors and a priori error covariance matrices of the temperature and humidity profiles were compared to obtain the accuracy of the hyperspectral microwave inversion temperature and humidity profiles.

[0016] As a preferred technical solution, clear sky sounding profiles are screened out, and the specific steps include:

[0017] When there is an atmospheric layer with a relative humidity greater than a set ratio in the profile of the ground-based sounding observation data, it is determined that there are clouds or precipitation in the corresponding profile of the current ground-based sounding observation data, and the corresponding profile is removed from the ground-based sounding observation data to screen out the clear sky sounding profile;

[0018] Perform linear interpolation on the clear sky sounding profile according to the atmospheric pressure, set the minimum pressure in the profile, divide the atmospheric pressure values ​​into multiple layers, select the sounding profile interpolation within the preset pressure range, or use the standard atmospheric profile linear interpolation to supplement the missing high-altitude atmosphere in the clear sky sounding profile.

[0019] As a preferred technical solution, the prior error covariance matrix of the temperature profile and the humidity profile is constructed based on the clear sky sounding profile, specifically including:

[0020] The priori error covariance matrix of the temperature profile is constructed. The priori error covariance matrix of the temperature profile is symmetric along the diagonal line and is expressed as:

[0021]

[0022] Among them, S TP，arepresents the prior error covariance matrix of the temperature profile, T represents the atmospheric temperature, the subscript represents the number of atmospheric layers corresponding to the profile, var represents variance, and cov represents covariance;

[0023] The prior error covariance matrix of the humidity profile is constructed. The prior error covariance matrix of the humidity profile is symmetric along the diagonal line and is expressed as:

[0024]

[0025] Among them, S RH，a represents the prior error covariance matrix of the humidity profile, RH represents the atmospheric humidity, and the subscript represents the number of the atmospheric layer corresponding to the profile.

[0026] As a preferred technical solution, the instrument error is the channel sensitivity of the instrument, which is specifically expressed as:

[0027]

[0028] Where NEDT is the channel sensitivity, ΔBW is the channel bandwidth, and T sys represents the system noise temperature, and τ represents the integration time of the hyperspectral microwave radiometer.

[0029] As a preferred technical solution, the ideal brightness temperature corresponding to the ideal center frequency and the rectangular spectrum response function is calculated, which is specifically expressed as:

[0030]

[0031] Among them, TB ideal represents the ideal brightness temperature, f1 and f2 are the starting frequency and ending frequency of the channel respectively, ΔBW represents the bandwidth, Δb represents the set frequency, and TB k Indicates the corresponding radiation brightness temperature within the bandwidth.

[0032] As a preferred technical solution, the actual brightness temperature corresponding to the center frequency drift and the actual measured non-ideal channel spectrum response function is calculated, which is specifically expressed as:

[0033]

[0034] Among them, TB real represents the actual brightness temperature, f1 and f2 are the starting frequency and ending frequency of the channel respectively, TB k It represents the corresponding radiation brightness temperature within the bandwidth, Δf represents the center frequency drift, SRF represents the non-ideal channel spectral response function, and ΔBW represents the bandwidth.

[0035] As a preferred technical solution, the instrument observation error covariance matrix is ​​constructed based on the instrument error and the observation error, which is specifically expressed as:

[0036]

[0037] Among them, S e represents the instrument observation error covariance matrix, sqrt represents the square root calculation, NEDT represents the channel sensitivity, that is, the instrument error, and ΔTB represents the observation error.

