Haiyan Li

Target Detection on the Ocean With the Relative Phase of Compact Polarimetry SAR

This paper discusses the potential for automatic ocean surveillance using compact linear polarization (CL-pol) synthetic aperture radar (SAR), with large area coverage. Here, the target is a wind farm in the North Sea. The relative phase, as derived from CL-pol SAR, is employed for detection of the wind turbines, apart from the wind turbines wakes, based on fine-mode quad-polarization (quad-pol) RADARSAT-2 (RS-2) images. The relative phase of CL-pol measurements improves the contrast between the wind turbines and their wakes, because it has opposite signs for these two entities. Moreover, there is almost no variation in the relative phase with respect to wind speed or incidence angle. The results are verified by high-sea-state cases, up to 8.7-m significant wave height and 24.3-m/s wind speed, and also 641 quad-pol RS-2 SAR images collocated with 52 National Data Buoy Center buoys at different incidence angles and sea states. Thus, the relative phase of CL-pol SAR provides new light into the problem of operational autodetection of man-made targets, under high-sea-state conditions, over large areas.

Analysis of the Polarimetric SAR Scattering Properties of Oil-Covered Waters

An analysis of the polarimetric scattering properties of oil-covered waters is conducted using the classic Poincaré ellipticity parameter chi (χ) and the degree of polarization (m) from the Stokes parameters of hybrid polarized mode synthetic aperture radar (SAR). Oil spills reduce m for all four cases considered in this study. However, for the natural oil seep case considered, χ has a change in signs, comparing oil-covered waters with the “clean” ocean surface. The χ sign reversal is basic for “sign difference oil spill detection methods.” However, a problem is that the oil spill related to the deep water horizon (DWH) disaster did not exhibit a reversal in χ signs, comparing the “clean” ocean surface to the area contaminated by crude oil. The scattering mechanism related to the oil seep is different from that of the DWH oil spill; the former is dominated with even bounce scattering and the latter is dominated by Bragg scattering, similar to that of the clean oil-free ocean surface, in the imaging area.

Sea Ice Characterization and Classification Using Hybrid Polarimetry SAR

A characteristic analysis of sea ice is conducted using the classic Poincare ellipticity parameter chi (χ) and the degree of polarization (m) from the Stokes parameters of the compact polarized (CP) mode of Synthetic Aperture Radar (SAR). The polarimetric characteristics of sea ice for the simulated CP mode are compared with the classic polarimetric parameters, specifically the polarimetric entropy (H) and polarimetric scattering angle (α), for quad-polarized SAR systems. The physically based m-χ decomposition, obtained from the Stokes parameters of hybrid polarimetry mode CP SAR, is applied to sea ice classification, because the colorcoded images of the m-χ decomposition can be related to the basic scattering processes directly. Therefore, this methodology is able to provide more details in terms of sea ice classification than is possible with single polarimetic parameters. The classification results with the m-χ decomposition from CP SAR are shown to be comparable to those resulting from the Pauli RGB decomposition from quadpol SAR systems. Therefore, due to the planned wide swath widths for future CP SAR missions, such as the Canadian RADARSAT Constellation Mission, we suggest that hybrid polarimetry is an attractive mode for future CP SAR-based sea ice monitoring.

On polarimetric characteristics in SAR images of mesoscale cellular convection in the marine atmospheric boundary layer

[1]xa0Convection is an important phenomenon in the marine atmospheric boundary layer (MABL). Previous spaceborne radar studies of such have been limited to single polarization data, and therefore their focus was on the variation in intensity of the radar return, which was constrained by the existence of a single polarization image pattern, representing different atmospheric and oceanic phenomena. In this paper, we study the polarimetric characteristics of mesoscale cellular convection (MCC) in the MABL using high-resolution data from fully polarimetric (HH, VV, HV, and VH) RADARSAT-2 (RS-2) synthetic aperture radar (SAR) images, in conjunction with closely collocated mesoscale atmospheric model simulations, to identify the MCC signatures. To compare the polarimetric characteristics of MCC with those of the ocean surface, our analysis also includes 641 open ocean surface quad-polarization RS-2 SAR images collocated with 52 National Data Buoy Center buoys. The open ocean surface SAR images exhibit different polarimetric characteristics from those of MCC. Thus, we differentiate MCC from other open ocean phenomena, based on identifiable polarimetric SAR characteristics.

Ship detection with the fuzzy c-mean clustering algorithm using fully polarimetric SAR

A fuzzy c-mean clustering algorithm to detect ships is proposed using fully polarimetric SAR data. The algorithm is unsupervised. It does not need the statistical decision and the performance is not data specific, as often arises with CFAR methods. A distance measure, based on a complex Wishart distribution, is applied using the fuzzy c-means clustering algorithm. The algorithm makes use the statistical properties of polarimetric data, and takes advantage of a clustering algorithm. It is thus expected that the algorithm could include fully polarimetric backscattering information for ship detection. Its effectiveness is demonstrated by applying it to detect the targets in a set of AIRS AR data.

