Project Topics
Get Complete Project Material File(s) Now! »
Multi satellite altimetry record and global mean sea level budget
Satellite altimetry missions have transformed the way we view Earth and its ocean and has been the main tool for continuously and precisely monitoring the sea surface height (SSH) with quasi-global coverage and short revisit time. In Chapter 1, we have seen the evolution of global mean sea level (GMSL) as measured by satellite altimeters since 1993. In this chapter, we first discuss the evolution of various altimetry satellites used for ocean observation, principle behind the altimetry-based SSH measurement and different error corrections involved in precise SSH measurements. We then move on to the estimation of global mean sea level budget since a decade using various observational and in-situ data.
Evolution of altimetry satellites
Satellite altimetry was developed in the 1960s soon after the flight of artificial satellites became a reality (Fu and Cazenave, 2001). The first altimetry satellite, GEOS 3 (Geodynamics Experimental Ocean Satellite) was launched in 1975 and carried instruments to yield useful measurements of sea level and its variability. However, its performance was not good enough to extract useful scientific information from its measurements. This was then followed by Seasat (SEA faring SATellite) and Geosat (GEOdetic SATellite) missions in 1978 and 1985 respectively. While Seasat gave us the first global view of ocean circulation, waves and winds, Geosat was the first mission to provide long-term (over 3 years) of high quality altimetry data. However, until the early 1990s, satellite altimetry has been more useful to marine geophysics rather than oceanography. This is mainly because the orbital error (see section 2.1.2 for details)for these missions was so large (from several decimeters to ~1m) that the altimetry range uncertainty prevented detection and precise measurements of phenomena associated with ocean dynamics such as dynamic topography, tides, sea level etc. (Fu and Cazenave, 2001, Palanisamy et al., 2015a).
The era of precise satellite altimetry dedicated to space oceanography began in the early 1990s with the launch of ERS-1, a European Space Agency (ESA) mission in 1991 and theNational Aeronautics and Space Agency (NASA)/ Centre National d’Etudes Spatiales (CNES)joint venture TOPEX/Poseidon in 1992. In fact, TOPEX/Poseidon revolutionized the study of Earth’s oceans by providing the first continuous, global coverage of ocean surface topography.
Its data made a huge difference in our understanding of the oceans and their effect on global climatic conditions. Its repeat cycle (i.e. the time taken for the satellite to pass vertically over the same location) of ~10 days provided more information than in-situ measurements over hundred years. The mission far exceeded the expectations in terms of both mission duration (initial design life was 5 years but stayed in operation for 13 years) and measurement system performance (Buis et al., 2006).
During the period of TOPEX/Poseidon (1992-2006), several other new altimetry missions were launched: ERS-2 (ESA, 1995-2011), Jason-1 (NASA/CNES, 2001-2012), ENVISAT (ENVIronmental SATellite, ESA, 2002-2012) and Jason-2 (NASA/CNES, 2008- ). The combination of several satellites available at the same time enabled a good compromise between the spatial and temporal resolutions for ocean monitoring. For example, while the repeat cycle of TOPEX/Poseidon was ~10 days (good temporal resolution), its inter-track spacing at the Equator was in the order of 315 km; whereas in the case of ENVISAT, its repeat cycle was 30 days with inter-track spacing in the order of 80 km at the Equator (good spatial resolution).
Moreover, while TOPEX/Poseidon, Jason-1 and Jason-2 flew at the non- sun synchronous orbit (up to 66°N/S), ENVISAT’s polar sun synchronous orbit further enabled high latitude oceancoverage (up to 81.5° N/S) thereby complementing to the other existing missions.
