Elsevier

Advances in Space Research

Volume 55, Issue 11, 1 June 2015, Pages 2641-2656
Advances in Space Research

The German Bight: A validation of CryoSat-2 altimeter data in SAR mode

https://doi.org/10.1016/j.asr.2015.02.014Get rights and content

Highlights

  • Coherent and incoherent processing of SAR altimetry give consistent unbiased results.

  • Sea level and wave height from SAR are twice as precise as Pseudo Pulse Limited data.

  • Look Up Tables significantly benefit SWH retrieval during SAR altimeter processing.

Abstract

The retrieval of the three geophysical parameters – sea surface height above the reference ellipsoid (SSH), significant wave height (SWH) and wind speed at 10 m above the sea surface (U10) – is the main goal of satellite altimetry and of primary importance for climate research. The Synthetic Aperture Radar (SAR) altimetry is expected to provide improved precision and along-track resolution compared to the conventional low-resolution mode (LRM) radar altimetry.

CryoSat-2 enables a quantitative comparison of SAR and Pseudo-LRM (PLRM) data derived respectively from a coherent and an incoherent processing of the same SAR echoes. In this paper we perform their cross-validation and validation against in situ and model data to derive precision and accuracy at 1 Hz in open ocean, at distances larger than 10 km from the coast. The analysis is performed in the German Bight during 2011 and 2012.

Both the PLRM and the SAR scheme include waveform zero-padding and identical environmental, geophysical, and atmospheric attenuation corrections. A Look Up Table is additionally used in SAR to correct for approximations of the Point Target Response (PTR) applied in the retracking procedure.

The regional cross-validation analysis proves the good consistency between PLRM and SAR data, with no bias and rms differences of 3 cm, 21 cm, and 0.26 m/s for SSH, SWH, and U10, respectively. The precision of SSH and SWH is higher in SAR than in PLRM (by a factor of 2), while the precision of U10 is 1.4 times better in PLRM than in SAR. At 2 m waveheight, the SAR precision is 0.9 cm for SSH, 6.6 cm for SWH. and 5.8 cm/s for U10.

The in situ analysis shows that SSH and U10 have comparable accuracy in SAR and PLRM, while SWH has a significantly higher accuracy in SAR. With a maximum distance of 20 km between altimeter and in situ data, the minimum values obtained for their rms differences are 7 cm, 14 cm, and 1.3 m/s for SAR and 6 cm, 29 cm, and 1.4 m/s for PLRM.

Introduction

The CryoSat-2 satellite, launched in 2010, carries the first satellite radar altimeter with a delay/Doppler mode (Raney, 1998). The altimeter can operate in a burst mode similar to a Synthetic Aperture Radar (SAR) with markedly different characteristics than the previously used pulse-limited Low Repetition Mode (LRM). Conventional pulse limited altimeters employ incoherent processing, meaning that power is averaged without regard to phase. In SAR mode the altimeter employs coherent processing, by which is meant that the phase information in each echo is exploited as well as the echo’s power. In LRM the echoes are incoherently summed up to reduce the noise (multi-looking); analogously in SAR mode, the multi-looked waveform is built by an incoherent sum of the stack of echoes staring at the same ground cell (Wingham et al., 2004). This greater use of information (i.e. number of looks) can have advantages, in theory, and the present study is a test of that theory.

Another important difference between LRM and SAR lies in the transmission and reception modes. In LRM mode, the pulses are transmitted continuously and reflection from the transmitted pulses is processed on a pulse-by-pulse basis incoherently; this is called interleaved mode of transmission. In CryoSat-2 LRM mode, the Pulse Repetition Frequency (PRF) of 1970 Hz is close to the Walsh bound (Walsh, 1982) for which sequential returns are uncorrelated. This PRF gives the highest number of looks per second obtainable in LRM and is therefore optimal for incoherent processing. The PRF selected for the CryoSat-2 SAR mode is 17,800 Hz. In this mode CryoSat-2 produces closed bursts, i.e. every radar cycle a group of pulses (a burst) is transmitted at high PRF and subsequently the instrument waits silently until the pulses are all received prior to transmit a new burst. The alternative to this approach is the open burst or interleaved mode option, in which the transmitted and received pulses are interleaved but this approach requires a lower PRF to allow interleaved transmitted and received pulses. The Walsh upper bound generalized to burst-mode transmission gives a lower bound for the duration of the burst period, which is three times shorter than the CryoSat-2 burst period. The SAR CryoSat-2 PRF it is therefore sub-optimal and there is a potential for three times more uncorrelated looks in SAR by using the open burst option (Raney, 2012), as planned for the Jason-CS mission.

