Sea surface salinity inversion algorithm based on wind and cloud meteorological satellite
Technical Field
The invention relates to the technical field of sea surface salinity observation, in particular to a sea surface salinity inversion algorithm based on a wind cloud meteorological satellite.
Background
Sea surface salinity is an important physical and chemical parameter of the ocean and is one of important parameters for determining the basic properties of the seawater. Sea surface salinity is not only an essential environmental variable in the exploration of ocean phenomena such as hot salt circulation, global sea level change and the like, but also provides a forced reference for water mass analysis, climate research, global ocean pattern research and the like. The monitoring of sea surface salinity is also widely applied in various fields such as petroleum, import and export trade, military channel measurement, ocean fishery and the like. However, the sea surface salinity measured by the traditional ship navigation and drifting buoy is discrete in time and space, and the change of the sea surface salinity in space and time distribution cannot be accurately detected.
The medium-resolution spectral imaging instrument (MERSI-2) carried on the second generation wind cloud polar orbit meteorological FY-3D satellite in China integrates the functions of two imaging instruments (MERSI-1 and VIRR) of the original wind cloud three-satellite, is the first imaging instrument in the world capable of acquiring information of a global 250-meter resolution infrared split window area, can acquire a global 250-meter resolution true color image seamlessly every day, realizes high-precision quantitative inversion of atmospheric, land and ocean parameters such as cloud, aerosol, water vapor, land surface characteristics, ocean water color and the like, and forms observation of sea surface high space-time resolution by utilizing FY-3D/MERSI high space-time resolution (250m, scanning a target area once every day).
Artificial Intelligence (AI) is a science, machine learning is the most mainstream artificial intelligence implementation method at present, and deep learning is a branch of machine learning. Deep learning is a method based on data characterization learning in machine learning, is proposed in 2006 by Hinton et al, is a machine learning method capable of simulating a neural structure of a human brain, and brings hopes for solving optimization problems related to deep structures. The invention develops a sea surface salinity inversion method for a resolution ratio spectral imager (FY-3D/MERSI) in a second generation wind cloud polar orbit meteorological satellite in China by utilizing a deep learning technology.
Disclosure of Invention
The invention aims to provide a sea surface salinity inversion algorithm based on a wind cloud meteorological satellite so as to solve the problems in the background technology.
In order to achieve the purpose, the invention provides the following technical scheme: a sea surface salinity inversion algorithm based on a wind cloud meteorological satellite comprises the following steps:
s1 training argument (X): the method is characterized in that the number of hidden layers is increased according to regression task requirements, a model for carrying out sea surface salinity satellite inversion algorithm on a deep neural network is built, 4 independent variables including remote sensing reflectivity (Rrs), aCDOM, Sea Surface Temperature (SST) and suspended matter concentration (TSM) are observed for 4-waveband satellites, and the 4 independent variables are obtained in the following steps:
s11, satellite observation remote sensing reflectivity (Rrs);
s12, performing aCDOM inversion;
s13 SST inversion;
s14, inversion of TSM;
s2 measurement of sea surface salinity (Y) inversion of a at FY-3D/MERSICDOMA is established through a deep learning model on the basis of Sea Surface Temperature (SST) and suspended matter concentration (TSM)CDOMAnd the relation between SST, TSM and sea surface salinity realizes the global sea surface salinity observation by utilizing the global observation capability of FY-3D/MERSI.
Preferably, the S11 satellite observation remote sensing reflectivity (Rrs): the reflectivity can be directly obtained by an FY-3D/MERSI satellite after atmospheric correction.
Preferably, the S12: aCDOM inversion: in the process of obtaining the aCDOM by using remote sensing data, firstly, an empirical formula is established by using the total absorption anw (443) of non-water at 443nm, the backscattering coefficient bbp (555) at 555nm and remote sensing reflectivity parameters to deduce the absorption coefficient ad (443) of the non-phytoplankton particles:
ad(443)=0.60×σ0.90
wherein the content of the first and second substances,
anw (443) and bbp (555) consist ofThe QAA algorithm.