Cryosat-2 and HY-2A are two other satellite altimetry missions launched by ESA and China in 2010 and 2011 respectively. The Cryosat-2 orbit at an inclination of about 92 degree and an altitude of 717 km covers almost all the polar region and henceforth is dedicated to polar observation while HY-2A at sun synchronous and geodetic orbit helps in monitoring the dynamics of the ocean. Since the beginning of 2013, a new altimetry mission, SARAL-AltiKa, a joint venture between CNES and Indian Space Research Organization (ISRO), now ensures the continuity of high precision sea level data along with the existing Jason-2 measures (Verron et al., 2015). It is the first oceanographic mission using a high frequency Ka band altimeter for improved spatial and temporal resolution and flies in the same orbit as that of ENVISAT (a detailed explanation on SARAL-AltiKa is presented in Section 2.1.4). Another mission, Jason-3 in the framework of a co-operation between CNES, NASA, Eumetsat (EUropean METeorology SATellite) and NOAA (National Oceanic and Atmospheric Administration) is proposed to be launched soon. This mission is similar to those of TOPEX/Poseidon, Jason-1 and Jason-2 with the same kind of payload and orbital parameters.
Since satellite altimetry has proven to be a valuable source of data for ocean monitoring and understanding, several future missions are foreseen such as Sentinel-3 Jason-CS, SWOT (Surface Water Ocean Topography). Continuous monitoring with the help of such satellite altimeters will further enhance our understanding of oceans and their mechanisms.
Principle of satellite altimetry
The radar altimeters on board the satellite transmit signals at high frequency (over 1700 pulses per second) towards the sea surface which is partly reflected back to the satellite.
Measurement of the round-trip travel time is then converted to obtain the distance of the satellite above the instantaneous sea surface, called as ‘range’. SSH measurement is deduced from the difference between the satellite distance to the Earth’s centre of mass (deduced from precise orbitography and called ‘satellite/orbit altitude’) and the satellite distance (range) above the sea
surface (Fig.2.1) Besides sea surface height, wave height and wind speed over the oceans can also be measured from the amplitude and waveform of the return signal.
Corrections involved in Sea Surface Height measurement
The determination of precise SSH measurements from satellite altimetry is influenced by many factors. Amongst the most important are sensor characteristics, long-term stability of the altimeter, and the methods used in altimeter data processing (Fernandes et al., 2006). While the sensor characteristics of the altimeter are pre-determined depending on the altimetry missions, instrumental and orbital errors, differences in the methodology for processing altimetry data and the corrections involved are crucial for sea level monitoring. High precision SSH measurements require (1) precise orbit determination of the satellite and (2) application of accurate geophysical, corrections to the raw altimetry signal. The latter accounts for the interactions of altimeter signal with the atmosphere and sea surface. Slight differences in the models used or processing approaches for the above mentioned corrections can seriously impact the SSH estimates and corresponding uncertainties. Therefore, from Section 2.1.1, the precise SSH measurement from satellite altimetry can now be expressed as In this section, we discuss several of these corrections involved for precise SSH estimation.
1) Orbital correction
The quality of the altimetry data depends on the ability to precisely determine a satellite’s position on orbit. Orbital error is caused by imperfect knowledge of the spacecraft position in the radial orbit direction and is the largest error source on altimetry measurements for SSH monitoring. Therefore, a very precise knowledge of the satellite orbit with respect to Earth’s reference ellipsoid is essential. The satellite’s orbit can be also perturbed by gravitational forces related to the non-uniform distribution of Earth’s gravity field, and those caused by the sun, moon and other planets. Forces on the satellite’s surface such as atmospheric drag (e.g. low orbit satellites experience more atmospheric drag) and radiative pressure (e.g. solar radiation, Earth’s IR radiation etc.) also play a role in the perturbation of the satellite’s orbit. This indicates that a detailed knowledge of the satellite and its variations due to maneuvers, fuel consumption, solar panel orientation etc. are also necessary in order to precisely model the above mentioned forces acting on it. Precise orbit determination is done through a combination of satellite tracking and dynamic modelling. Satellite tracking in general involves the use of Satellite Laser Ranging (SLR), Global Positioning System (GPS) and Doppler Orbitography and Radio positioning System (DORIS). A dynamical model taking into account the forces acting on the satellite is then fitted through the tracking data to obtain the precise orbit of the satellite.