[0038] As a preferred technical solution, the temperature Jacobian matrix is ​​calculated based on atmospheric radiation transfer, which is specifically expressed as:

[0039]

[0040] Among them, K TP，jl represents the temperature Jacobian matrix, j represents the number of atmospheric layers, l represents the number of channels of the hyperspectral microwave radiometer, TB1 represents the brightness temperature of the lth channel of the hyperspectral microwave radiometer, TB2 represents the corresponding brightness temperature calculated under the temperature disturbance of the jth layer in the atmospheric profile, represents temperature disturbance;

[0041] The humidity Jacobian matrix is ​​calculated based on atmospheric radiation transfer, which is specifically expressed as:

[0042]

[0043] Among them, K RH，jl represents the humidity Jacobian matrix, TB3 represents the corresponding brightness temperature calculated under the humidity disturbance of the jth layer in the atmospheric profile, Represents humidity disturbance.

[0044] As a preferred technical solution, the temperature profile posterior error covariance matrix is ​​expressed as:

[0045]

[0046] Among them, K TP is the temperature Jacobian matrix, S e is the instrument observation error covariance matrix, S TP，a is the prior error covariance matrix of the temperature profile;

[0047] The posterior error covariance matrix of humidity profile is expressed as:

[0048]

[0049] Among them, K RH is the humidity Jacobian matrix, S e is the instrument observation error covariance matrix, S RH，a is the humidity prior error covariance matrix;

[0050] By comparing the a posteriori error and a priori error covariance matrix of temperature and humidity, the accuracy of the hyperspectral microwave inversion temperature and humidity profiles is obtained, which is specifically expressed as:

[0051] ΔS TP =S TP -S TP，a

[0052] ΔS RH =S RH -S RH，a

[0053] Among them, S TP represents the temperature profile posterior error matrix, S RH represents the humidity profile posterior error matrix.

[0054] The present invention also provides a satellite-borne microwave hyperspectral temperature and humidity profile inversion accuracy assessment system, comprising: a data preprocessing module, a priori error calculation module, an instrument observation error processing module, a Jacobian matrix calculation module, a posteriori error processing module, and a temperature and humidity inversion accuracy assessment module;

[0055] The data preprocessing module is used to obtain ground-based sounding observation data, screen and obtain clear sky sounding profiles, and perform linear interpolation on the clear sky sounding profiles;

[0056] The prior error calculation module is used to construct a prior error covariance matrix of the temperature profile and the humidity profile based on the clear sky sounding profile;

[0057] The instrument observation error processing module is used to calculate the ideal brightness temperature corresponding to the ideal center frequency and the rectangular spectrum response function, calculate the actual brightness temperature corresponding to the center frequency drift and the actually measured non-ideal channel spectrum response function, compare the difference between the ideal brightness temperature and the actual brightness temperature to obtain the observation error, and construct the instrument observation error covariance matrix based on the instrument error and the observation error;

[0058] The Jacobian matrix calculation module is used to calculate the temperature Jacobian matrix and the humidity Jacobian matrix based on atmospheric radiation transmission;

[0059] The a posteriori error processing module is used to construct a temperature profile a posteriori error covariance matrix based on the temperature Jacobian matrix, the instrument observation error covariance matrix and the a priori error covariance matrix of the temperature profile;

[0060] The humidity profile posterior error covariance matrix is ​​constructed based on the humidity Jacobian matrix, the instrument observation error covariance matrix and the humidity profile prior error covariance matrix;

[0061] The temperature and humidity inversion accuracy assessment module is used to compare the a posteriori errors and a priori error covariance matrices of the temperature profile and the humidity profile to obtain the accuracy of the hyperspectral microwave inversion temperature and humidity profile.

[0062] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0063] (1) The present invention quantifies the influence of brightness temperature error caused by channel sensitivity, channel center frequency drift and channel spectral response function of the hyperspectral microwave radiometer on the temperature and humidity profile inversion results, achieving the technical effect of accurately evaluating the inversion accuracy of the hyperspectral temperature and humidity profile.

