阅读说明：本技术 一种使用后向散射截面校正粗糙海面亮温的方法 (Method for correcting brightness and temperature of rough sea surface by using backscattering cross section ) 是由 马文韬 于暘 刘桂红 杨晓峰 杜延磊 李紫薇 于 2019-10-14 设计创作，主要内容包括：本发明公开了一种使用后向散射截面校正粗糙海面亮温的方法,该方法应用于海洋微波遥感中的盐度遥感,用于粗糙海面亮温校正,适用于Aquarius卫星的主被动联合观测体制,与原有的先利用散射计反演海表风速或获取辅助风速,再用风速校正海面粗糙度对亮温的影响不同,该方法直接利用散射计观测的NRCS和辅助的海表风向信息,校正海表粗糙度对海面亮温的影响,无需使用辅助的风速数据对海表面亮温进行校正,该方法使用的参数更少、精度更高,且该方法处理过程简单、成本低、精度高、易于操作。(The invention discloses a method for correcting rough sea surface brightness temperature by using a backscattering cross section, which is applied to salinity remote sensing in ocean microwave remote sensing, is used for correcting rough sea surface brightness temperature, is suitable for an active and passive combined observation system of an Aquarius satellite, is different from the original method that sea surface wind speed is inverted or auxiliary wind speed is obtained by using a scatterometer, and sea surface roughness is corrected by using wind speed to influence brightness temperature.)

1. A method for correcting rough sea surface brightness temperature by using backscattering cross section is characterized by comprising the following steps:

the method comprises the following steps: reading and preprocessing scatterometry data;

extracting the backscattering cross section data of the horizontal emission level receiving scatterometer from the L2 level data of the Aquarius satellite, and recording the backscattering cross section data as scat _ HH _ toa; then, reducing the scat _ HH _ toa by adopting a logarithmic reduction model to obtain reduced horizontal emission level receiving backscattering cross section data R sigma _ HH;

extracting scatterometer vertical transmitting and vertical receiving backscattering section data from Aquarius satellite L2 level data and recording the backscattering section data as scat _ VV _ toa; then, reducing the scat _ VV _ toa by adopting a logarithmic reduction model to obtain R sigma _ VV of the reduced vertical transmitting and vertical receiving back scattering cross section data;

step two: acquiring wind direction data and adjusting a wind direction angle;

sea surface wind direction data are extracted from the Aquarius satellite L2 level data and recorded as anc _ wind _ dir;

extracting an instrument azimuth angle marked as celphi from L2 level data of the Aquarius satellite;

then using the wind direction correction model

If it is

If it is

Step three: calculating the sea surface emissivity increment by using a scattering emissivity model;

calculating the sea surface emissivity increment by using a scattering emissivity model as follows:

ew_NRCSindicating the emissivity increase.

beam represents the angle of incidence;

p represents the polarization mode of the radiometer;

r σ represents the reduced backscatter cross-section data;

representing the wind direction angle relative to the satellite observation;

A_n(beam, p, R sigma) represents a cosine coefficient function, the lower corner mark n represents a coefficient identification number, and the value of n is 0, 1, 2 and 4); wherein A is₀(beam, p, R σ) represents a cosine coefficient function of order 0; a. the₁(beam, p, R σ) represents a first order cosine coefficient function; a. the₂(beam, p, R σ) represents a second order cosine coefficient function; a. the₄(beam, p, R σ) represents a fourth order cosine coefficient function;

a is described_nThe calculation of (beam, p, R σ) is:

a_n,iexpressing a cosine emissivity model coefficient, wherein a lower subscript n expresses a coefficient identification number, and a lower subscript i expresses an order identification number;

step four: obtaining sea surface temperature data;

extracting corrected sea surface temperature of brightness temperature data at the same time and the same longitude and latitude from the L2 level data of the Aquarius satellite, and recording the sea surface temperature as anc _ surface _ temp;

step five: performing sea surface brightness temperature correction;

the horizontally polarized sea surface brightness temperature for removing the influence of the rough sea surface is as follows:

TB_flat,H＝rad_TbH-ew_NRCS×anc_surface_temp

TB_flat,Hcalculating the horizontal polarization calm sea surface brightness temperature;

rad _ TbH is horizontal polarization sea surface brightness temperature data;

the vertical polarization sea surface brightness temperature for removing the influence of the rough sea surface is as follows:

TB_flat,V＝rad_TbV-ew_NRCS×anc_surface_temp

TB_flat,Vcalculating the vertical polarization calm sea surface brightness temperature;

rad _ TbV is vertical polarization sea surface light temperature data.