2) Propagation corrections
As the radar altimetry signal travels through the atmosphere to the sea surface, it is slowed down due to the presence of various elements in its path. The delay in the propagation of the radar signal thus needs to be corrected in order to estimate the precise range, i.e. the distance between the satellite and sea surface. There are three types of propagation delays that need to be accounted for: (1) ionospheric, (2) wet and (3) dry tropospheric corrections.
a) Ionosphere correction
This correction takes into account the path delay in the radar signal due to the presence of electrons in the ionosphere. The ionosphere is the uppermost layer of the atmosphere (ranging between 60 and 800 km) and the electrons are produced as a result of ionization attributed to solar radiation. As a result, it causes the altimeter to slightly over estimate the range by 1 to 20 cm. This amount can vary with respect to the seasons, solar cycle and occurrences of geomagnetic storms with the maximum influence occurring at the tropical band. Since the range delay due to the presence of electrons is related to electromagnetic radiation frequency, the correction can be estimated using two different radar frequencies (for example, C-band and Ku-band for TOPEX and Jason-1). This correction can also be taken into account from models of the vertically integrated electron density (Fu and Cazenave, 2001, Callahan, 1984, Imel, 1995).
b) Wet troposphere correction
The wet tropospheric correction is the correction for the delay of the radar signal due to the presence of water vapor content in the atmosphere. This is a difficult correction to be accounted for as the wet atmospheric effect highly varies both spatially and temporally with maximum occurring in the tropical convergence zones and magnitudes ranging from 5cm to 30 cm. Over the oceans, the wet tropospheric correction is in general computed using on-board microwave radiometer measurements. But, such radiometric measurements generally fail near the coasts where the signal coming from the surrounding land surface contaminates the radiometer measurements (Desportes et al., 2007). In such cases, the correction is computed from meteorological model outputs such as ECMWF (European Center for Medium Range Weather Forecast) or NCEP (National Centers for Environmental Prediction) models (Legeais et al., 2014).
c) Dry troposphere correction
The mass of dry air molecules in the atmosphere causes a range delay known as the dry tropospheric effect. It is directly proportional to sea level pressure and is the largest adjustment that has to be applied to altimetry measurement as its order of magnitude is about 2.3m. However, its temporal variations are low and range a few centimeters only. The dry tropospheric correction is computed using atmospheric model pressure forecasts such as ECMWF.
3) Geophysical corrections
The gravity forces generated by the Sun and Moon on the Earth can create perturbations on the earth’s interior and also sea surface elevations of few meters. These geophysical effects are called as tidal effects and can be classified into four types: (1) ocean, (2) solid Earth, (3) polar tidal and (4) loading effects.
a) Ocean, solid Earth tidal and loading corrections
Ocean tides and their variations are the result of the combined attraction of Sun and Moon and represent more than 80% of the surface variability in open ocean. The tidal periods (half a day or one day) are shorter than the repeat periods of an altimetry satellite. Tidal correction is therefore essential since they contaminate the low frequency altimetry signals. Furthermore, for the estimation of dynamic sea surface height, the magnitude of ocean tides are very large and must therefore be removed as they are considered as noise. Ocean tidal corrections are performed using assimilated hydrodynamic and statistical models that estimate tides globally with high spatial resolution (Fu and Cazenave, 2001, Ray, 1999).
