Coastal SAR and PLRM altimetry in German Bight and West Baltic Sea
Introduction
The objective of this paper is to quantify the capacity of the novel SAR (Synthetic Aperture Radar) altimetry concept in measuring the coastal dynamic topography processes when a special tailoring of both the Delay-Doppler processing and the SAR waveform retracking is carried out and when the state of the art auxiliary data and range and geophysical corrections are used.
This work is meant to be a continuation of the work carried out in the open sea (i.e. distance to coast larger than 10 km) by Fenoglio-Marc et al. (2015a). That previous study has shown that over open sea the SAR and PLRM (pseudo-LRM) altimetric measurements feature a good level of consistency between each other and against in situ data, SAR mode data having, anyhow, a higher level of precision. In the present study, we focus instead essentially on the coastal zone, defined here as the band of 0–10 km from the coast. Our goal is to assess the capacity of SAR altimetry to bring the altimetric measurements closer to the coast, in comparison to collocated PLRM pulse-limited altimetry, when a tailored coastal processing is used for both the SAR and pulse-limited altimetry.
In the first place, we emphasize the importance of having an accurate global altimetric dataset in the coastal zone: this is a highly dynamic and lively part of the marine environment, where processes and interactions between sea and land occur very quickly and at very short scales. The coastal zone plays a key role in the life of millions of people, being that more than 10% of the Earth population live in the low-elevation coastal zone area (Neumann et al., 2015) and hence potentially exposed to the effects of sea level rise, global warming and climate change. It is therefore mandatory to develop tools and techniques to monitor the sea level rise in the coastal zone, to include in the near future the coastal data in the global sea level budget and to demonstrate that coastal altimetry data can contribute effectively to the monitoring of regional sea level trends. The importance of coastal altimetry has been indeed recognized by the major space agencies, which have been supporting research and development (R&D) in the field. Thanks to this effort, progress has been achieved in the last years in the frame of projects such as ALTICORE (Lebedev et al., 2008), COASTALT (Cipollini et al., 2009), X-TRACK (Roblou et al., 2007), PISTACH (CLS Report, 2015), eSurge (Cipollini et al., 2012) and CP4O (Cotton, 2015). These projects aimed at improving conventional and SAR altimetry and at improving the range and geophysical corrections (in particular the wet tropospheric path delay and ocean tide) in the coastal zone. In particular, in the COASTALT project, a coastal-enhanced wet tropospheric correction, designated Global Navigation Satellite Systems (GNSS) derived Path Delay (GPD), was developed (Fernandes et al., 2010, Fernandes et al., 2015, Fernandes and Lázaro, 2016) that has been subsequently selected as the default wet tropospheric correction in the sea level calculation for the Sea Level CCI (Climate Change Initiative) project (Ablain et al., 2015).
In the last years, relevant advancement has been achieved in the field of pulse-limited waveform retracking (i.e. the process of extracting the geophysical altimetric measurements from the received waveform) in the coastal zone by adopting a sub-waveform retracking approach (as Guo et al., 2010, RED3 in CLS CLS Report, 2015, Idris and Deng, 2012) and an adaptive sub-waveform retracking (as ALES in Passaro et al., 2014). The latter authors succeeded in mitigating the impact of land contamination (off-nadir returns from the land synchronous with nadir return from the ocean, see (Vignudelli et al., 2011) and bright targets (as wetlands, mud flats, reefs, etc.) on the retrieved coastal measurements, by retracking only a subset of the waveform. Despite these advances in conventional altimetry, due to the novelty of the SAR altimetry mode, not much work has yet been carried out in the field of SAR coastal retracking.
In spite of its importance for climate change monitoring, dedicated coastal retracking is not yet integrated in the operational altimetry Payload Data Ground Segment (PDGS), coastal datasets are being produced only on a prototypal basis (see http://www.coastalt.eu/#datasets for available coastal datasets), there is still no general consensus on the optimal approach to process altimetric measurements in coastal zone (Vignudelli et al., 2011), and the sea state bias in coastal zone is still a major issue (Cipollini et al., 2010). Consequently, most altimetry data collected in coastal zone over the last 25 years are still largely unexploited or flagged as invalid in the ground segment altimetric products.