[0064] (2) The technical solution of the present invention, based on error covariance matrix analysis, improves the computational efficiency and constructs an evaluation system for the inversion accuracy of microwave hyperspectral temperature and humidity profiles, objectively describing the effectiveness of on-orbit detection of hyperspectral microwave radiometers. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] Figure 1 It is a schematic diagram of the process architecture of the satellite-borne microwave hyperspectral temperature and humidity profile inversion accuracy assessment method of the present invention;

[0066] Figure 2 It is a schematic diagram of the architecture of the satellite-borne microwave hyperspectral temperature and humidity profile inversion accuracy assessment system of the present invention. DETAILED DESCRIPTION

[0067] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0068] Example 1

[0069] like Figure 1 As shown, this embodiment provides a method for evaluating the accuracy of satellite-borne microwave hyperspectral temperature and humidity profile inversion, including the following steps:

[0070] S1: Select data;

[0071] S11: Acquire ground-based sounding observation data;

[0072] According to the latitude and longitude information of the ground sounding station, the observation data of the spaceborne hyperspectral microwave radiometer that matches the time and space of the sounding station are selected, and the observation data of the sounding station are used to eliminate the data with clouds and precipitation.

[0073] When the distance between the observation pixel of the satellite-borne hyperspectral microwave radiometer and the sounding site is less than 15 km, and the transit time of the satellite at the sounding site is within 1 hour of the flight time of the sounding balloon, the observation pixel of the satellite is considered to match the sounding data in time and space.

[0074] S12: Eliminate the profiles corresponding to the cloud and rain scenarios and select the clear sky sounding profiles;

[0075] When there is an atmospheric layer with a relative humidity greater than 95% in the profile of the ground-based sounding observation data, it is considered that there are clouds or precipitation in the corresponding profile of the sounding data, and the corresponding profile is removed from the sounding data set of the ground sounding station.

[0076] S13: Linear interpolation of clear sky sounding profiles based on atmospheric pressure;

[0077] To ensure the acquisition of observation information of the hyperspectral microwave radiometer, the minimum pressure in the profile is set to 5 mbar. The atmospheric pressure P value is divided into 30 layers, and the P values ​​are 5, 25, 50, 75, 100, 125, 150, 175, 200, 225, 300, 350, 400, 450, 500, 550, 600, 650, 700, 750, 775, 800, 825, 850, 875, 900, 925, 950, 975, and 1000 mbar.

[0078] Since the maximum flight altitude of the sounding balloon is 20-30 km, the sounding profile interpolation is selected in the range of 125-1000 mbar, and the linear interpolation of the US Standard Atmosphere (1976) profile is used in the range of 5-125 mbar to supplement the missing upper atmosphere in the sounding profile in the pressure range of 5-125 mbar.

[0079] S2: Construct the a priori error covariance matrix of temperature profile and humidity profile;

[0080] In the clear sky sounding profile obtained in step S1, the atmospheric temperature and humidity are expressed as Wherein, the superscript i is the i-th profile obtained according to step S1; the subscript j corresponds to the j-th atmospheric layer of the profile, and there are 30 layers in total.

[0081] A priori error covariance matrix S of the temperature profile TP，a Symmetrical along the diagonal, expressed as follows:

[0082]

[0083] Among them, var represents variance and cov represents covariance;

[0084] The prior error covariance matrix S of the humidity profile RH，a Symmetrical along the diagonal, expressed as follows:

[0085]

[0086] S3: Construct the instrument observation error covariance matrix S e ;

[0087] The instrument observation error is the sum of the instrument error and the observation error. The instrument error is the channel sensitivity NEDT of the instrument, and the observation error is caused by factors such as the channel center frequency drift and the channel bandwidth spectrum response function.

[0088] S31: Calculate the channel sensitivity NEDT of the instrument;

[0089] Channel sensitivity NEDT is related to channel bandwidth, and the relationship between the two is as follows:

[0090]

[0091] Wherein, ΔBW is the channel bandwidth, which is in the range of 3-200 MHz according to the channel design of the hyperspectral microwave radiometer; τ is the integration time of the hyperspectral microwave radiometer, in seconds; T sys is the system noise temperature, in K.

