2:00pm - 2:20pm
Iceberg Detection and Analysis using Cryosat Modes
1IFREMER, France; 2CNES, France; 3Eumetsat, Germany
Conventional pulse limited (LRM) altimetry is a powerful tool to detect and characterize small (<3km) icebergs and to measure the profile of large ones (>16km). The Cryosat-2 SIRAL is the first altimeter that can operate in two new modes over the ocean besides the classical LRM: the Delay Doppler or SAR and the SAR Interferometric (SARin) modes. It offers thus a unique opportunity to test, validate and compare the capabilities of the three modes for the detection and analysis of small icebergs and the estimation of large iceberg topography. Over most of the ocean, SIRAL operates in LRM mode and the classical iceberg detection algorithm can be applied without modification. It can also be applied to the pseudo-LRM data, i.e. sequence of echoes derived from reducing SAR or SARin data into a sequence which looks like LRM data.
SAR altimeters have high pulse repetition frequency to ensure pulse-to-pulse coherence. The correlation of the returning echoes is used by the data processor to separate the echoes into strips arranged across the track by exploiting the slight Doppler frequency shifts, in the forward- and aft-looking parts of the beam. The strips laid down by successive bursts can therefore be superimposed on each other and averaged to reduce noise. This leads to an along-track resolution around 300m, improved SNR and enhanced ranging performance. The reduction of the noise level of the thermal noise part of the waveform used for detection facilitates the iceberg detection. The iceberg signatures, i.e. parabolas in the classical altimetry waveform space, reduce to bright spots in SAR waveform. They can be easily detected using classical image processing connected components and region properties algorithms. The iceberg area can be estimated using the signature's along-track and across-track lengths.
SARin is the most advanced mode, primary used around the ice sheet margins. SIRAL performs synthetic aperture processing and uses a second antenna as an interferometer to determine the across–track angle to the earliest radar returns. The SARin mode provides the exact surface location being measured. SAR and pseudo-LRM echoes can be used to detect icebergs. Coherence can be used to further improve the detection by limiting the probability of false alarm and by insuring the presence of a surface above the sea surface. In this mode the range analysis window (240m) is four times larger than that in LRM and SAR mode, which double the altimeter's detection swath. The main interest of SARin mode is the possibility, for the first time for a satellite sensor, to precisely locate the surface scatterer and to allow the estimation of the iceberg free-board from the phase difference and thus the iceberg volume. The high across-track accuracy allows to map the iceberg topography at an unprecedented resolution.
In SARin mode, the interferometric capabilities also allows to estimate the topography of large icebergs with an unprecedented resolution. Several passes over the B17A icebergs have been analyzed and the evolution of the surface topography has been analyzed during the lass year of life of the iceberg.
2:20pm - 2:40pm
Scientific Applications of Fully-Focused SAR Altimetry
1NOAA / UMD, United States of America; 2NOAA, United States of America
The delay/Doppler algorithm implemented in CryoSat-2 and Sentinel-3 applies a coherent processing the 64 echoes within each burst (about 3.5 milliseconds of flight), which allows narrowing the footprint in the direction along the track to about 300 m. However, by accounting for the phase evolution of the targets in the scene, it is possible to focus the complex echoes along the aperture, and perform inter-burst coherent integration potentially as long as the target illumination time. This process, similar to SAR imaging systems, reduces the along-track resolution down to the theoretical limit equal to L/2, where L is the antenna length. We call this the fully focused SAR (FF-SAR) Altimetry processing. For the development of the technique we have used the CryoSat-2 SAR Mode data, but our methods could also be used with similar data from Sentinel-3 or Sentinel-6/Jason-CS.