Multi-mission SSH altimetry data
We carried out global and regional sea level studies by making use of multi-mission merged SSH altimetry data. The purpose of using multi-mission altimetry data is to obtain the most stringent accuracy requirements that are needed for climate research. Moreover, as mentioned in Section 2.1, the combination of several satellites enables a better compromise on spatial and temporal resolution. Merging multiple altimeter data sets is not easy. It requires homogenous, inter-calibrated data sets; correcting of orbit error for the less precise altimeter missions; extracting the sea level anomaly using a common reference surface; and combining the data through a mapping (or assimilation) method (Le Traon et al., 1998). After calculating the along track sea level measures for each of the satellite missions, the main steps consists of: combining all missions together, reducing the orbit and the long wavelength errors, computing the gridded sea level anomalies using an objective analysis approach (Ducet et al., 2000, Le Traon et al., 2003), and generating mean sea level products (e.g. GMSL time series, gridded sea level time series) dedicated for climate studies (Ablain et al., 2015). Owing to the homogenization of the altimetry corrections between all the missions, the uncertainty in multi-mission GMSL trend over 1993-2010 is now reduced to 0.5 mm/yr. The main source of this error remains to be the wet tropospheric correction with a drift uncertainty in the range of 0.2-0.3 mm/yr (Legeais et al., 2014). Orbit error and altimeter parameters error also add an uncertainty in the order of 0.1 mm/yr. Furthermore, imperfect links between various altimetry missions cause a GMSL trend error of about 0.15 mm/yr (Ablain et al., 2015).
Table of contents :
Chapter 1 Introduction
1.1 Paleo sea level
1.2 Instrumental era sea level
1.2.1 Tide gauge records
1.2.2 Satellite altimetry
1.3 Contributors to global mean sea level rise during the instrumental era
1.3.1 Ocean temperature and salinity changes
1.3.2 Glaciers melting
1.3.3 Ice sheets
1.3.4 Land waters
1.4 Thesis objectives
Chapter 2 Multi satellite altimetry record and global mean sea level budget
2.1 Evolution of altimetry satellites
2.1.1 Principle of satellite altimetry
2.1.2 Corrections involved in Sea Surface Height measurement
1) Orbital correction
2) Propagation corrections
a) Ionosphere correction
b) Wet troposphere correction
c) Dry troposphere correction
3) Geophysical corrections
a) Ocean, solid Earth tidal and loading corrections
b) Polar tidal correction
4) Surface corrections
a) Inverse barometric (IB) correction
b) Sea State Bias (SSB) correction
5) Other potential errors
2.1.3 Multi-mission SSH altimetry data
2.1.4 SARAL-AltiKa, the new altimetry mission
2.2 Global mean sea level (GMSL) budget since altimetry era
2.2.1 Future needs
Chapter 3 Regional sea level variability and total relative sea level change
3.1 Regional sea level trend variability: Causes
3.1.1 Climate related regional sea level variability
1) Thermal expansion and salinity changes
2) Ocean mass changes
3.1.2 Non climatic causes for regional sea level variability
3.1.3 Vertical Land Motions
3.2 Long term regional sea level variability, total relative sea level change and coastal impacts
3.2.1 Indian Ocean
3.2.2 Caribbean Sea
3.2.3 South China Sea
3.2.4 The vulnerable zones: a synthesis
Chapter 4 The role of internal climate variability and external forcing on regional sea level variations
4.1 Internal climate variability
4.1.1 El Niño-Southern Oscillation (ENSO)
4.1.2 Pacific Decadal Oscillation (PDO)/ Interdecadal Pacific Oscillation (IPO)
4.1.3 Indian Ocean Dipole (IOD)
4.1.4 North Atlantic Oscillation (NAO)
4.1.5 Other modes of internal climate variability
4.2 Externally-forced climate variability
4.2.1 Natural external forcing
4.2.2 Anthropogenic external forcing
4.3 Detection and attribution of climate change
4.3.1 Detection and attribution on global mean sea level variations
4.3.2 Detection and attribution on regional sea level variability
4.4 The case of the Pacific Ocean
4.5 Role of external anthropogenic forcing on internal climate modes – A synthesis
4.6 Internal climate variability uncertainty in CMIP5 models
Conclusion and perspectives