[0092] Different channel bandwidths correspond to different channel sensitivities. When the channel bandwidth of the satellite-borne hyperspectral microwave radiometer changes within the range of 3MHz-200MHz, the channel sensitivity NEDT is recalculated according to the above formula.

[0093] S32: Calculate observation error;

[0094] Since the center frequency may be offset (1-2MHz), the spectral response function is a non-rectangular window. The center frequency and spectral response function errors cause the observed brightness temperature to deviate from the ideal brightness temperature, which is the observation error.

[0095] First, calculate the ideal brightness temperature TB corresponding to the ideal center frequency and rectangular spectrum response function iaeal According to the atmospheric radiation transmission simulation, the corresponding radiation brightness temperature within the microwave hyperspectral bandwidth is simulated with a frequency resolution of 0.1 MHz. Since the ideal spectrum response function is a rectangular window, the channel brightness temperature TB is the average brightness temperature within the channel bandwidth, which is specifically expressed as:

[0096]

[0097] Where f1 and f2 are the start and end frequencies of the channel (in MHz), bandwidth ΔBW = f2-f1, Δb is 0.1MHz, TB k is the corresponding radiation brightness temperature within the bandwidth.

[0098] Then, the actual brightness temperature TB corresponding to the center frequency drift Δf and the actual measured non-ideal channel spectral response function SRF is calculated. real , specifically expressed as:

[0099]

[0100] Finally, the difference between the ideal brightness temperature and the actual brightness temperature is compared to obtain the observation error, which is set as ΔTB.

[0101] ΔTB=TB real -TB ideal

[0102] S33: Construct instrument observation error covariance matrix S e ;

[0103] According to step S31 and step S32, the instrument observation error covariance matrix is ​​obtained. The instrument observation error covariance matrix S e It is expressed as:

[0104]

[0105] S4: Construct Jacobian matrix;

[0106] S41: construct temperature Jacobian matrix;

[0107] The temperature Jacobian matrix is ​​calculated using atmospheric radiation transfer simulation. First, the ground sounding profile is input into the atmospheric radiation transfer simulation, and the brightness temperature of the lth channel of the hyperspectral microwave radiometer is calculated as TB1. Assume that the temperature perturbation of the jth layer in the atmospheric profile is The corresponding brightness temperature is TB2. The corresponding temperature Jacobian matrix element is expressed as:

[0108]

[0109] The jth row corresponds to the jth atmospheric layer, and the lth column is the lth channel of the hyperspectral microwave radiometer.

[0110] S42: construct humidity Jacobian matrix;

[0111] In the atmospheric radiation transmission simulation, the ground sounding profile is input and the brightness temperature of the lth channel of the hyperspectral microwave radiometer is calculated as TB1. is 2%, and the corresponding brightness temperature is calculated as TB3, then the humidity Jacobian matrix element is obtained as:

[0112]

[0113] S5: Evaluate the accuracy of temperature and humidity profiles;

[0114] S51: Construct the posterior error covariance matrix of temperature and humidity profiles;

[0115] The accuracy of the inverted temperature and humidity profiles is evaluated based on the posterior error covariance matrix. TP It is expressed as:

[0116]

[0117] Among them, S TP is the temperature profile posterior error matrix, K TP is the Jacobian matrix, S e is the instrument observation error covariance matrix, S TP，a is the a priori error covariance matrix of the temperature profile.

[0118] For the humidity profile, the posterior error covariance matrix S RH for

[0119]

[0120] Among them, S RH is the humidity profile posterior error, K RH is the Jacobian matrix, S e is the instrument observation error covariance matrix, S RH，a is the humidity prior error covariance matrix.