The footprint of an FF-SAR altimeter measurement is an elongated strip on the surface, which is pulse-limited across-track and SAR focused along-track. The technique has been demonstrated using transponder data, showing an achievable along-track resolution of 0.5 meters. Despite the asymmetry of the altimeter footprint, the fully focused technique may be useful for applications in which one needs to separate specific targets within highly heterogeneous scenes, such as in the case of sea-ice leads detection, hydrology, and coastal altimetry applications. Applying this technique on CryoSat-2 data over land and sea-ice, we can correctly measure the along-track extent of water bodies and ice-leads only a few meters long in the along-track dimension. On a random rough surface, independent FF-SAR waveforms can be obtained, potentially, every 0.5 meters, leading to an increase on the effective number of looks that can be obtained of the surface, with respect to delay/Doppler altimetry.
This paper concentrates on the cryospheric applications of the FF-SAR altimetry technique, and reviews the results that we have obtained so far from CryoSat-2 SAR mode observations over sea-ice, where a consistent performance improvement of square root of 2 with respect to the ESA L2 product is obtained for sea surface height estimates from sea-ice leads. In addition, due to the finer along-track resolution obtained with FF-SAR altimetry, we are now able to precisely determine the width of sea-ice leads from a nadir-looking altimeter. The performance improvement is, in any case, lower than expected for an ideal FF-SAR, as the closed burst operation mode of the CryoSat-2 SIRAL instrument imposes a lacunar sampling of the Doppler spectrum. This results in side lobes in the full along-track point target response, which introduces correlation in the successive looks of the ocean. This effect and possible mitigation strategies are also discussed in this work.
2:40pm - 3:00pm
As Assessment of Sea Ice Freeboard Derived from Fully-Focussed SAR Altimetry
1NASA Jet Propulsion Laboratory; 2NOAA Laboratory for Satellite Alimetry
We assess the performance of fully-focussed synthetic aperture radar (FF-SAR) altimetry from CryoSat-2 over Arctic sea ice. The altimeter onboard CryoSat-2 is operated in a closed-burst configuration, whereby coherent bursts of 64 echoes are processed to form 64 'beams' in the along-track direction. The aperture duration of 3.5ms narrows the along track footprint to around 300m, and successive looks at the same strip on the ground are summed (multi-looked) as the satellite passes overhead. On the other hand, by accounting for the phase evolution of scatterers in the scene, the FF-SAR technique applies an inter-burst coherent integration, potentially over the entire duration that a scatterer remains in the altimeter footprint. It has been demonstrated that applying FF-SAR processing to CryoSat-2 data can narrow the along track resolution to just 0.5m, or half the antenna diameter (the theoretical limit for SAR imaging). Further, the effective number of looks for the multi-looked echoes is approximately doubled, leading to better precision in the retrieved geophysical parameters.
Here, we explore the potential for FF-SAR processing to improve sea ice freeboard retrievals from radar altimetry. Further narrowing of the along-track footprint, and increasing the number of waveforms, allows more stringent waveform filtering, and retention of only the 'cleanest' lead/floe waveforms and smaller leads should become resolvable. For the SSH retrieval we will investigate whether this has the potential to reduce contamination by off-nadir leads, which can bias the SSH retrieval low. The increased effective number of looks should also improve the local sea level and sea ice elevation precision. We compare the freeboard retrieval from FF-SAR against the conventional Level-1b SAR data from CryoSat-2 and we make use of a coincident Operation IceBridge underflight to assess the results.
3:00pm - 3:20pm
Deriving IMO Polar Code Risk Index Outcome from Cryosat-2
1Finnish Meteorological Institute, Finland; 2Finnish Meteorological Institute, Ice Service, Finland; 3Alfred Wegener Institut, Germany
We present different methods to derive Polar Operational Limit Assessment Risk Indexing System (POLARIS) Risk Index Outcomes (RIO) from CryoSat-2 data. This work contributes towards the use of satellite radar altimeters in operation planning and tactical sea ice navigation.