2. The method of using backscattering cross-sections to correct rough sea surface light temperature according to claim 1, wherein: the log reduction model is:

r σ represents the reduced backscatter cross-section data; σ represents backscattering cross-sectional data; "Λ" represents the sign of the power.

3. The method of claim 1, wherein the method further comprises the step of correcting the rough sea surface brightness temperature using a backscattering cross section: the rough sea surface brightness temperature is directly corrected by using the backscattering section without inverting the wind speed firstly.

4. The method of claim 1, wherein the method further comprises the step of correcting the rough sea surface brightness temperature using a backscattering cross section: the scatterometer and radiometer observe the sea surface using the same angle of incidence and the same azimuth.

5. The method of claim 1, wherein the method further comprises the step of correcting the rough sea surface brightness temperature using a backscattering cross section: the radiometer model is an Aquarius satellite L-band microwave radiometer, and the scatterometer model is an Aquarius satellite L-band microwave scatterometer.

Technical Field

The invention relates to a sea surface salinity inversion method in the field of ocean microwave remote sensing, in particular to a method for correcting rough sea surface brightness temperature by using a backscattering section.

Background

The observation of sea surface salinity has important significance for researching and predicting global climate change, monitoring and forecasting of ocean circulation, ocean water mass and the like, and the distribution of the sea surface salinity is closely related to seawater evaporation and rainfall space-time change. Satellite salinity remote sensing is the most effective means for acquiring sea surface salinity data with global scope, long time sequence and wide coverage. The Aquarius satellite (i.e. the Baozhen satellite) is a satellite specially designed for salinity remote sensing, adopts an active and passive combined observation means, combines a scatterometer of 1.26GHz and a radiometer of 1.413GHz to observe the same position of the sea surface at the same angle, and plans and inverts to obtain the global sea surface salinity data with the resolution of 150km and the monthly average precision superior to 0.2psu (practical standard salinity unit).

The brightness temperature observed by the radiometer not only contains the influence of salinity change, but also comprises the influence of other factors, wherein the influence of sea surface roughness on the brightness temperature is the most main limiting factor for improving the salinity remote sensing precision. The conventional roughness correction method needs to firstly utilize a scatterometer to invert the sea surface wind speed or obtain other auxiliary wind speeds, and then utilizes the wind speed to correct the brightness temperature, so that the method has complex steps on one hand, and is easier to introduce model errors on the other hand. And the Aquarius satellite uses a radiometer and a scatterometer to observe the same position of the sea surface by using the same angle, so that backscattering cross section data (NRCS for short) observed by the scatterometer can be used for directly correcting the bright temperature without inverting the wind speed firstly.

Disclosure of Invention

In order to correct the sea surface roughness influence which is the most main influence on the sea surface brightness temperature during sea surface salinity inversion, the invention provides a method for correcting the rough sea surface brightness temperature by using a backscattering section. The method is based on an active and passive united observation system of an Aquarius satellite, utilizes a radiometer and a scatterometer which observe at the same incident angle and the same azimuth angle, utilizes backscattering cross section and auxiliary wind direction data observed by the scatterometer, and directly calculates and obtains rough sea surface bright temperature increment caused by wind by the method. Compared with the traditional correction algorithm, the method does not need to acquire auxiliary wind speed data and does not need to invert the sea surface wind speed of the intermediate variable firstly. The correction accuracy of the inventive method is better than that of the data using the auxiliary NCEP (national environmental prediction center).

The invention relates to a method for correcting brightness and temperature of a rough sea surface by using a backscattering section, which is characterized by comprising the following steps of:

the method comprises the following steps: reading and preprocessing scatterometry data;

extracting the backscattering cross section data of the horizontal emission level receiving scatterometer from the L2 level data of the Aquarius satellite, and recording the backscattering cross section data as scat _ HH _ toa; then, reducing the scat _ HH _ toa by adopting a logarithmic reduction model to obtain reduced horizontal emission level receiving backscattering cross section data R sigma _ HH;

extracting scatterometer vertical transmitting and vertical receiving backscattering section data from Aquarius satellite L2 level data and recording the backscattering section data as scat _ VV _ toa; then, reducing the scat _ VV _ toa by adopting a logarithmic reduction model to obtain reduced vertical transmitting and vertical receiving backscattering cross section data R sigma _ VV;

step two: acquiring wind direction data and adjusting a wind direction angle;

sea surface wind direction data are extracted from the Aquarius satellite L2 level data and recorded as anc _ wind _ dir;

extracting an instrument azimuth angle marked as celphi from L2 level data of the Aquarius satellite;

then using the wind direction correction model

Calculating to obtain the wind direction angle relative to the satellite observation

If it is

Less than 0 deg. and then through adding 360 deg. to regulate to

Within the range.

