INTRODUCTION
The main goal of satellite altimetry is the study and observation of the processes and properties of the marine environment, so that when utilizing altimetry data the monitoring of phenomena like the mean sea level variations and changes, the ice transfer, the wind speed, the wave height and the water temperature would be feasible. It is possible, through various techniques, to derive information about the marine gravity field and the ocean tides as well (Emadi et al., 2003).
Altimetric measurements have a vital importance for geodesy, since we are able to provide a direct measurement of the main estimation quantity of geodesy i.e., geoid heights. Since the altimetric observations, the Sea Surface Heights (SSHs), correspond to the separation of the sea surface from the reference ellipsoid is very close to geoid undulations. The difference between Mean sea surface height and geoid is called Sea Surface Topography (SST). The deflections of the vertical are calculated directly by differentiating the sea surface height observations in the along track direction, or by differentiating the geoid undulations in the regular grid and then the freeair gravity anomalies, the difference between the magnitude of the actual gravity (W_{0}) on the geoid and the magnitude of the normal gravity (γ_{0}) on the ellipsoid’ were computed using the inverse VeningMeinesz formula.
2D of deflections of vertical components and gravity anomaly with satellite altimetry data: Deflection of vertical is defined as the spatial angel between the normal gravity vector on the reference ellipsoid and the actual gravity vector on the geoid (θ).
It can also be interpreted as the maximum slope of the geoid with respect to the reference ellipsoid at the point of interest.
By using gradients of the sea surface height many of the long wavelength altimetry
error sources are limited. An obvious advantage is that the task of having to
perform a crossover reduction can be avoided. The deflections of the vertical
are calculated directly by differentiating the sea surface height observations
in the along track direction and then computing the deflections of the vertical
(ξ, η) from equation (1)At the crossover location where ascending
and descending ground tracks intersects from either the same of different satellites
a much more stable determination of the slopes can be obtained.
Vertical deflections from alongtrack slopes: Taking the first horizontal derivatives of the altimetersensed sea surface heights alongtrack yields the negative deflections of the vertical at the geoid. These are interpolated onto a regular grid and can be converted to gravity anomalies using the inverse VeningMeinesz formula.
The derivative of the geoid height N with respect to time t along the ascending profile is (Smith and Sandwell, 1994):
and along the descending profile is:
Where
n λ are the geodetic latitude and geodetic longitude of data points, respectively. At the crossover point the following relationships are accurate to better than 0.1%.
The geoid gradient (deflection of the vertical) is obtained by solving (1) using (4).Then we can write:
When two or more satellites with different orbital inclinations are available, the situation is slightly more complex but more stable.
Gridding the deflection of vertical components: The computed the deflections
of vertical components were discrete values, which were randomly distributed
over Arabian sea and that the distances between them were not uniform. As required
by FFT techniques that the input data should be uniformly spaced on a regular
grid (Dadzie, 2005).
We used Shepard’s method of gridding procedure to interpolate values of deflection of vertical components at locations where no data existed, using a grid spacing of 2.5’ arcminute in both longitude and latitude directions. the computational steps as follow:
Knowing the coordinates of the discrete crossover point positions (φ_{i}, λ_{i}) in a spherical coordinate system(φ, λ ) and the corresponding values of the components of the deflection of vertical, f_{i} = f (φ_{i}, λ_{i}) (I = 1, 2,..., N) the differences in longitude (Δλ_{i}) and latitude (Δφ_{i}) between the coordinates of each of the discrete crossover points (φ_{i},λ_{i}) within a gird cell and the coordinates (φ_{0}, λ_{0}) of the grid node ( i.e., the centre point of the grid cell where Z value is to be interpolated ) are computed;
Where r_{i} is the distance, between the grid node, termed the computation point P and the ith discrete crossover point within the cell, termed the running point Q, is computed by :
where R is the mean radius of the Earth, ψ, is the spherical distance between P and Q and
the weight function can be calculated as (Jiang, 2001) :
where r is the search radius and S is the spherical cap, which was set at 2°. The interpolator was finally given by:
where N is the number of data points used in fitting the interpolated value at the grid node and μ is the smoothing factor, which was set at 2 for this study.
Computation gravity anomaly using inverse veningmeinesz formula: The concepts of determining marine gravity anomalies from satellite radar altimetry are as follows. The altimeter essentially measures the distance between the satellite and the instantaneous sea surface along the nadir using pulselimited radar at a series of footprints along the subsatellite tracks (Fu and Cazenave, 2001).
The Inverse Vening Meinesz formula used in this study to compute gravity anomalies using altimeterderived components of deflection of vertical (Eqs. 15) as input data and is given as (Cheng et al., 2001; Hwang, 1998):
where the component of deflection of vertical along the azimuth α is related
to the geoid slope ( )
where R is the mean radius of the earth ξ and are η the altimeterderived northsouth and eastwest components of the deflection of the vertical, respectively.
substituting Eq. (13) into Eq. (12)
yields
considering the Eq. (14) and (15)
Eq. (17) is written as:
where
and
so
the 1D convolution of Eq. (17) is expressed
as:
The corresponding spectral expressions for 2D spherical convolution and 2DFFT
of Eq. (17) given below:
where
where and represent the 2DFFT operator and its inverse. With Eq.
(28) the gravity anomalies at all gridded points are
computed simultaneously.
RESULTS AND ANALYSES
The differences between the altimeterderived gravity anomalies (Fig.
1 and 2) and the EGM96derived gravity anomalies are equal
to the residual gravity anomalies computed utilizing the inverse Vening Meinisz
formula and they represent the shortwavelength component of the gravity anomaly
signal.
Table 1: 
Statistics
of the differences between EGM96generated and altimeterderived gravity
anomalies via 1DFFT over the Arabian Sea (unit, mgal) 

Table 2: 
Statistics
of the differences between EGM96generated and altimeterderived gravity
anomalies via 2DFFT over the Arabian Sea (unit, mgal) 


Fig. 1: 
2.5'x2.5'Map
of altimeterderived gravity anomalies 

Fig. 2: 
2.5'x2.5'
3D map of altimeterderived gravity anomalies 
Table 3: 
Statistics
of the deflection of vertical over the Arabian Sea via 1DFFT(unit, mgal) 

Table 4: 
Statistics
of the deflection of vertical over the Arabian Sea via 2DFFT(unit, mgal) 

Table 1 and 2 show the descriptive
statistics of the differences between the ltimeterderived gravity anomalies
and the EGM96 derived gravity anomalies for 1D and 2Dspherical FFT, respectively, Table 3 and 4 show the statistics of the
gravity anomalies over the Arabian Sea computed from deflections of vertical
via 1D and 2Dspherical FFT, respectively.
CONCLUSIONS
This study describes the procedure and accuracy for marine gravity anomalies over Arabian Sea from multisatellite altimetry. for the present research, it could concluded that the influence of the Dynamic Ocean Topography on the deflection of vertical should be taken into account in order to improve the accuracy of the determination of gravity anomalies. The gridded residual vertical deflections over unobserved area can be supplied by the corresponding model value when the inverse VeningMeinesz formula is used to derive the marine gravity anomalies.