The International Maritime Organization (IMO) adopted the international code for ships operating in polar waters or Polar Code in 2014. The polar code covers several aspects of navigation in ice covered seas spanning from ship design to training and environmental impacts. An important part of the polar code is the POLARIS system. The core of POLARIS is a number called Risk Index Outcome or RIO, which is an easy to interpret index depending on ship's ice class and sea ice conditions. Mathematically POLARIS is a 2d look-up-table of risk values (RV) for different ice conditions and ship ice classes. RIO is then calculated as a mean of these RV's weighted with their partial ice concentrations. RIO also takes into account potential ice breaker assistance and summer ice decay. RIO is easy to interpret with RIO > 0 standing for “operation permitted” and -10 > RIO > 0 and RIO < -10 for limited operation permitted and operation not permitted consequently.
Cryosat-2 thickness estimates are not trivial to interpret as RIO values. Gridded CryoSat-2 thickness products are averages of successful thickness retrievals falling within the averaging grid cell and do not include information on thickness distributions or lead detections. This information however is present in the L2 CryoSat-2 products. We present different approaches for deriving RIO from CryoSat-2 L2 products from the Pysiral software and compare them to RIO values calculated directly from the Canadian Ice Service operational ice charts. First we present methodology to derive RIO from CryoSat-2 L2 files from the ESA CCI Pysiral processor. We then explore the possibility to use a coupled sea-ice-atmosphere model NEMO-LIM3 for spatial interpolation of Cryosat-2 derived ice information. Finally we present a possibility to use an unsupervised classifier to derive adjusted IMO ice classes from Cryosat-2 data and convert these into RIO values.
3:20pm - 3:40pm
An Assessment of the State-of-the-Art Mean Sea Surface and Geoid Models of the Arctic Ocean: Implications for Sea Ice Freeboard Derivation
1National Space Institute, DTU Space, Denmark; 2NOAA Center for Weather and Climate Prediction, College Park, MD, USA; 3ESSIC, University of Maryland, College Park, MD, USA; 4Alfred Wegener Institute, Bremerhaven, Germany; 5IFREMER, Brest, France; 6CPOM, University College London, UK; 7York University, Canada; 8University of Bremen, Germany; 9Mullard Space Sciences Lab/University College London, UK
State-of-the-art Arctic Ocean mean sea surface (MSS) models and global geoid models (GGMs) are used to support sea ice freeboard estimation from satellite altimeters, as well as in oceanographic studies such as mapping sea level anomalies and mean dynamic ocean topography. However, errors in a given model in the high frequency domain, primarily due to unresolved gravity features, can result in errors in the estimated freeboard. These errors are exacerbated in areas with a sparse lead distribution in consolidated ice pack conditions. Additionally model errors can impact ocean geostrophic currents, derived from satellite altimeter data, while remaining biases in these models may impact longer-term, multi-sensor oceanographic time-series of sea level change in the Arctic.
Here we present results arising from the ESA CryoVal Sea Ice project, focusing on an assessment of four state-of-the-art Arctic MSS models (UCL13, DTU15/13/10) and a commonly used GGM (EGM2008). We describe errors due to unresolved gravity features and inter-satellite biases and their impact on the derivation of sea ice freeboard based on data obtained from CryoSat-2. The latest MSS models, incorporating CryoSat-2 sea surface height measurements, show improved definition of gravity features, such as the Gakkel Ridge. The differences between the models are analyzed and the results can be used to support improvements in future models. We find that the impact of MSS or GGM errors on freeboard can reach several decimeters in some parts of the Arctic. The maximum observed freeboard difference was 0.59 m (UCL13 MSS minus EGM2008 GGM).
So far the impact of the ocean mean dynamic topography (MDT) has largely been neglected in the freeboard processing chain, since existing MSS models have been so divergent. However, improvements in our ability to measure the sea surface height of the polar oceans with advanced altimetry techniques has improved our knowledge and the precision of the MDT. We compare Arctic MDT based on the differences between the MSS and GGM models, to understand the impact on MDT quality derived from historical versus current models. The results support algorithm decision for deriving freeboard from current and future missions such as CryoSat-2, Sentinel-3 and ICESat-2.