Progress in satellite remote sensing for studying physical processes at the ocean surface and its borders with the atmosphere and sea ice

TLDR: Physical oceanography is the study of physical conditions, processes and variables within the ocean, including temperature-salinity distributions, mixing of the water column, waves, tides, currents.

Abstract: Physical oceanography is the study of physical conditions, processes and variables within the ocean, including temperature–salinity distributions, mixing of the water column, waves, tides, currents...

Figures

Figure 3 The Agulhas return current as seen from altimeter and passive microwave data on January 1, 2004. Top: Surface currents derived from altimeter sea surface heights (SSHs) and overlaid on the microwave sea surface temperature data. Bottom: the associated temperature gradient differences (labelled norm) with the same surface currents overlaid. Source: Reproduced with the permission of the authors from figure 1 of (Tandeo et al., 2014).

Figure 2: Example satellite remote sensing data of physical oceanography a) – 5 March 2014 Aqua-MODIS data showing the surface signatures of internal waves in the Andaman Sea (source: NASA Ocean Colour Biology Group); b) Surface currents inferred from visible spectrum ocean colour data using maximum cross correlation. Source: European Space Agency GlobCurrent project and Dr Weidong Xu.

Figure 1: Timelines for all the key ocean remote-sensing instruments used to studying and monitoring physical oceanography, with names given when different from the spacecraft. The passive microwave sensors do not all share the same frequency channels, as these depend upon the parameters of interest, and those co-manifested with an altimeter are nadir-viewing only rather than providing swath coverage. All the satellites are in low Earth orbits, except for those asterisked, which are in geostationary orbits. GLAS was a laser altimeter, whilst CHAMP, GRACE and GOCE measure the gravitational field. ERS-2 lost its last on-board tape recorder in June 2003, so subsequent coverage was limited to regions with direct reception. Source: constructed by co-author G. Quartly using Adobe Illustrator.

Figure 6 Global atmosphere-ocean CO2 gas fluxes derived from satellite remote-sensing, model and in situ data by the ESA OceanFlux Greenhouses project. a) global annually averaged mean air-sea CO2 flux in grams of carbon per metre squared per day (gC m-2 day-1). b) The position of persistent oceanic currents as derived satellite remote-sensing used to help interpret the global gas fluxes in a). Colours are only provided to indicate the strength and position of the surface features. Light blue suggests a stronger surface current, whereas dark blue suggests no current. Images courtesy of the European Space Agency.

Figure 5 The air-sea flux of gases and the various processes that are believed to control these fluxes. Original image by Jayne Doucette, Woods Hole Oceanographic Institution, and Wade McGillis, Columbia University.

Figure 4 Example of satellite altimetry verification of model performance. Cryosat-2 observations (filled circles) and tide gauge observations (filled squares) are overlaid the Danish Meteorological Institute (DMI) storm surge model field. The surge occurred during the Bodil storm 6-7 December 2013 and resulted in a 1000 years event in Roskilde, Denmark. Source: Generated by co-author J. Høyer.