If it is

If greater than 360 DEG, the angle is adjusted to

Within the range.

Step three: calculating the sea surface emissivity increment by using a scattering emissivity model;

calculating the sea surface emissivity increment by using a scattering emissivity model as follows:

cosine coefficient function A_nThe calculation of (beam, p, R σ) is:

step four: obtaining sea surface temperature data;

extracting corrected sea surface temperature of brightness temperature data at the same time and the same longitude and latitude from the L2 level data of the Aquarius satellite, and recording the sea surface temperature as anc _ surface _ temp;

step five: performing sea surface brightness temperature correction;

the horizontally polarized sea surface brightness temperature for removing the influence of the rough sea surface is as follows:

TB_flat,H＝rad_TbH-ew_NRCS×anc_surface_temp

TB_flat,Hcalculating the horizontal polarization calm sea surface brightness temperature;

rad _ TbH is horizontal polarization sea surface brightness temperature data;

the vertical polarization sea surface brightness temperature for removing the influence of the rough sea surface is as follows:

TB_flat,V＝rad_TbV-ew_NRCS×anc_surface_temp

TB_flat,Vcalculating the vertical polarization calm sea surface brightness temperature;

rad _ TbV is vertical polarization sea surface light temperature data.

In the present invention, the log reduction model is:

in the invention, the rough sea surface brightness temperature is directly corrected by using the backscattering section without inverting the wind speed firstly.

In the present invention, the scatterometer and radiometer observe the sea surface using the same angle of incidence and the same azimuth.

Compared with the traditional method, the method for correcting the rough sea surface brightness temperature by using the backscattering cross section has the advantages that:

① the method of the invention directly establishes a relation model of rough sea surface emissivity and scatterometer backscattering cross section, reduces model error introduced by inverting wind speed and then calculating emissivity, and has simpler model and faster operation speed.

② the method of the present invention can directly take advantage of the combined observation of radiometer and scatterometer, and accurately correct the portion of radiometer affected by roughness by using the characteristic that scatterometer is sensitive to sea surface roughness and has higher sensitivity.

③ the invention corrects the influence of roughness on sea surface brightness temperature by using emissivity increment and sea surface temperature obtained by calculating backscattering cross section, and the corrected calm sea surface brightness temperature can be directly used for sea surface salinity inversion, thus obviously reducing the error of salinity inversion.

Drawings

FIG. 1 is a flow chart of the present invention using backscattering cross-sections to correct for the effect of sea surface roughness on light temperature.

FIG. 2A is a graph of emissivity increase for vertical polarization as a function of R σ for vertical transmission and vertical reception and wind direction.

Fig. 2B is a graph of the emissivity increase for vertical polarization as a function of horizontal transmit-horizontal receive R σ and wind direction.

FIG. 2C is a graph of horizontally polarized emissivity gain versus vertical received R σ and wind direction for vertical transmission.

FIG. 2D is a graph of horizontally polarized emissivity increase as a function of R σ and wind direction received at the horizontal transmission level.

FIG. 3 is a plot of the emissivity gain calculated by the model as a function of wind direction and NRCS (beam3, V polarization emissivity gain calculated by the HH polarized NRCS).

FIG. 4A is a diagram of the variation of the root mean square error of the increment of the vertical polarization emissivity obtained by model calculation and the actually measured increment of the vertical polarization emissivity with the wind direction and the R sigma of vertical transmission and vertical reception.

FIG. 4B is a diagram of the variation of the RMS error of the increment of the vertical polarization emissivity calculated by the model and the actually measured increment of the vertical polarization emissivity with the wind direction and the R sigma received horizontally by the horizontal transmission.

FIG. 4C is a diagram of the variation of the RMS error of the incremental horizontal polarization emissivity calculated by the model and the incremental measured horizontal polarization emissivity with the wind direction and the R σ received vertically by the vertical transmission.

FIG. 4D is a diagram of the RMS error of the model calculated increase in horizontal polarization emissivity versus the measured increase in horizontal polarization emissivity as a function of wind direction and horizontal transmit level received R σ.

Fig. 5A is a graph comparing the corrected beam1 vertical polarization calm sea surface light temperature to a theoretical estimate of the light temperature.

Fig. 5B is a graph comparing the corrected beam2 vertical polarization calm sea light temperature to a theoretical estimate of the light temperature.