[0121] S52: perform profile accuracy assessment;

[0122] By comparing the a posteriori error and a priori error covariance matrix of the temperature profile and humidity profile, the accuracy of the hyperspectral microwave inversion temperature and humidity profile can be obtained, which is specifically expressed as:

[0123] ΔS TP =S TP -S TP，a

[0124] ΔS RH =S RH -S RH，a

[0125] The present invention takes into account the influence of instrument bandwidth and instrument observation error on the inversion accuracy of temperature and humidity profiles, proposes a profile accuracy evaluation method based on instrument observation error designed based on channel bandwidth, and realizes the detection performance and detection potential evaluation of hyperspectral microwave radiometers.

[0126] Example 2

[0127] like Figure 2 As shown, this embodiment proposes a satellite-borne microwave hyperspectral temperature and humidity profile inversion accuracy assessment system, including: a data preprocessing module, a priori error calculation module, an instrument observation error processing module, a Jacobian matrix calculation module, a posteriori error processing module, and a temperature and humidity inversion accuracy assessment module;

[0128] In this embodiment, the data preprocessing module is used to obtain ground-based sounding observation data, screen and obtain clear sky sounding profiles, and perform linear interpolation on the clear sky sounding profiles;

[0129] In this embodiment, the prior error calculation module is used to construct a priori error covariance matrix of the temperature profile and the humidity profile based on the clear sky sounding profile;

[0130] In this embodiment, the instrument observation error processing module is used to calculate the ideal brightness temperature corresponding to the ideal center frequency and the rectangular spectrum response function, calculate the center frequency drift and the actual brightness temperature corresponding to the non-ideal channel spectrum response function actually measured, compare the difference between the ideal brightness temperature and the actual brightness temperature to obtain the observation error, and construct the instrument observation error covariance matrix based on the instrument error and the observation error;

[0131] In this embodiment, the Jacobian matrix calculation module is used to calculate the temperature Jacobian matrix and the humidity Jacobian matrix based on atmospheric radiation transmission;

[0132] In this embodiment, the a posteriori error processing module is used to construct a temperature profile a posteriori error covariance matrix based on the temperature Jacobian matrix, the instrument observation error covariance matrix and the a priori error covariance matrix of the temperature profile;

[0133] The humidity profile posterior error covariance matrix is ​​constructed based on the humidity Jacobian matrix, the instrument observation error covariance matrix and the humidity profile prior error covariance matrix;

[0134] In this embodiment, the temperature and humidity inversion accuracy assessment module is used to compare the a posteriori errors and a priori error covariance matrices of the temperature profile and the humidity profile to obtain the accuracy of the hyperspectral microwave inversion temperature and humidity profile.

[0135] The above embodiments are preferred implementation modes of the present invention, but the implementation modes of the present invention are not limited to the above embodiments. Any other changes, modifications, substitutions, combinations, and simplifications that do not deviate from the spirit and principles of the present invention should be equivalent replacement methods and are included in the protection scope of the present invention.