Full text

1
Please note that this is an author-produced PDF of an article accepted for publication following peer review. The definitive
publisher-authenticated version is available on the publisher Web site.
Progress In Physical Geography
April 2016, Volume 40 Issue 2 Pages 215-246
http://dx.doi.org/10.1177/0309133316638957
http://archimer.ifremer.fr/doc/00333/44405/
© 2016 by SAGE Publications
Achimer
http://archimer.ifremer.fr
Progress in satellite remote sensing for studying physical
processes at the ocean surface and its borders with the
atmosphere and sea ice
Shutler Jamie D.
1, *
, Quartly Graham D.
2
, Donlon Craig J.
3
, Sathyendranath Shubha
2
, Platt Trevor
2
,
Chapron Bertrand
4
, Johannessen Johnny A.
5
, Girard-Ardhuin Fanny
4
, Nightingale Philip D.
2
,
Woolf David K.
6
, Hoyer Jacob L.
7
1
Univ Exeter, Exeter EX4 4QJ, Devon, England.
2
Plymouth Marine Lab, Plymouth, Devon, England.
3
European Space Agcy, F-75738 Paris 15, France.
4
Inst Francais Rech Exploitat Mer IFREMER, Nancy, France.
5
NERSC, Bergen, Norway.
6
Heriot Watt Univ, Edinburgh EH14 4AS, Midlothian, Scotland.
7
Danish Meteorol Inst, Odense, Denmark.
* Corresponding author : Jamie D. Shutler, email address : j.d.shutler@exeter.ac.uk
Abstract :
Physical oceanography is the study of physical conditions, processes and variables within the ocean,
including temperature-salinity distributions, mixing of the water column, waves, tides, currents and air-
sea interaction processes. Here we provide a critical review of how satellite sensors are being used to
study physical oceanography processes at the ocean surface and its borders with the atmosphere and
sea ice. The paper begins by describing the main sensor types that are used to observe the oceans
(visible, thermal infrared and microwave) and the specific observations that each of these sensor types
can provide. We then present a critical review of how these sensors and observations are being used to
study: (i) ocean surface currents, (ii) storm surges, (iii) sea ice, (iv) atmosphere-ocean gas exchange
and (v) surface heat fluxes via phytoplankton. Exciting advances include the use of multiple sensors in
synergy to observe temporally varying Arctic sea ice volume, atmosphere-ocean gas fluxes, and the
potential for four-dimensional water circulation observations. For each of these applications we explain
their relevance to society, review recent advances and capability, and provide a forward look at future
prospects and opportunities. We then more generally discuss future opportunities for oceanography-
focused remote sensing, which includes the unique European Union Copernicus programme, the
potential of the International Space Station and commercial miniature satellites. The increasing
availability of global satellite remote-sensing observations means that we are now entering an exciting
period for oceanography. The easy access to these high quality data and the continued development of
novel platforms is likely to drive further advances in remote sensing of the ocean and atmospheric
systems.

2
Please note that this is an author-produced PDF of an article accepted for publication following peer review. The definitive
publisher-authenticated version is available on the publisher Web site.
Keywords : Atmosphere-ocean interface, sea ice, remote sensing, surface currents, storm surge,
surface heat fluxes, atmosphere-ocean gas fluxes, oceanography
I Introduction
The oceans provide a plethora of services to society from climate regulation, the provision of food and
minerals, and the transport of goods. For instance, proteins derived from fish, crustaceans and molluscs
account for between 14% and 17% of the animal protein intake of the worlds human population (WHO,
2014) and 90% of the worlds trade is carried across the oceans (IAEA, 2014). Thus, studying and
monitoring the health of the oceans and their physical, chemical and biological components is key for
predicting future climate and maintaining life on Earth.
Physical oceanography is the study of the physical conditions, processes and variables within the
ocean. In this context the word ‘physical’ refers to the physical parameters of structure and motion. This
includes a large range of oceanic properties and characteristics including temperature-salinity
distribution, mixing, waves, tides, surface and sub-surface currents, and the exchanges (fluxes) of heat,
momentum and gas between the ocean and the atmosphere.
Space-based remote-sensing observations have an important role to play in oceanography research
and monitoring through providing quasi-synoptic and reproducible data for investigating processes on
global scales (Land et al., 2015). The present period (2016) is an exciting period of satellite remote-
sensing with sustained investment in long-term monitoring programmes scientifically driven from
international organisations, directives, space agencies and commercial enterprises. Satellites with a
large range of sensors are now in orbit and these new observations along with historical datasets are in
the most part freely and routinely

