MEAN SEA SURFACE (MSS) MODEL DETERMINATION FOR MALAYSIAN SEAS USING MULTI-MISSION SATELLITE ALTIMETER

The advancement of satellite altimeter technology has generated many evolutions to oceanographic and geophysical studies. A multimission satellite altimeter consists with TOPEX, Jason-1 and Jason-2, ERS-2, Envisat-1, CryoSat-2 and Saral are extracted in this study and has been processed using Radar Altimeter Database System (RADS) for the period of January 2005 to December 2015 to produce the sea surface height (hereinafter referred to SSH). The monthly climatology data from SSH is generated and averaged to understand the variation of SSH during monsoon season. Then, SSH data are required to determine the localised and new mean sea surface (MSS). The differences between Localised MSS and DTU13 MSS Global Model is plotted with root mean square error value is 2.217 metres. The localised MSS is important towards several applications for instance, as a reference for sea level variation, bathymetry prediction and derivation of mean dynamic topography. 1. RESEARCH BACKGROUND


Introduction
Over the past years, the progression of satellite altimeter has been tremendously improved.Previously, the study of ocean variability is limited to certain ocean area and only came from tidal gauges, buoys and ships.With the presence of satellite altimeter, information such as the height of the oceans over nearly 60% of the earth's surface are able to be obtained.Majumdar and Bhattacharyya (2005) stated that satellite altimeter measures the instantaneous sea surface height at the nadir with good precision.A pulse-limited radar is used to measure the altitude of the satellite above the closest sea surface point which, to a first approximation (is a measure of) geoid height or gravitational potential as shown in Figure 1.The difference between the satellite's height above the ellipsoid and above the Earth's surface gives the height of the topography relative to the reference ellipsoid.The precision over land areas is poorer than oceans, but over smooth land features like desserts and inland water bodies, an accuracy of better than a meter is achievable (Lowrie, 2007).
According to Peacook and Laxon (2004), the satellite altimeter has been used in the measurement of the time invariant component of sea surface height or mean sea surface, on a global basis.Furthermore, Jin et al. (2016) stated the satellite altimetry has greatly improved the spatial and temporal resolution of sea surface height (SSH) observations since the first satellite Geo-3 is launched in the 1970s.A series of global and regional mean sea surface (MSS) models has been established, however, only two institutions are still publishing MSS models (the Centre National d'Etudes Spatiales (CNES) and the Space Research Centre of the Technical University of Denmark (DTU)).The Mean Sea Surface (MSS) is an essential and important product of satellite altimeter for supporting to oceanographic and geophysical studies (Hwang et al., 2002).
It is the time-averaged physical height of the ocean's surface.In principle, a complete separation of the oceans means and a variable part requires uninterrupted infinite sampling in both time and space (Andersen and Knudsen, 2009).For some applications in physical oceanography, the MSS is generally used as a reference field for altimeter missions (Schaeffer et al., 2012).The determination of the MSS for Malaysian Seas integrates several applications namely: i.
Global tide Modeling ii.
Sea Level Rise study iv.
Vertical Datum in the field of ocean engineering v.
Derivation of Mean Dynamic Topography Figure 1.Principle of satellite altimetry (Fu and Cazenave, 2001) This study aims to determine the MSS for Malaysian Seas.First the data and the methodology used are described, in particular the method used to process the SSH.This step is crucial for processing the altimetry data because it is intended for defining the average content of the SSH that will serve later as main data.Other parts of this study are also dedicated to mapping the monthly climatological data from the sea surface height data derivation for certain years.The mapping of SSH variability during Malaysia's seasonal monsoons are presented.Finally, the comparison of the new localised MSS model with global model the quality of this new MSS by various validations is defined.

Study Area
In order to meet the requirements of the study, the selected study area is chosen includes four Malaysian Seas; the Straits of Malacca in Peninsular Malaysia, South China Sea and Sabah seas (the Celebes Sea and Sulu Sea).The geographical data boundary used for all data extraction is between 0 ° ≤ Latitude ≤ 14°N and 95°E ≤ Longitude ≤ 126°E.