Claims

1. A method for evaluating the accuracy of satellite-borne microwave hyperspectral temperature and humidity profile inversion, characterized in that: The steps include: Obtain ground-based sounding observation data, filter out clear sky sounding profiles, and perform linear interpolation on the clear sky sounding profiles; The prior error covariance matrix of temperature profile and humidity profile is constructed based on the clear sky sounding profile; Calculate the ideal brightness temperature corresponding to the ideal center frequency and rectangular spectrum response function, calculate the actual brightness temperature corresponding to the center frequency drift and the actual measured non-ideal channel spectrum response function, compare the difference between the ideal brightness temperature and the actual brightness temperature to obtain the observation error, and construct the instrument observation error covariance matrix based on the instrument error and observation error; The instrument error is the channel sensitivity of the instrument, which is expressed as: Where NEDT is the channel sensitivity, ΔBW is the channel bandwidth, and T sys represents the system noise temperature, τ represents the integration time of the hyperspectral microwave radiometer; Calculate the ideal brightness temperature corresponding to the ideal center frequency and rectangular spectrum response function, which is specifically expressed as: Among them, TB ideal represents the ideal brightness temperature, f1 and f2 are the starting frequency and ending frequency of the channel respectively, Δb represents the set frequency, TB k Indicates the corresponding radiation brightness temperature within the bandwidth; The actual brightness temperature corresponding to the center frequency drift and the actual measured non-ideal channel spectrum response function is calculated, which is specifically expressed as: Among them, TB real represents the actual brightness temperature, Δf represents the center frequency drift, and SRF represents the non-ideal channel spectral response function; Calculate the temperature Jacobian matrix and humidity Jacobian matrix based on atmospheric radiation transfer; The temperature profile posterior error covariance matrix is ​​constructed based on the temperature Jacobian matrix, the instrument observation error covariance matrix and the temperature profile prior error covariance matrix; The humidity profile posterior error covariance matrix is ​​constructed based on the humidity Jacobian matrix, the instrument observation error covariance matrix and the humidity profile prior error covariance matrix; The a posteriori errors and a priori error covariance matrices of the temperature and humidity profiles were compared to obtain the accuracy of the hyperspectral microwave inversion temperature and humidity profiles.

2. The satellite-borne microwave hyperspectral temperature and humidity profile inversion accuracy assessment method according to claim 1 is characterized in that: Screen out the clear sky sounding profiles. The specific steps include: When there is an atmospheric layer with a relative humidity greater than a set ratio in the profile of the ground-based sounding observation data, it is determined that there are clouds or precipitation in the corresponding profile of the current ground-based sounding observation data, and the corresponding profile is removed from the ground-based sounding observation data to screen out the clear sky sounding profile; Perform linear interpolation on the clear sky sounding profile according to the atmospheric pressure, set the minimum pressure in the profile, divide the atmospheric pressure values ​​into multiple layers, select the sounding profile interpolation within the preset pressure range, or use the standard atmospheric profile linear interpolation to supplement the missing high-altitude atmosphere in the clear sky sounding profile.

3. The satellite-borne microwave hyperspectral temperature and humidity profile inversion accuracy assessment method according to claim 1 is characterized in that: The prior error covariance matrix of the temperature profile and the humidity profile constructed based on the clear sky sounding profile specifically includes: The priori error covariance matrix of the temperature profile is constructed. The priori error covariance matrix of the temperature profile is symmetric along the diagonal line and is expressed as: Among them, S TP,a represents the prior error covariance matrix of the temperature profile, T represents the atmospheric temperature, the subscript represents the number of atmospheric layers to which the profile corresponds, var represents variance, and cov represents covariance; The prior error covariance matrix of the humidity profile is constructed. The prior error covariance matrix of the humidity profile is symmetric along the diagonal line and is expressed as: Among them, S RH,a represents the prior error covariance matrix of the humidity profile, RH represents the atmospheric humidity, and the subscript represents the number of the atmospheric layer corresponding to the profile.

4. The satellite-borne microwave hyperspectral temperature and humidity profile inversion accuracy assessment method according to claim 1 is characterized in that: The instrument observation error covariance matrix is ​​constructed based on the instrument error and observation error, which is specifically expressed as: Among them, S e represents the instrument observation error covariance matrix, sqrt represents the square root calculation, NEDT represents the channel sensitivity, that is, the instrument error, and ΔTB represents the observation error.

5. The satellite-borne microwave hyperspectral temperature and humidity profile inversion accuracy assessment method according to claim 1 is characterized in that: The temperature Jacobian matrix is ​​calculated based on atmospheric radiation transfer, which is specifically expressed as: Among them, K TP,jl represents the temperature Jacobian matrix, j represents the number of atmospheric layers, l represents the number of channels of the hyperspectral microwave radiometer, TB1 represents the brightness temperature of the lth channel of the hyperspectral microwave radiometer, TB2 represents the corresponding brightness temperature calculated under the temperature disturbance of the jth layer in the atmospheric profile, represents temperature disturbance; The humidity Jacobian matrix is ​​calculated based on atmospheric radiation transfer, which is specifically expressed as: Among them, K RH,jl represents the humidity Jacobian matrix, TB3 represents the corresponding brightness temperature calculated under the humidity disturbance of the jth layer in the atmospheric profile, Represents humidity disturbance.