Fig. 5C is a graph comparing the corrected beam3 vertical polarization calm surface light temperature to a theoretical estimate of the light temperature.

Fig. 5D is a graph comparing the corrected beam1 horizontal polarization calm sea light temperature with the theoretical estimate of light temperature.

Fig. 5E is a graph comparing the corrected beam2 horizontal polarization calm sea light temperature to the theoretical estimate of light temperature.

Fig. 5F is a graph comparing the corrected beam3 horizontal polarization calm sea light temperature to the theoretical estimate of light temperature.

Detailed Description

The present invention will be described in further detail with reference to the accompanying drawings and examples.

The invention aims at a system that a radiometer and a scatterometer are adopted to observe the same position at the same incidence angle and azimuth angle, backscattering cross section data (NRCS for short) obtained by the scatterometer is used for correcting sea surface roughness influence in bright temperature obtained by the radiometer, and at present, a satellite using the system is mainly an Aquarius satellite.

Description of the data

The radiometer model is an Aquarius satellite L-band microwave radiometer, and the scatterometer model is an Aquarius satellite L-band microwave scatterometer. The same position of the sea surface was observed using the Aquarius satellite, which observed the sea surface using three angles of incidence, a first angle of incidence designated beam1, a second angle of incidence designated beam2, and a third angle of incidence designated beam3, which were 28.7 degrees, 37.8 degrees, and 45.6 degrees, respectively. The radiometer used in the present invention includes a horizontal polarization state H and a vertical polarization state V, and the radiometer used includes a horizontal transmission and horizontal reception state HH and a vertical transmission and vertical reception state VV. The data used by the invention is Aquarius satellite L2 level data, which specifically comprises the following data: recording the horizontal polarization sea surface brightness temperature data as rad _ TbH, and recording the vertical polarization sea surface brightness temperature data as rad _ TbV; the horizontal emission horizontal receiving backscattering section data of the scatterometer is recorded as scat _ HH _ toa, and the vertical emission vertical receiving backscattering section data of the scatterometer is recorded as scat _ VV _ toa; the sea surface wind direction data is recorded as anc _ wind _ dir, the sea surface temperature is recorded as anc _ surface _ temp, the instrument azimuth angle is recorded as celphi, the sea surface wind speed data is recorded as anc _ wind _ speed, the calm sea surface horizontal polarization brightness temperature data is recorded as rad _ exp _ TbH0, and the calm sea surface vertical polarization brightness temperature data is recorded as rad _ exp _ TbV 0.

Referring to fig. 1, a method for correcting rough sea surface brightness temperature using backscattering cross section according to the present invention comprises the following steps:

the method comprises the following steps: reading and preprocessing scatterometry data;

extracting the backscattering cross section data of the horizontal emission level receiving scatterometer from the L2 level data of the Aquarius satellite, and recording the backscattering cross section data as scat _ HH _ toa; then, reducing the scat _ HH _ toa by adopting a logarithmic reduction model to obtain reduced horizontal emission level receiving backscattering cross section data R sigma _ HH;

extracting scatterometer vertical transmitting and vertical receiving backscattering section data from Aquarius satellite L2 level data and recording the backscattering section data as scat _ VV _ toa; then, reducing the scat _ VV _ toa by adopting a logarithmic reduction model to obtain R sigma _ VV of the reduced vertical transmitting and vertical receiving back scattering cross section data;

in the invention, as scat _ HH _ toa and scat _ VV _ toa exist in a decibel form, a logarithmic reduction model is adopted for reduction to obtain an observed quantity R sigma of backscattering section data;

in the present invention, scat _ HH _ toa and scat _ VV _ toa are expressed in a collective form as σ ═ { scat _ HH _ toa, scat _ VV _ toa }.

In the present invention, the log reduction model is:

r σ represents the reduced backscatter cross-sectional data (i.e., observed quantity), which can be denoted as R σ _ HH for scat _ HH _ toa and R σ _ VV _ toa for scat _ VV _ toa; σ represents backscattering cross-sectional data; "Λ" represents the sign of the power.

And reducing the scat _ HH _ toa by adopting a logarithmic reduction model to obtain reduced horizontal emission level receiving backscattering cross section data R sigma _ HH.

And reducing the scat _ VV _ toa by adopting a logarithmic reduction model to obtain reduced vertical transmitting and receiving backscattering cross section data R sigma _ VV.