Secondly, after obtaining ad (443), the value of ad in any band can be calculated by using the correlation between the bands of ad:
after obtaining ad, the total absorption value aphg of CDOM and phytoplankton was determined using the total absorption of nonaqueous anw:
aphg(λ)=anw(λ)-ad(λ)
finally, the absorption coefficient aCDOM (443) of the yellow substance at 443nm is separated by aph at 412nm, 443nm and 490 nm:
aCDOM(443)=aphg(443)/(1+9.56×104×e-11.13×ψ)。
preferably, in S12:
λ
1=412,λ
0=443,λ
2=490。
preferably, the S13 SST inversion: sea surface temperature is an important climate and weather parameter, and has a great influence on sea surface salinity distribution. The infrared band of the satellite sensor can be used for inverting the sea surface temperature, and a multivariate nonlinear inversion algorithm based on atmospheric correction is adopted for inverting the SST:
SST=a+bT11+c(T11-T12)Tsfc+d(T11-T12)(sec(θ)-1)
wherein, T11And T12Respectively the brightness temperature values of 11um and 12um wave bands directly observed by the satellite, theta is the zenith angle observed by the satellite, and T12For the SST inception field (the united states national environmental prediction center (NCEP) was used in this patent study to re-analyze the sea surface temperature data set). a, b, c and d are constants, and the transmission of infrared radiation in the atmosphere can be simulated based on a numerical mode to obtain corresponding values.
Preferably, the S14 TSM inversion: and (3) inverting the concentration of the suspended matters in the water body by utilizing an algorithm for inverting the concentration of the suspended matters in the water body based on the Intrinsic Optical Properties (IOP) of the water body.
TSM=1.73*bp(443)
bp(443)=bbp(443)/0.015。
Preferably, the S14 is B in TSM inversionp(443) And bbp(443) Scattering coefficient and backscattering coefficient of particulate matter, respectively, bp(443) And bbp(443) Can be calculated by QAA (Quasi-analytical algorithm) semi-analytical algorithm (see Table 1 for algorithm flow).
Compared with the prior art, the invention has the beneficial effects that:
1. the sea surface salinity data coverage rate is low due to the limitations of low space-time resolution (30-300km, a scanning period of more than 3 days) and data quality of SMOS and Aquarius/SAC-D satellites, so that the sea surface salinity cannot be observed with high space-time resolution;
2. the invention solves the problems that sea surface salinity is inverted by SMOS and Aquarius satellites by utilizing the relation among sea surface temperature, sea water dielectric constant and sea surface salinity detected by microwave bands, the inversion accuracy of the sea surface salinity is influenced by factors such as atmosphere and background radiation in the transmission process, sea surface radiation brightness temperature gain caused by sea surface roughness, influence of white canopy on sea surface brightness temperature and the like, and the requirement of high-space-time resolution sea surface salinity monitoring is difficult to form due to the limitation of low space-time resolution (30-300km, scanning period of more than 3 days) of microwave sensor data.
Drawings
FIG. 1 is an overall flow chart of a sea surface salinity inversion algorithm based on a wind cloud meteorological satellite according to the present invention;
FIG. 2 is a diagram of a sea surface salinity inversion algorithm neural network structure based on a wind cloud meteorological satellite according to the present invention;
FIG. 3 shows sea surface salinity inversion algorithm FY-3D/MERSI inversion a based on wind cloud meteorological satelliteCDOM(443) A drawing;
FIG. 4 is a sea temperature map inverted by sea surface salinity inversion algorithm FY-3D/MERSI based on a wind cloud meteorological satellite according to the invention;
FIG. 5 is a diagram of sea surface salinity inverse algorithm FY-3D/MER inversion TSM based on a wind cloud meteorological satellite according to the invention;
FIG. 6 is a diagram of a sea surface salinity inversion algorithm FY-3D/MERSI global salinity inversion product based on a wind cloud meteorological satellite.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Referring to fig. 1-6, the present invention provides a technical solution: a sea surface salinity inversion algorithm based on a wind cloud meteorological satellite comprises the following steps:
s1 training argument (X): the method is characterized in that the number of hidden layers is increased according to regression task requirements, a model for carrying out sea surface salinity satellite inversion algorithm on a deep neural network is built, 4 independent variables including remote sensing reflectivity (Rrs), aCDOM, Sea Surface Temperature (SST) and suspended matter concentration (TSM) are observed for 4-waveband satellites, and the 4 independent variables specifically obtain the following steps:
s11 satellite observed remote sensing reflectance (Rrs): the reflectivity can be directly obtained by an FY-3D/MERSI satellite observation after atmospheric correction;
s12 aCDOM inversion: in the process of obtaining the aCDOM by using remote sensing data, firstly, an empirical formula is established by using the total absorption anw (443) of non-water at 443nm, the backscattering coefficient bbp (555) at 555nm and remote sensing reflectivity parameters to deduce the absorption coefficient ad (443) of the non-phytoplankton particles:
ad(443)=0.60×σ0.90
wherein the content of the first and second substances,
anw (443) and bbp (555) are derived by the QAA algorithm.