!
3!
available to a broad range of users over the Internet (figure 1). Individual sensors used for
remote-sensing may exploit visible light, infra-red or microwave radiation — the common theme
for all of these sensors is that the electromagnetic waves that they observe must interact with
the ocean, thus providing information on some property of the ocean surface or water column,
whilst not being completely absorbed or scattered by the atmosphere or ocean.
This critical review and forward look begins by providing an overview of the existing remote-
sensing technologies and the ocean properties that they observe, with a particular emphasis on
their use for studying physical processes at the ocean surface and its borders with the
atmosphere and sea-ice. These technologies are able to make observations at different spatial
scales of processes occurring in the oceans, from microscale (< 10 km), through mesoscale
(10-1000 km) to synoptic scale (>1000 km) and this paper presents examples of all of these.
We then focus on five areas where considerable research effort has been focused over the last
5 years with exciting developments for ocean and atmosphere science. These areas are
surface currents (section 2), storm surges (section 3), sea-ice (section 4), atmosphere-ocean
gas fluxes (section 5) and surface heat fluxes (section 6). The research in all of these areas
has exploited multiple sensors and in many cases has also begun to exploit multi-sensor
synergy approaches. Each of the sections in this paper begins with an overview of the
oceanography theory, briefly highlights the societal relevance of the application, and then
describes recent developments and advances. The paper then concludes in Section 7 with a
forward look and highlights future research and development opportunities, including the
European Union Copernicus programme and commercial satellite activities.
[Figure 1 here]
1 Visible spectrum sensors
Visible spectrum remote-sensing is commonly used for showing real-time environmental
conditions in the media, such as over-head views of hurricanes and weather systems. Such
images illustrate the challenges in visible spectrum remote-sensing: clouds, sea mist, or aerosol
contamination such as desert dust can all block the view of the (often more subtler) colour
changes within the surface ocean. These “ocean colour” sensors thus operate at a number of
narrow spectral bands, each of which covers specific wavelength regions of the visible spectrum

!
4!
(for example 8 spectral bands for the Sea Viewing Wide Field of View Sensor, SeaWiFS and 15
for Medium Resolution Imaging Spectrometer, MERIS) that permit a careful screening for
intervening haze, allow the calculation and removal of effects of the atmosphere, and relate the
spectra of water-leaving radiances to surface properties. Although the majority of ocean colour
data are used to study ocean biology such as surface chlorophyll-a concentrations (which is not
covered in this review, but a plethora of useful reviews are contained within the series of books
available from the International Ocean Colour Co-ordinating Group, IOCCG,
http://www.ioccg.org), information on some physical properties can also be derived. This is
because the biological signal in the ocean colour data can be used as a proxy for the structure
and/or motion of the water (as covered in sections 2 and 5). For instance areas of locally high
chlorophyll concentration may indicate regions of upwelling (upwelled water supplies nutrient-
rich water for the biology) or vertical lifting of near-surface layers caused by internal waves e.g.
(Muacho et al., 2013). The ocean colour signal can also act as an indicator of a particular water
mass, with boundaries between optically different water bodies being indicators of fronts
between regions of different temperature and salinity properties e.g. (Fratantoni and Glickson,
2002). Sequential ocean colour data can also be used to infer the velocity of features through
maximum cross correlation of high-resolution images (Yang et al., 2014) as described in section
2 of this paper.
2 Thermal infra-red sensors
Thermal infra-red sensors observe the thermal radiation that is emitted from the skin of the sea
surface, where the skin is considered to be top millimetre or so of the water. As is the case for
visible spectrum sensors, clouds and haze may obscure the sea surface or affect the signal that
reaches the satellite sensor. Thus thermal infra-red radiometers operate at multiple channels
(wavelengths) so that an atmospheric correction algorithm can be used to correct for
atmospheric attenuation, allowing the sea surface temperature (SST) to be determined e.g.
(McClain et al., 1985). As water is a strong emitter of thermal infra-red radiation, the received
signal corresponds to a skin depth of a few microns. This sea surface skin temperature controls
heat and gas exchange with the atmosphere, and, provided there is some turbulent mixing by
wind and waves is often close to the temperature a few metres below the surface, i.e. that
recorded by ships and floats. As well as providing temperature observations and contributing to