Sea Surface Height (SSH) Computation from Satellite Altimeter
Satellite altimeter transmits microwave radiation to the sea surface and the signal is reflected back to the altimeter (Andersen and Scharroo, 2011;Fu et. al., 1988).The range, is measured from the travel time taken by the signal to propagate from the transmitter to the sea surface and back to the receiver on-board.The instantaneous height of the sea surface, at particular time is relative to a reference ellipsoid while the altitude, of the altimeter above the corresponding reference ellipsoid is given by an independent tracking system (Din and Omar, 2009).
The sea surface height, can in its simplest form be described according to the following expression, instantaneous sea surface height, (1) where = the instantaneous height of sea surface height (SSH) above the reference ellipsoid = the altitude of the altimeter above the corresponding reference ellipsoid = the travel time taken from the transmitter to the sea surface and back to the receiver on-board However, the satellite altimeter measures the range to the sea surface and the data initially corrected for various range corrections.These range corrections are performed to model the behaviour of the speed of the radar through the atmosphere and the interaction with the sea surface through the sea state correction (Andersen, 2012).Fu and Cazenave (2001) stated that the height of the spacecraft is determined relative to the reference ellipsoid through Precise Orbit Determination and Global Positioning System (GPS) on-board.Combining the knowledge of the height of the spacecraft with the corrected range gives the sea surface height relative to the reference ellipsoid.For this study, all atmospheric, instrument and geophysical corrections is included (see Table 1) to obtain SSH; (2)   11 years (2005-2015) satellite altimeter (ERS-2, Envisat-1, CryoSat-2 and Saral) data have been used to obtain the sea surface height for Malaysian Seas.While the TOPEX-class satellites (TOPEX, Jason-1 and Jason-2) fix as a reference to the other-class satellites and are adjusted simultaneously throughout the processing.Table 2 summaries the details of the satellite altimeter and its status.

Table 2. Satellite Altimetry Data Processed in RADS
In the RADS, certain shell script files are created and used in accessing and analysing desired results that suit to the requirement of this study.Firstly, a file called "getraw.nml" is created to enable the user to define the geographical region of interest in latitude and longitude, including other parameters.Then, the crossover adjustment is performed by applying altimeter corrections and removing bias in RADS processing.Schrama (1989) suggested since the TOPEX and JASON satellites can give better accuracy compares to other satellites, the sea surface heights (SSH) from different satellite missions need to be adjusted to a "standard" surface.
The time frame covered by the individual crossovers is pegged at 18 days to increase the likelihood of capturing all available real oceanic signals and sea level trends (Trisirisatayawong et al., 2011).A distance-weighted gridding is applied to obtain useful values for points situated between footprints.The Gaussian distribution, is relied on for the weighting function (Singh et al., 2004): (3) where sigma, σ = weight given to a value at a normal point with location of a distance from the grid point The sigma 1.5 weighting function is used to filter and grid the data to sea surface height bins using the Gaussian weighting function.A square mesh with block size of 0.25° (spatial) and an 9 days cut-off (temporal) is factored into the gridding step that used for spatial and temporal weighting on SSH anomaly.

Mean Sea Surface (MSS) Model Determination
In order to generate localised Mean Sea Surface (MSS) Model for Malaysian seas, the climatology data of the SSH for 11 years has been spatially averaged.However the climatology data from SSH has been subtracted from Localised Mean Sea Surface (new model) to provide the anomaly as mapping shown because of the changes from SSH is slightly small and does not appear to be significant when it been plotted.Figure 4 shows a SSH anomaly for the month of January -June (Figure 5 for July -December), from a new SSH climatology derived exclusively from the 11-year dataset.
It can be seen that there are some areas of the SSH anomaly where the quantity or quality of data are varies for each month.Figure 6 and Figure 7 illustrated that variation of SSH anomaly during the 1 st and 2 nd Inter monsoon, which, are appeared to be calm and not much increase in the height of the sea surface.