SAR altimetry is expected to have a higher precision than LRM, thanks to the noise reduction obtained by averaging the large number of averaged pulses. Earlier simulations by Raney, 1998, Jensen and Raney, 1998 predict a twofold noise reduction, they however assume the delay-Doppler operation to be in open burst instead of closed burst. A higher spatial resolution along-track is also expected, due to the smaller along-track footprint. The nominal along-track resolution is about 300 m and does not depend on the significant wave height as in conventional altimetry. This high resolution should allow resolving shorter scale ocean features.

Pseudo-LRM (PLRM) data can be derived from the CryoSat-2 Full Bit Rate (FBR) data by processing the pulse-limited echoes incoherently, as in the conventional LRM concept (Scharroo et al., 2013b, Scharroo, 2014, Smith and Scharroo, 2015). This technique is also known as reduced SAR (RDSAR). In this way both SAR and PLRM data are obtained from the same sampling and can be compared under the same sea state conditions. It is worth to emphasize that in the PLRM the waveforms are not statistically equivalent to the CryoSat-2 LRM waveforms, because of their lower precision caused by the smaller number of uncorrelated looks accumulated in building the multi-looked echo in PLRM in the closed burst option.

For the derivation of the geodetic parameters from the SAR and PLRM waveforms, different methods need to be used in SAR and PLRM, due to the different shape of the waveforms. PLRM waveforms are characterized by slowly decaying trailing edge and pronounced tilt of the leading edge for high SWH, while SAR are peakier, with fast decaying trailing edge and with peak broadening at high SWH. In LRM the analytical formulation of the standard model (Brown, 1977) is generally used, which consists of the double convolution of Flat Sea Surface Response (FSSR), the Radar Point Target Response (PTR) and the Probability Density Function (PDF) of surface elevation. In SAR altimetry, the reflected power is a two dimensional function of time (delay) and of the distance (Doppler frequency) and the Brown model is not suitable for fitting its shape. Numerical (Boy and Moreau, 2013), semi-analytical (Wingham et al., 2004), fully analytical (SAMOSA (Ray et al., 2015)) and empirical models for deriving the relevant parameters (range, significant wave height, and backscatter coefficient) have been developed.

In this paper we validate in open sea the three basic geophysical parameters in altimetry sea surface height (SSH), significant wave height (SWH), and near-surface wind speed (U10) derived from SAR and PLRM CryoSat-2 data. The main goal is to test whether the greater use of information made in the SAR processing has the expected advantages in terms of both precision and data accuracy. A second goal is to estimate possible biases occurring in SAR mode with respect to the PLRM Mode and to tune-up the SAR re-tracking scheme.

The German Bight has been selected as validation area for various reasons. First, the CryoSat-2 mission is operated in SAR mode in this area. Second, the German Bight it is characterized by generally low sea states. This shall allow us to assess the measurement capacity of SAR technology in regime of low sea state, which is nevertheless considered particularly challenging to be detected by retracking algorithms, since a sharp leading edge in a waveform is described by fewer gates and therefore the derived estimate of the rising time is proportionally less accurate. Third, the German Bight hosts a large network of in situ stations of high quality. The in situ data are from a network of stations co-located with Global Navigation Satellite Systems receivers (GNSS) and thus precisely located in a geodetic reference frame. The in situ sea level measurements are easily referred to the reference ellipsoid like the altimeter measurements and an absolute validation of the altimeter range is therefore possible (Jin et al., 2013). Finally various regional models of high quality are available in the area. We use two regional wave and wind speed models.

In Section 2 we describe the data used and the validation methodology. In Section 3 the validation is made by comparison to in situ and to model data. In Section 4 we discuss our results and provide an outlook for future investigations.

Section snippets

Altimeter data

CryoSat-2 FBR data in SAR mode taken over the German Bight during the year 2012 and 2013 have been processed to produce both PLRM and SAR waveforms at 20 Hz. In both cases zero-padding has been applied before range compression, as suggested by Smith and Scharroo (2015). This step avoids signal aliasing for low SWH regimes and ultimately produces more accurate retrieval of mainly SWH at low sea states.