!
5!
calculation of heat and gas fluxes (see sections 5 and 6), thermal infra-red sensor data can also
be used to provide estimates of ocean currents through the apparent migration of features,
using techniques such as maximum cross-correlation (section 2).
[Figure 2 here]
3 Passive microwave sensors
In the frequency range 1-200 GHz range (wavelengths of 30 cm to 0.15 cm) the surface
emissivity of the ocean is affected by properties such as temperature, salinity and roughness.
Microwave sensors operating at these frequencies are often stated to have an "all weather
capability". This is not completely true, as although microwave signals are unperturbed by
clouds (offering a great advantage with respect to visible and thermal infra-red sensors),
significant atmospheric liquid (rain) will prevent the recovery of accurate information on the sea
surface. Emissions at 85 GHz act as a good discriminator for the presence of sea-ice (section
4.0) and enable observations at a spatial resolution of ~5 km (Comiso et al., 2003). Passive
microwave sensors typically have channels at 37 GHz and 22 GHz, which are used for the
detection and quantification of rain and water vapour respectively. In the absence of rain, wind
is an important influence through its effect on sea surface roughness; sensors typically use
combinations involving either 37 GHz or 11 GHz to infer wind speed (Wentz, 1997). We provide
case studies of its use in sections in sections 2 and 5. Microwave frequencies in the range 6.5-
11 GHz have a good sensitivity to SST although the contribution of wind speed needs to be
estimated and removed to determine accurate estimates of SST. The footprint size is typically
~40-60 km, and hence not suitable for recording filamentary (microscale) structures or any
features near the coast (as the land-sea adjacency effect contaminates the signal). Microwave
frequencies of about 1.4 GHz can be used to observe sea surface salinity (Gabarró et al., 2004).
However, both temperature and wind-roughening of the surface cause greater variability at this
frequency, so both wind speed and SST need to be recorded at the same time, or else a
reliable model estimate is required to allow the surface salinity to be determined. Only two
satellite sensors have been launched to specifically study ocean salinity. These are the Surface
Moisture Ocean Salinity, SMOS sensor and the NASA sensor Aquarius and they are already
providing new insights beyond their original mission focus e.g. Land et al. (2015); Reul et al.

Frequently asked questions

Q1. What are the contributions mentioned in the paper "Progress in satellite remote sensing for studying physical processes at the ocean surface and its borders with the atmosphere and sea ice " ?

Physical oceanography is the study of physical conditions, processes and variables within the ocean, including temperature-salinity distributions, mixing of the water column, waves, tides, currents and airsea interaction processes. Here the authors provide a critical review of how satellite sensors are being used to study physical oceanography processes at the ocean surface and its borders with the atmosphere and sea ice. The paper begins by describing the main sensor types that are used to observe the oceans ( visible, thermal infrared and microwave ) and the specific observations that each of these sensor types can provide. The authors then present a critical review of how these sensors and observations are being used to study: ( i ) ocean surface currents, ( ii ) storm surges, ( iii ) sea ice, ( iv ) atmosphere-ocean gas exchange and ( v ) surface heat fluxes via phytoplankton. For each of these applications the authors explain their relevance to society, review recent advances and capability, and provide a forward look at future prospects and opportunities. The authors then more generally discuss future opportunities for oceanographyfocused remote sensing, which includes the unique European Union Copernicus programme, the potential of the International Space Station and commercial miniature satellites. Exciting advances include the use of multiple sensors in synergy to observe temporally varying Arctic sea ice volume, atmosphere-ocean gas fluxes, and the potential for four-dimensional water circulation observations. The easy access to these high quality data and the continued development of novel platforms is likely to drive further advances in remote sensing of the ocean and atmospheric systems. 

Q2. What are the future works mentioned in the paper "Progress in satellite remote sensing for studying physical processes at the ocean surface and its borders with the atmosphere and sea ice " ?