Localised MSS Model for Malaysian Seas
The mean sea surface (MSS) relates to the sum of mean geoid and mean ocean dynamic topography.However, for this study, the MSS is geometrically obtained from the difference between the SSH and the sea level anomaly.Figure 10 presents the new MSS model derived from satellite altimeter data acquired from TOPEX, Jasons-1&2, ERS-2, Envisat, Cryosat-2 and Saral missions covering the Malaysian Seas.
From Figure 10, it can be comprehended that the mean sea surface increases eastwards from West Malaysia (Peninsular) towards West Malaysia (Sabah and Sarawak).Grant Scheme (GUP) Vot No. 12H99 given to complete this research project.

Comparison of New
where = the instantaneous height of sea surface height (SSH) above the reference ellipsoid = the travel time taken from the transmitter to the sea surface and back to the receiver onboard = the altitude of the altimeter above the corresponding reference ellipsoid = ionospheric correction = dry tropospheric correction (Oxygen and Nitrogen) = wet tropospheric correction (water vapour) = sea state bias correction = inverse barometric effect = load tide effect = solid earth tide effect = pole tide effect = ocean tide effect Atmospheric Corrections Model/Sensor Flows: Anderson (2012) highlighted that altimetry data are accessible and well-distributed through agencies, like the European Organisation for the Exploitation of Meteorological Satellites (EUMETSAT), AVISO, the Physical Oceanography Distributed Active Archive Center (PO.DAAC) and the National Oceanic and Atmospheric Administration (NOAA).In addition to these operational data centres, the Radar Altimeter Database System (RADS) is a joint effort by the Department for Earth Observation and Space Systems (DEOS), Delft University of Technology and the NOAA Laboratory for Satellite Altimetry in establishing a harmonised, validated and cross-calibrated sea level database from all altimeter missions.

Figure 6 .
Figure 6.The Sea Surface Height Variation during 1 st Inter Monsoon

Figure 7 .
Figure 7.The Sea Surface Height Variation during 2 nd Inter Monsoon On the other hand, the Southwest Monsoon variation in Figure 8 reveals a slightly increasing in the height of the sea surface at the Straits of Malacca area since the monsoon season produced an average rainfall throughout the 4 months (Sang et al., 2015a).

Figure 8 .
Figure 8.The Sea Surface Height Variation during Southwest Monsoon However, the pattern intensified during the Northeast Monsoon where the variation of sea surface height at its most highest among other season at South China Sea (Figure 9).It is due to the characteristics of the Northeast Monsoon which brings heavy rainfall, particularly in the east coast states of Peninsular Malaysia and western Sarawak (Sang et al., 2015b).

Figure 9 .
Figure 9.The Sea Surface Height Variation during Northeast Monsoon Localised MSS Model with DTU13 MSS Global Model Data from DTU13 Global MSS Model is utilised to evaluate the constructed new and localised MSS Model.For this MSS Model for Malaysian Seas, the difference between mean sea surface heights (local) and global model (DTU13) have to be computed.Figure 11 signified the difference between Localised MSS Model and DTU13 MSS Global Model.Meanwhile,Table 3 represents a statistical analysis of the difference between DTU13 Global MSS Model and Localised MSS Model .-13.426 7.443 -0.30004 -2137.82.217 Table 3. Statistic of the difference between DTU13 Global MSS Model and Localised MSS Model In this paper, multi-mission satellite altimeter observations are combined to produce the monthly climatological SSH data.From the SSH data, the variation of sea surface is generated and mapped during the monsoon season.Moreover, the SSH data are required to establish a localised mean sea surface for Malaysian Seas using the mean of gridded SSH spanning the 11 years data, between 2005 and 2015.The MSS is an important reference for sea level variation.Besides, MSS is capable to integrate other oceanographic and geophysical studies, such as for global tide modeling, bathymetry prediction and vertical datum determination in the field of ocean engineering.

Figure 10 .
Figure 10.Localise Mean Sea Surface for Malaysian Sea

Table 1 . List of corrections applied 2.2.1 Satellite Missions: In this study,
IonosphereSmoothed dual-frequency valueDry TroposphereEuropean Centre for Medium-Range Weather Forecasts (ECMWF) Wet Troposphere RadiometerSignal-Surface InteractionSea State Bias CLS non-parametric model