Oil spill detection on the ocean surface using hybrid polarimetric SAR imagery

Hybrid-polarimetric SAR (synthetic aperture radar) is a new SAR mode, with relatively simple architecture, low cost, and wide swath, which will be carried by several Earth-observing systems from now to the near future. Here, we show how the second Stokes parameter of hybrid-polarimetric SAR can be employed to detect oil on the ocean surface using the classic well-known Otsu threshold methodology, in relation to contributions from different polarizations and dampening effects on backscatter intensity, neglecting the specific scattering mechanisms and oil types for an oil-covered surface. The detection methodology is demonstrated to be reliable in three example cases: oil-on-water experiments conducted by the Norwegian Clean Seas Association, natural oil seeps from the Gulf of Mexico, and observations from the Deep Water Horizon oil spill disaster in 2010.

Ocean surface wave measurements from fully polarimetric SAR imagery

A new method for the retrieval of ocean wave parameters from SAR imagery is developed, based on the shape-from-shading (SFS) technique. Previously, the SFS technique has been used in the reconstruction of 3D landform information from SAR images, in order to generate elevation maps of topography for land surfaces. Here, in order to retrieve ocean wave characteristics, we apply the SFS methodology, together with a method to orient the angular measurements of the azimuth slope and range slope, in the measurement of ocean surface waves. This method is applied to high resolution fine-quad polarization mode (HH, VV, VH and HV) C-band RADARSAT-2 SAR imagery, in order to retrieve ocean wave spectra and extract wave parameters. Collocated in situ buoy measurements are used to validate the reliability of this method. Results show that the method can reliably estimate wave height, dominant wave period, dominant wave length and dominant wave direction from C-band SAR images. The advantage of this method is that it does not depend on modulation transfer functions (MTFs), in order to measure ocean surface waves. This method can be used in monitoring ocean surface wave propagation through open water areas into ice-covered areas, especially the marginal ice zone (MIZ) in polar oceans.

An analysis of the polarimetric scattering properties of oil spills on the ocean surface with hybrid polarimetry SAR

An analysis of the polarimetric scattering properties of oil-covered waters is conducted using the classic Poincaré ellipticity parameter chi (χ) from the Stokes parameters of hybrid polarized mode Synthetic Aperture Radar (SAR). For natural oil seeps, χ has a change in signs, comparing oil-covered waters with the `clean ocean surface. The χ sign reversal is basic to sign difference oil spill detection methods. However, a problem is that oil spills related to the Deep Water Horizon (DWH) disaster did not exhibit a reversal in χ signs, when `clean ocean surfaces are compared to the area contaminated by crude oil. Therefore, the scattering mechanism related to the oil seeps is different from that of the DWH oil spill; the former is dominated with even bounce scattering and the latter is dominated by Bragg scattering, similar to that of the clean oil-free ocean surface.

Detection of wind farm using the relative phase of compact polarimetry SAR

This paper discusses the potential for automatic ocean surveillance using compact linear polarization (CL-pol) SAR, with large area coverage, including high sea state conditions. Here, the target is a wind farm in the North Sea. The relative phase derived from CL-pol SAR is employed for detection of the wind turbines, apart from the wind turbines wakes, based on fine mode quad-polarization (quad-pol) RADARSAT-2 (RS-2) images. The relative phase of CL-pol measurements improves the contrast between the wind turbines and their wakes, because its sign is opposite for these two entities. Therefore, the relative phase of CL-pol SAR provides new light on the problem of operational auto-detection of man-made targets, under high sea state conditions, over large areas.

Wind Speed Retrieval From Hybrid-Pol Compact Polarization Synthetic Aperture Radar Images

This paper presents an attempt to retrieve wind speed from hybrid-pol compact polarization (CP) synthetic aperture radar (SAR) data. Cross-polarization (cross-pol, denoted by HV or VH) facilitates wind speed retrieval and improves the accuracy of the results, especially with respect to high wind speeds. Although the cross-pol data cannot be obtained from CP SAR data directly, these data can be reconstructed from CP SAR data. However, the existing reconstruction algorithms cannot satisfy the quantitative wind-speed retrieval requirements for specifying the critical parameter, denoted “N” in reconstruction algorithms, either as a constant, or as a variable with limited range. Here, N is defined in terms of the ratio between cross- and co-polarization channels and the coherence coefficient between co-polarization (denoted by HH and VV) channels. Thus, we have improved the empirical reconstruction algorithm for the modified N based on a data set of more than 2000 RADASAT-2 (RS-2) quad-polarization images and collocated buoy observations. The algorithm improves the accuracy for the reconstruction of cross-pol data and, ultimately, gives improved wind speed retrievals from the hybrid-pol CP SAR data. With the new algorithm, results show that the wind speed retrievals from reconstructed cross-pol data can approximate the accuracy of VH observations collected by RS-2.