Fortunately both of these gases have similar gas transfer velocity parameterisations to that of CO2 ; indeed CO2 transfer velocity parameterisations and data are often used to study N2O e. g. ( Rees et al., 2011 ). A relatively large in situ climatology of in-water DMS concentrations already exists ( Lana et al., 2011 ) and recent work has developed methods for using this climatology and satellite remotes sensing ( altimeter derived wind and mean square slope, and SST ) data to study global sea-air DMS fluxes ( Land et al., 2014 ; GoddijnMurphy et al., 2012 ). Until these datasets are publically available the potential of satellite remotesensing derived fluxes of these other gases is limited to regional and short-term studies where individual scientists have collected suitable in situ data. 

Q3. What is the main product that can be retrieved from remote-sensing?

Chlorophyll concentration is the one of the main products that can be retrieved from remote-sensing by visible spectral radiometry (ocean colour). 

Q4. What are the main absorbers of visible light in the open ocean?

The principal absorber of visible light in the open ocean are phytoplankton along with associated detritus and dissolved organic matter. 

Q5. What is the advantage of relating altimeter backscatter to transfer velocity?

Since the backscatter at nadir is closely related to mean square slope, there is a theoretical advantage in directly relating altimeter backscatter to transfer velocity. 

Q6. What is the main reason for the increase in thickness measurements?

The successful launches of more advanced satellite sensors, including ICESat-1 and Cryosat-2, is helping to reduce uncertainties in these thickness measurements through providing much richer and denser datasets, which is particularly important as the rate of decline in ice thickness appears to be increasing e.g. (Kwok and Rothrock, 2009). 

Q7. What are the key measurements for successful storm forecasting?

Key measurements for successful surge forecasting include timely and accurate observations of the storm track, water level, surface wind speeds and surface atmospheric pressure. 

Q8. What are the main uses of surface current measurements?

Surface current measurements over a broad range of spatial and temporal scales are important for a wide variety of applications including industry (e.g. oil drilling, shipping), provision of food (e.g. fishing), recreation (e.g. sailing), safety (e.g. search and rescue) and environmental research (e.g. climate modelling). 

Q9. What has been used to aid the understanding of gas flux data?

A range of studies have also used satellite remote-sensing to aid the understanding of gas flux data (e.g. figure 6b), as a proxy for in situ data that were not collected at the time of a research campaign, to replace in situ data that contained known errors, or to determine the suitability or sensitivity of the in situ data to describe the larger spatial and temporal variations. 

Q10. What other instruments and sensors are used for studying sea ice thickness?

Altimeter sensors, including RA2, ERS-1 and -2, Cryosat-2 and ICESat-1 (table 1) are the main instruments and sensors used for studying sea ice thickness. 

Q11. What is the effect of the presence of phytoplankton on the seasonality of mixed?

Wu et al. (2007) showed that, as well as advancing the phenology, the influence of phytoplankton on the seasonal progression of mixed-layer temperature and depth was such as to increase the maximum SST for the year by 1.5 °C in the Labrador Sea, and 2.5 °C over the Grand Bank of Newfoundland. 

Q12. What is the recent study of sea ice thickness?

Sea ice thickness is studied using submarine sonar data, airborne surveys, in situ measurements and satellite observations e.g. Kwok and Rothrock (2009). 

Q13. Why is there a lack of exploitation of satellite data?

This lack of exploitation of satellite data is due to the transient nature of the sea level and the often sparse spatio-temporal sampling characteristics of altimeter data. 

Q14. What was the first space-borne demonstration of the ATI technique?

The first space-borne demonstration was possible with the single-pass interferometric SAR system installed on space shuttle Endeavour for the Shuttle Radar Topography Mission (SRTM) in February 2000 (Rabus et al., 2003; Romeiser et al., 2010). 

Q15. What are the key uses of satellite observations for deriving disaster mitigation strategies?

Satellite observations are also key for deriving disaster mitigation strategies, for example to identify where the surge energy may be focused. 

Q16. What prompted Salter et al. (2011) to use historical in situ cruise data as?

All of this work prompted Salter et al. (2011) to used historical remote-sensing derived primary production data as a proxy for surfactant coverage to reanalyse historical in situ cruise data.