Step two: acquiring wind direction data and adjusting a wind direction angle;

from Aquarius satellites at level L2Extracting sea surface wind direction data from the data, and recording the sea surface wind direction data as anc _ wind _ dir and an instrument azimuth angle as celphi; then, the wind direction angle observed relative to the satellite is calculated by utilizing a wind direction correction model

The wind direction correction model is as follows:

in the present invention, if

Less than 0 deg. and then through adding 360 deg. to regulate to

Within the range.

In the present invention, if

If greater than 360 DEG, the angle is adjusted to

Within the range.

In the invention, the acquisition anc _ wind _ dir and celphi refers to the corrected brightness temperature data at the same time and at the same longitude and latitude.

Step three: calculating the sea surface emissivity increment by using a scattering emissivity model;

according to the invention, according to the load working characteristics of the Aquarius satellite, aiming at the advantage that the active and passive parameters can be simultaneously obtained, the characteristics that the scatterometer is sensitive to the sea surface roughness only and has higher sensitivity are utilized, and the R sigma and the scattering emission which are actively obtained by the scatterometer are usedObtaining sea surface emissivity increment ew by calculating through a rate model_NRCS. In the traditional seawater salinity inversion algorithm, a wind speed inversion algorithm is required to be introduced for obtaining the sea surface emissivity increment, the wind speed is obtained by inversion, and then the sea surface emissivity increment is obtained by calculation, so that the error caused by a wind speed inversion model is introduced, and the calculation time of seawater salinity inversion is increased. Therefore, the method has strong business application significance for directly calculating the sea surface emissivity increment by using the parameters obtained by satellite observation.

Calculating the sea surface emissivity increment by using a scattering emissivity model as follows:

ew_NRCSindicating the emissivity increase.

beam represents the angle of incidence; in the present invention, the first incident angle may be denoted as beam1, the second incident angle may be denoted as beam2, and the third incident angle may be denoted as beam 3.

p represents the polarization mode of the radiometer; in the present invention, it may be horizontally polarized (denoted as H) or vertically polarized (denoted as V).

R σ represents the reduced backscatter cross-section data;

representing the wind direction angle relative to the satellite observation;

A_n(beam, p, R σ) represents a cosine coefficient function, and the subscript n represents a coefficient identification number (in the present invention, n takes a value of 0, 1, 2, 4). Wherein A is₀(beam, p, R σ) represents a cosine coefficient function of order 0; a. the₁(beam, p, R σ) represents a first order cosine coefficient function; a. the₂(beam, p, R σ) represents a second order cosine coefficient function; a. the₄(beam, p, R σ) represents a fourth order cosine coefficient function.

Because A is_nThe calculation of (beam, p, R σ) is:

a_n,iand the coefficient of a cosine emissivity model is represented, the lower subscript n represents a coefficient identification number, and the lower subscript i represents an order identification number. In the present invention, through a_n,iThe training data are fitted and tables about the angle of incidence beam and the polarization mode are made for lookup, i.e., table 1, table 2 and table 3.

Step four: obtaining sea surface temperature data;

and extracting corrected sea surface temperature of the brightness temperature data at the same time and the same longitude and latitude from the L2 level data of the Aquarius satellite, and recording the sea surface temperature as anc _ surface _ temp.

In the invention, the anc _ surface _ temp is cited to construct the relation between the sea surface emissivity increment and the sea surface bright temperature increment.

Step five: performing sea surface brightness temperature correction;

in the invention, in order to obtain the calm sea surface brightness temperature which can be directly used for salinity inversion, sea surface brightness temperature increment is obtained by calculating the sea surface emissivity increment and the sea surface temperature data in the third step and the fourth step, and then the sea surface brightness temperature increment part caused by rough sea surface is removed from the total sea surface brightness temperature data obtained by the Aquarius satellite. When the factor influencing the inversion accuracy of the seawater salinity, namely the brightness temperature increment of the rough sea surface, is eliminated, the characteristic that the Aquarius satellite is used for simultaneously obtaining the active and passive parameters is used, and the model error caused by sea surface wind speed inversion performed by the traditional method is reduced, so that the precision of the brightness temperature of the calm sea surface obtained by calculation is higher, and the inversion accuracy of the seawater salinity is improved.

The horizontally polarized sea surface brightness temperature for removing the influence of the rough sea surface is as follows:

TB_flat,H＝rad_TbH-ew_NRCS×anc_surface_temp

TB_flat,Hand (4) calculating the horizontal polarization calm sea surface brightness temperature.

The vertical polarization sea surface brightness temperature for removing the influence of the rough sea surface is as follows:

TB_flat,V＝rad_TbV-ew_NRCS×anc_surface_temp

TB_flat,Vand calculating the obtained vertical polarization calm sea surface brightness temperature.