6. The satellite-borne microwave hyperspectral temperature and humidity profile inversion accuracy assessment method according to claim 1 is characterized in that: The posterior error covariance matrix of the temperature profile is expressed as: Among them, K TP is the temperature Jacobian matrix, S e is the instrument observation error covariance matrix, S TP,a is the prior error covariance matrix of the temperature profile; The posterior error covariance matrix of humidity profile is expressed as: Among them, K RH is the humidity Jacobian matrix, S RH,a is the humidity prior error covariance matrix; By comparing the a posteriori error and a priori error covariance matrix of temperature and humidity, the accuracy of the hyperspectral microwave inversion temperature and humidity profiles is obtained, which is specifically expressed as: ΔS TP =S TP -S TP,a ΔS RH =S RH -S RH,a Among them, S TP represents the temperature profile posterior error matrix, S RH represents the humidity profile posterior error matrix.

7. A satellite-borne microwave hyperspectral temperature and humidity profile inversion accuracy assessment system, characterized in that: include: Data preprocessing module, a priori error calculation module, instrument observation error processing module, Jacobian matrix calculation module, a posteriori error processing module, temperature and humidity inversion accuracy assessment module; The data preprocessing module is used to obtain ground-based sounding observation data, screen and obtain clear sky sounding profiles, and perform linear interpolation on the clear sky sounding profiles; The prior error calculation module is used to construct a priori error covariance matrix of temperature profile and humidity profile based on clear sky sounding profile; The instrument observation error processing module is used to calculate the ideal brightness temperature corresponding to the ideal center frequency and the rectangular spectrum response function, calculate the actual brightness temperature corresponding to the center frequency drift and the actually measured non-ideal channel spectrum response function, compare the difference between the ideal brightness temperature and the actual brightness temperature to obtain the observation error, and construct the instrument observation error covariance matrix based on the instrument error and the observation error; The instrument error is the channel sensitivity of the instrument, which is expressed as: Where NEDT is the channel sensitivity, ΔBW is the channel bandwidth, and T sys represents the system noise temperature, τ represents the integration time of the hyperspectral microwave radiometer; Calculate the ideal brightness temperature corresponding to the ideal center frequency and rectangular spectrum response function, which is specifically expressed as: Among them, TB ideal represents the ideal brightness temperature, f1 and f2 are the starting frequency and ending frequency of the channel respectively, Δb represents the set frequency, TB k Indicates the corresponding radiation brightness temperature within the bandwidth; The actual brightness temperature corresponding to the center frequency drift and the actual measured non-ideal channel spectrum response function is calculated, which is specifically expressed as: Among them, TB real represents the actual brightness temperature, Δf represents the center frequency drift, and SRF represents the non-ideal channel spectral response function; The Jacobian matrix calculation module is used to calculate the temperature Jacobian matrix and the humidity Jacobian matrix based on atmospheric radiation transmission; The a posteriori error processing module is used to construct a temperature profile a posteriori error covariance matrix based on the temperature Jacobian matrix, the instrument observation error covariance matrix and the a priori error covariance matrix of the temperature profile; The humidity profile posterior error covariance matrix is ​​constructed based on the humidity Jacobian matrix, the instrument observation error covariance matrix and the humidity profile prior error covariance matrix; The temperature and humidity inversion accuracy assessment module is used to compare the a posteriori errors and a priori error covariance matrices of the temperature profile and the humidity profile to obtain the accuracy of the hyperspectral microwave inversion temperature and humidity profile.

Images