From the PLRM and SAR waveforms we derived values of SSH, SWH, and U10 at 20 Hz, which were

Regional cross-validation of CryoSat-2 SAR and PLRM

As indicator for the precision of each parameter at 1 Hz, we consider the standard deviation of the parameter itself (SSH, SWH or U10) over 20 consecutive 20 Hz measurements and evaluate those as a function of SWH. We then compute the median of the standard deviations for each SWH to obtain the performance curve and we particularly consider the value for a SWH of 2 m.

Fig. 3 shows the results for SSH. Both SAR SSH and SWH measurements have a higher precision than PLRM; for a SWH of 2 m, the

Discussion of results and conclusion

We have validated in open sea the three basic geophysical parameters for altimetry, namely sea surface height, significant wave height, and near-surface wind speed, derived from SAR and Pseudo-LRM CryoSat-2 data. The data were generated from the same echoes using both coherent SAR and the incoherent LRM processing, which allow a comparison under same sea state conditions.

The main goal was here to test whether the greater use of information made in the SAR processing has the expected advantages

Acknowledgments

The authors acknowledge the European Space Agency (ESA) for the CryoSat-2 Full Bit Rate Data Products and the National Oceanic and Atmospheric Administration (NOAA) for the PLRM processed data. In-situ data are provided by the German Waterway and Shipping Administration (WSV), the Bundesanstalt für Gewässerkunde (BfG), the German Federal Agency for Cartography and Geodesy (BKG), the Bundesamt fuer Seeschifffahrt und Hydrographe (BSH), the Bundesumweltministerium (BMU), Project Management

References (45)

  • Boy, F., Moreau, T., 2013. Algorithm Theoretical Basis Document (ATBD) of the CPP SAR numerical retracker for oceans,...
  • G.S. Brown

    The average impulse response of a rough surface and its applications

    IEEE Trans. Antennas Propag.

    (1977)
  • Dick, S., Kleine, E., Müller Navarra, Klein, H., Komo, H., 2001. The Operational Circulation Model of BSH (BSHcmod)...
  • Dinardo S., Lucas B., Benveniste J., 2013. A toolkit for CryoSat investigations by the ESRIN EOP-SER altimetry team....
  • ESA ESRIN and MSSLUCL Product Handbook,...
  • ESA ESRIN, 2013. Guidelines for reverting Waveform Power to Sigma Nought for CryoSat-2 in SAR mode, Technical...
  • L. Fenoglio-Marc et al.

    Vertical land motion in the Mediterranean sea from altimetry and tide gauge stations

    Mar. Geod.

    (2004)
  • Fenoglio-Marc, L., Weiss, R., Dinardo, S., Becker, M., Sudau, A., 2013a. A study on the conformance of altimetry and in...
  • Fenoglio-Marc L., Dinardo S., Scharroo R., Roland A., Lucas B., Weiss R., Dutour M., Becker M., Benveniste, J.A.,...
  • Fenoglio-Marc L., Dinardo S., Scharroo R., Roland A., Lucas B., Weiss R., Dutour M., Becker M., Benveniste, J., 2014....
  • Goffinet, P., Blasi, J., Sudau, A., Liebsch, G., 2011. National Report of Germany. GLOSS-Report...
  • C. Gommenginger et al.

    Retracking altimeter waveforms near the coasts

  • Cited by (66)

    • The technical progress and achievements of coastal altimetry waveform retracking over the Asia seas

      2023, Coastal Altimetry: Selected Case Studies from Asian Shelf Seas
    • Interference-sensitive coastal SAR altimetry retracking strategy for measuring significant wave height

      2022, Remote Sensing of Environment
      Citation Excerpt :

      Intrinsic noise is defined as the standard deviation of 20-Hz SWH data within a 1-Hz distance. This definition is based on the assumption that the variability at the 20-Hz posting rate (equivalent to an along-track distance of ∼330 m) is mostly dominated by noise (Ardhuin et al., 2019; Fenoglio-Marc et al., 2015). In order to compute the STD of the twenty 20-Hz measurements, we impose the constraint that there must be least 17 valid measurements.

    View all citing articles on Scopus
    View full text