Secondly, after obtaining ad (443), the value of ad in any band can be calculated by using the correlation between the bands of ad:
after obtaining ad, the total absorption value aphg of CDOM and phytoplankton was determined using the total absorption of nonaqueous anw:
aphg(λ)=anw(λ)-ad(λ)
finally, the absorption coefficient aCDOM (443) of the yellow substance at 443nm is separated by aph at 412nm, 443nm and 490 nm:
aCDOM(443)=aphg(443)/(1+9.56×104×e-11.13×ψ)
wherein the content of the first and second substances,
λ
1=412,λ
0=443,λ
2=490;
s13 SST inversion: sea surface temperature is an important climate and weather parameter, and has a great influence on sea surface salinity distribution. The infrared band of the satellite sensor can be used for inverting the sea surface temperature, and a multivariate nonlinear inversion algorithm based on atmospheric correction is adopted for inverting the SST:
SST=a+bT11+c(T11-T12)Tsfc+d(T11-T12)(sec(θ)-1)
wherein, T11And T12Respectively the brightness temperature values of 11um and 12um wave bands directly observed by the satellite, theta is the zenith angle observed by the satellite, and T12For the SST inception field (the united states national environmental prediction center (NCEP) was used in this patent study to re-analyze the sea surface temperature data set). a, b, c and d are constants which can be based on numerical moduliSimulating the transmission of infrared radiation in the atmosphere to obtain a corresponding value;
s14 TSM inversion: and (3) inverting the concentration of the suspended matters in the water body by utilizing an algorithm for inverting the concentration of the suspended matters in the water body based on the Intrinsic Optical Properties (IOP) of the water body.
TSM=1.73*bp(443)
bp(443)=bbp(443)/0.015
Wherein b isp(443) And bbp(443) Scattering coefficient and backscattering coefficient of particulate matter, respectively, bp(443) And bbp(443) Can be calculated by QAA (Quasi-analytical algorithm) semi-analytical algorithm (see Table 1 for algorithm flow).
TABLE 1 QAA Algorithm calculation absorption coefficient and backscattering coefficient step
Wherein lambda is spectrum wavelength, rrs (lambda) is remote sensing emissivity of lambda waveband, chi, mu (lambda) and η are intermediate variables, g0 and g1 are model constants, aw(lambda) is the absorption coefficient of pure water in lambda wave band, a (lambda) is the total absorption coefficient of water body, bb(lambda) and bbp(lambda) is the water backscattering coefficient and the particulate backscattering coefficient respectively;
s2 measurement of sea surface salinity (Y) inversion of a at FY-3D/MERSICDOMA is established through a deep learning model on the basis of Sea Surface Temperature (SST) and suspended matter concentration (TSM)CDOMAnd the relation between SST, TSM and sea surface salinity realizes the global sea surface salinity observation by utilizing the global observation capability of FY-3D/MERSI.
It is noted that, herein, relational terms such as first and second, and the like may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus.
Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that changes, modifications, substitutions and alterations can be made in these embodiments without departing from the principles and spirit of the invention, the scope of which is defined in the appended claims and their equivalents.