^{1}

^{*}

^{1}

^{2}

^{1}

Inversion of seawater physical parameters (temperature, salinity and density) from seismic data is an important part of Seismic Oceanography, which was raised recent years to study physical oceanography. However present methods have problems that inversion accuracy is not high or inverted parameters are incomprehensive. To overcome these problems, this paper derives Allied Elastic Impedance (AEI), from which we can extract acoustic velocity and density of seawater directly. Furthermore this paper proposes a method to fit temperature and salinity with acoustic velocity and density respectively, breaking through the limitation that temperature and salinity can only be extracted from acoustic velocity. After applying it to model and real data, we find that this method not only solves the problem that ocean density is hard to extract, but also increases accuracy of other parameters, with the temperature and salinity resolution of 0.06°C and 0.02 psu respectively. All results show that AEI is promising in inversion of seawater physical parameters.

As a method to explore geologic structure and hydrocarbon resources by means of wave reflection, multichannel seismic (MCS) has been applied in ocean research for decades. But for a long time, geophysicists are just focusing on the reflections bellow seafloor, while those from seawater are ignored as noises [

Early researches of seismic oceanography were mainly focused on describing boundaries of water mass and thermohaline gradient directly by seismic reflections, and some success have been achieved in the study of marine fronts [

Zoeppritz equations are used to describe relationship between reflectivity of elastic interface and velocity, density and incident angle of adjacent layers. For liquid interface, it becomes to be [

where

where

Here we need a function

when

So

Then we substitute

Finally we integrate and index Equation (8), setting the integration constant to zero.

We can see the format of

However, AEI has an undesirable feature that its dimensionality varies with incidence angle θ and provides numerical values that change significantly with θ (_{0} which is the average value of acoustic velocity from XCTD [

This is the final form for inversion, and we can see that modifying the AEI function doesn’t affect the value of reflectivity through function (4).

The inversion of AEI is similar to that of acoustic impedance, which is also performed with the constraint of XCTD. While the input data is angle stack data, the impedance for constraint is calculated by function (10), and the wavelet is extracted from corresponding angle stack data. Low frequency model plays an important role in the process of inversion.

From function (10), we can know that, in order to extract acoustic velocity and density from AEI, at least two allied elastic impedance volumes are needed. So that they can compose simultaneous equations blew.

Here, we replace

Acoustic velocity and density is calculated at every sample separately, leading to the throb of results. Considering that ocean parameters vary very gently, we smooth the inverted profiles by means of cubical smoothing algorithm with five-point approximation, making them closer to the actual situation.

After getting acoustic velocity and density, we can extract temperature and salinity from them. Traditional method looks for the optimum temperature and salinity iteratively by means of empirical velocity equation (Wilson equation) and temperature-salinity relationship from XCTD. However the relationship between them is always complex and nonmonotonic, accurate relationship is hard to get. According to the conclusion that temperature is sensitive to acoustic velocity and salinity is easy to be affected by density [

Then we invert AEI from angle stack data with the constraint of XCTD data, resolve acoustic velocity and density from AEI, and compute temperature and salinity by the fitting relationships. ^{3}, 0.06˚C and 0.02 psu respectively. Taking their variation ranges into consideration, relative errors of density and salinity are bigger than acoustic velocity and temperature. This is because relative contribution to reflectivity of acoustic velocity and temperature are far greater than density and salinity, and all these results are inverted from reflectivity [

To highlight the variation of temperature and salinity on the profile, we also convert their data maps to contour maps, as shown in

This paper derives AEI from Zoeppritz equations of liquid and resolves acoustic velocity and density from it. Furthermore we propose a method to calculate temperature and salinity. After applying it to real data, we come to following conclusions:

1) Reflections from seawater also have obvious AVO features. AEI derived on this basis can provide more information (incidence angle) compared with acoustic impedance. So we can resolve more parameter (density) from AEI.

2) After attaining precise density, temperature and salinity can be calculated by acoustic velocity and density not only by acoustic velocity, which improves inversion accuracy of all parameters.

3) On the ground that input data are angle stack and incidence angle of water layer just below surface is large,

the shallow reflections are easy to be omitted. On the other hand, shallow reflections are mixed with primary waves, so physical parameters of shallow water must be got by other methods.

The authors would like to thank the Natural Science Foundation of China (41176077, 41230318), Subject of 863 (2013AA092501), Natural Science Foundation of Shandong (ZR2010DM012), Basic Research Special Foundation of the Third Institute of Oceanography affiliated to the State Oceanic Administration (TIOSOA, 2009004) and the Science Research Project for the South China Sea of Ocean University of China for their financial support to this work.

HuaishanLiu,XueqinLiu,JinqiangZhu,JiaWei, (2016) Inversion of Seawater Physical Properties Based on Allied Elastic Impedance. Journal of Water Resource and Protection,08,135-142. doi: 10.4236/jwarp.2016.82011