Here, we briefly introduce the concept of SAR altimetry and highlight its differences with respect to conventional pulse-limited altimetry. The main difference between these two measurement modes is related to the frequency used to transmit the pulses, which is called the Pulse Repetition Frequency (PRF). In pulse-limited mode, transmitted and received pulses are interleaved, i.e. pulses are received and transmitted continuously and reflections from the transmitted pulses are processed incoherently on a pulse-by-pulse basis (for more details see Fu and Cazenave, 2001). In SAR closed-burst mode, the pulses are transmitted and received in bursts with much higher PRF, so that successive echoes in a burst are correlated (Raney, 1998). This pulse-to-pulse coherence allows the application of the Delay-Doppler concept (Raney, 1998), which exploits the additional information coming from the relative Doppler phase between subsequent burst echoes, synthesizing a synthetic antenna aperture much larger that the real one. As result, a SAR altimeter will have a finer along track resolution than a pulse-limited altimeter but, since the sharpening is just in the along-track direction, the SAR altimeter and the pulse-limited altimeter share the same across-track resolution. Hence, after the acquisition of the high-rate unprocessed burst echoes, the echoes are coherently processed using the Delay-Doppler algorithm, producing the multi-looked SAR waveform (L1b). Because of the smaller footprint of a SAR altimeter, the ocean return waveform will have a different shape than the pulse-limited return waveform. This brought to the need to develop a novel return waveform physical model for SAR altimetry (see Ray et al. (2015) for an analytical physical model of SAR return waveform).
The first expected advantage of SAR altimetry over the ocean is a higher precision, due to the noise reduction made possible by the larger number of averaged echoes reflected from the same spot (looks) and higher SNR (signal to noise ratio) of the received signal. This benefit was demonstrated and reported by several authors e.g. Fenoglio-Marc et al., 2015a, Gommenginger et al., 2014. The second prominent advantage is a finer along track spatial resolution, thanks to the smaller footprint in the along-track direction which does not increase with the Significant Wave Height (SWH) as in conventional altimetry. Such a reduction of the altimeter footprint gives to SAR mode an intrinsic ability to resolve shorter scale ocean features and capacity to provide more robust and accurate sea state measurements in the coastal zone, being less sensitive to the land contamination.
Anyhow, because the footprint reduction occurs only in along track direction while the across track resolution remains basically unaltered with respect to the conventional altimetry case, it is expected that SAR altimetry brings altimetric measurements closer to coast especially when approaching the coast in orthogonal direction (Dinardo et al., 2011).
Hence, as a step forward from conventional altimetry, SAR altimetry is expected to provide the three basic geophysical parameters, sea surface height (SSH), significant wave height (SWH) and wind speed at 10 m height (U10), with enhanced accuracy and resolution in the coastal zone. These are indeed the three radar altimetry basic measurements which are estimated (in case of physics-based waveform retracking), by fitting the received return waveform with a waveform’s model that best-approximates the received waveform shape. The sea surface height is the elevation of the sea surface with respect to a reference ellipsoid, the significant wave height can be defined as four times the standard deviation of the sea surface elevation distribution and the wind speed at 10 m height is the wind speed over the sea surface computed at a reference altitude of 10 m (for more information refer to Fu and Cazenave, 2001).
In our study, the considered SAR mode data are those produced by the CryoSat-2 mission.
CryoSat-2 is an ESA (European Space Agency) Earth explorer mission, launched on 8 April 2010, devoted to the measurement of the sea ice thickness, land ice sheets and mountain glaciers elevations (ESA Report, 2007). CryoSat-2 flies on a Low Earth polar orbit with a long repeat cycle (369 days). In order to fulfil the tight cryospheric mission requirements, CryoSat-2 was equipped with a SAR interferometric radar altimeter (SIRAL) (Wingham et al., 2004). SIRAL is able to work in three measurement modes: SAR, SARin (SAR interferometry), and LRM (Low Rate Mode, i.e. the CryoSat-2 terminology for conventional pulse-limited mode). A distinctive feature of CryoSat-2 SIRAL, differently to all the former radar altimeters, is that it does not carry out any processing on board to build the multilooked waveforms when operating in SAR and SARin mode, but it downlinks to ground all the received pulses, after digitizing them. On ground, the unprocessed pulses are geo-located, time-referenced and combined with calibration quantities to form a L1A product named FBR (Full Bit Rate, the CryoSat-2 terminology for Level 1A).
Given the technical design of CryoSat-2 altimeter, simultaneous operations in SAR mode and LRM are not possible. However, when the radar altimeter is in SAR mode, it is possible to process the SAR burst data in the pulse-limited sense, to obtain a proxy of LRM, called Pseudo-LRM (PLRM) or also Reduced-SAR (RDSAR) (Scharroo, 2016, Martin-Puig et al., 2008). We point out that PLRM data are expected to feature lower performance (in terms of precision) than standard LRM mode data (Fenoglio-Marc et al., 2015a). When we assess the benefit of SAR altimetry against the pulse-limited altimetry, we use as reference this PLRM dataset and, hence, we are strictly assessing the performance of the SAR mode in comparison to the PLRM mode.
Section 2 presents the region of interest and the data (altimetry, auxiliary and range corrections). Section 3 presents the SAR and PLRM methodology; Section 4 shows the results from the cross-comparison between SAR, PLRM, ocean models and in situ data; finally, the conclusions are addressed in Section 5.
Section snippets
Region of interest
The region of interest, selected for the present study, is bound by the geographic coordinates (52°N to 60°N; 4°E to 16°E) and consists of the Eastern North Sea and the West Baltic Sea. CryoSat-2 operates in SAR mode over this region. The study area is depicted in Fig. 1.
The bathymetry is deep in the northern part, along the Norwegian coasts, and shallow in the rest of the region. The German Bight is the south-eastern bight of the North Sea bound by The Netherlands and Germany in the south and
SAR processing
In order to exploit the cloud computing power offered by ESA-ESRIN GPOD platform (https://gpod.eo.esa.int/), the developed SAR processor software has been compiled, deployed and run on this platform. Subsequently, the CryoSat-2 SAR mode L2 data have been extracted from the ESA-ESRIN GPOD platform.
The processor is also made available as a user service named SARvatore (Dinardo et al., 2014) at the web address https://gpod.eo.esa.int/services/CRYOSAT_SAR/. In this web page, different processing
Validation of SLAio
The validation metrics used in coastal zone for the cross-comparison of altimetric and in-situ data includes the following three parameters: regression slope, correlation and standard deviation of the differences (stdd) between in-situ and 1-Hz altimetric measurements. We perform also the comparison in open-ocean to analyse the data degradation when passing from open-ocean to coastal zone conditions. PLRM data have been retracked both with the SINC2 and TALES retrackers in order to analyse how
Conclusions
In this study, we have investigated, reported and discussed the improvements achieved by satellite SAR altimetry in terms of accuracy in the three basic geophysical parameters sea surface height, wave height and wind speed in the coastal zone. The improvement brought by SAR altimetry is validated versus “state of the art” PLRM coastal altimetry, against regional high-resolution ocean models and in-situ data in the German Bight and the West Baltic Sea.
For this purpose, we have implemented a
Acknowledgments
The authors acknowledge the European Space Agency (ESA) for the CryoSat-2 Full Bit Rate Data Products and the ESA/ESRIN G-POD (grid processing on demand) team for providing the grid computing capability. In-situ data are kindly provided by the German Waterway and Shipping Administration (WSV) and German Federal Institute of Hydrology (BfG). We acknowledge the German Federal Maritime and Hydrographic Agency (BSH), the Deutscher Wetterdienst (DWD) and the European Centre for Medium-range Weather
References (65)
- et al.
Regional in situ validation of satellite altimeters: Calibration and cross-calibration results at the Corsican sites
Adv. Space Res.
(2013) - et al.
Assimilation of altimetry data for nonlinear shallow-water tides: quarter-diurnal tides of the Northwest European shelf
Cont. Shelf Res.
(2010) - et al.
The German bight: a validation of CryoSat-2 altimeter data in SAR mode
Adv. Space Res.
(2015) - et al.
Improved wet path delays for all ESA and reference altimetric missions
Remote Sens. Environ.
(2015) - et al.
Automated daily process for global ionospheric total electron content maps and satellite ocean altimeter ionospheric calibration based on Global Positioning System data
J. Atmos. Sol. Terr. Phys.
(1999) - et al.
ALES: a multi-mission adaptive subwaveform retracker for coastal and open ocean altimetry
Remote Sens. Environ.
(2014) - et al.
Cross-calibrating ALES Envisat and CryoSat-2 Delay-Doppler: a coastal altimetry study in the Indonesian Seas
Adv. Space Res.
(2016) Ku-band radar altimeter surface wind speed algorithm
Mar. Geod.
(2012)- et al.
Improved sea level record over the satellite altimetry era (1993–2010) from the Climate Change Initiative project
Ocean Sci.
(2015) - et al.
Handbook of Mathematical Functions
(1964)