Research Article  Open Access
Microwave Emission and Scattering from Ocean Surface Waves in the Southern Beaufort Sea
Abstract
Estimates of the relationships between geophysical variables and microwave backscatter/emission are important for the evaluation of atmosphereocean interaction, as well as energy, and mass transfer across this interface. We evaluate shipbased passive microwave brightness temperatures at 37 and 89 GHz and active polarimetric backscatter at 5.5 GHz (Cband), as these relate to buoyderived ocean wave parameters for distinct wave regimes in the southern Beaufort Sea. Microwave emission and backscatter are shown to be sensitive to the ocean surface physical roughness as defined by the significant wave height , compared to wind speed. The shows significant correlation with , with the strongest correlation for the Hpolarization channel at 37 and 89 GHz. Active co and cross polarization ratios at 40° incidence angle are associated with , with the increase proportional to . The polarimetric coherence parameter at 20° also shows an inverse relationship with because of an expected decorrelation of complex returns with greater surface roughness.
1. Introduction
The southern Beaufort Sea is of interest because of its dynamic seasonal oceansea iceatmosphere interaction [1, 2] and the seasonal formation of flaw leads and polynyas [3, 4]. Microwave remote sensing has played a key role in investigations, with scientists studying the passive and active microwave signatures from the ocean surface for several decades. It has now been possible to estimate numerous geophysical parameters with sufficient accuracy based on microwave data collected using satellite sensors. However, relationship between surface waves and the radiative transfer across the atmosphereocean (AO) interface is, yet, not fully realized in the literature.
Ocean surface waves modify AO interaction and concurrently affect observed passive microwave brightness temperatures () and radar returns (hereinafter backscatter). Enhanced surface roughness from windgenerated surface waves significantly affects the formation and evolution of the oceanic boundary layer [5], which in turn causes greater atmospheric instability and turbulent exchange at the AO interface [6]. Physical roughness also acts as an important component of feedback processes which transfer heat and momentum across AO interface. Though there is good understanding of the transfer of wind energy to the ocean surface waves (aerodynamic roughness), the microwave behavior of related geophysical variables such as significant wave height () (hereinafter wave height), power spectral density (PSD), ocean surface temperature, and nearsurface water vapor is the subject of further study. Open water backscatter and observations represent composite signals that contain contributions from the full suite of these variables. In accordance, the application of electromagnetic (EM) signatures to geophysical inversion algorithms first demands consideration of the full suite of variables.
Previous research into AO surface roughness from both passive and active microwave remote sensing indicates considerable gaps in understanding the EM response to roughness elements (e.g., whitecaps, sea spray, and wave RMS (root mean square) height and correlation length). Polarimetric passive microwave signatures of the ocean surface at frequencies 10.7, 19.35, and 37 GHz are well modeled by a twoscale polarimetric emission model [7, 8]; however, the assimilation of effects of wind direction, windgenerated waves, atmospheric stability, and sea foam into polarimetric emission models of ocean surface is a topic requiring further research. The undulations of the ocean surface significantly affect the observed, frequencydependent, , which is dependent on the windinduced wave structure, that is, the sea state and foam coverage [9, 10]. Kravtsov and Churyumov [11] mathematically delineated the effect of steep irregularities and wave slope on the . Monahan and O’Muircheartaigh [12] used remote sensing to include whitecaps and foam in the wind retrieval algorithms. Wave height can also be linked to as a close function of surface wind speed [13]. Nordberg et al. [14] provided the measurements of microwave emission at 19 GHz from a foamcovered and winddriven sea. One observes an increase in the with an increase in foam coverage and whitecaps in all surface conditions mentioned above.
Microwave Cband (5.5 GHz, 5.6 cm wavelength) backscatter is more reliable for studying ocean surface waves than shorter wavelength Ku, Ka, and Xbands. Capillarity and molecular viscosity (water temperature) strongly influence the radar return in these bands [15, 16]. There are numerous studies readily found on relationships between Cband copolarization backscatter (HH and VV) and wind speed and direction, with geophysical model transfer functions (e.g., CMOD) widely used to retrieve wind information from backscatter [17–20]. A powerlaw relationship holds for backscatter and friction velocity [21] and between backscatter and wind speed [22]. Keller et al. [22] also observed backscatter increasing with decreasing AO temperature difference. Thompson et al. [23] found that the observed copolarization ratio (HH/VV) is larger than predicted by backscatter models, as these models inadequately address the nonlinearity of the longwave portion of a rough ocean surface. Thompson et al. [23] provided an incidenceangledependent polarization ratio model. Mouche et al. [24] extended this model to include wind speed, wind direction, sea state, and swell effects. The results of Mouche et al. [24] show that the copolarization ratio () is dependent more on wave steepness than on wave height and wind speed.
The waves under rough conditions are typically 3–10 times larger than the incident radar wavelength at Cband [25, 26]. Kudryavtsev et al. [25] suggested a composite backscatter model for describing backscatter (at moderate incidence angles 20°–60°) from ocean surface waves which included nonBragg scattering components caused by breaking waves (wavelength from few millimeters to few meters) based on wave breaking statistics proposed by Phillips [27]. One can use the to study the discrepancy between the Bragg model and observations, and it can explain the nonBragg component of scattering [25]. Valenzuela [28] provided a review of the theories addressing EM wave interaction with the small gravitycapillary waves, citing Bragg (resonant) scattering as the most important mechanism of ocean surface—EM interaction. Energy dissipation over different length scales and breaking wave mechanisms is largely linked to nonBragg scattering, though there is little understanding of their effects on backscatter. The and wave PSD measured from Waverider buoy are investigated here, which contain contributions from the nonBragg scattering elements, using the crosspolarization channel ratio () and polarimetric coherence parameters, in addition to .
This paper addresses open ocean surface waves (physical roughness) during the icefree season, with the overarching objective, to further our understanding of the role of and backscatter in estimating ocean surface waves and, by proxy, AO exchanges. We examine near surface (shipbased) passive microwave (37 and 89 GHz) and polarimetric Cband backscatter, as these relate to buoy measurements of and PSD. This comparison is rarely available in the literature. We investigate the utility of and , as well as polarimetric coherences from the ocean surface, to contribute to the realization of polarimetric synthetic aperture radar for deriving ocean surface wave information.
2. Materials and Methods
2.1. Study Area
The study area lies in Amundsen Gulf in the southern Beaufort Sea (Figure 1). The Cape Bathurst Polynya forms in the area and hosts several flaw leads throughout the winter [1] and the region becomes icefree in the summer. We acquired field data between July 1 and 28, 2008 under the auspices of the Circumpolar Flaw Lead (CFL) System Study, an International Polar Year (IPY) project that took place between October 2007 and August 2008 aboard the Canadian Research Icebreaker Amundsen [29].
2.2. Instruments’ Description and Data Collection
We acquired ocean wave height, wave steepness, wave period, and wave phase speed using a meteorological ocean (Datawell, Haarlem, The Netherlands) Mini Directional Waverider G4 (DWRG4) buoy. The buoy has a hull diameter of 0.4 m and weight in air of 17 kg. The buoy is capable of measuring wave height and direction for periods of 1.6–100 sec and covers the expected range of wave periods observed in the Arctic (~60 sec; see, e.g., [30]). Wave height accuracy is 1 cm in all directions while freely floating. We computed from the zeroth moment of the wave spectrum. The accuracy of wave PSD, wavelength, and other derived wave parameters relies on wave height accuracy. We deployed the buoy by small boat at a minimum distance of 200 m to avoid interference by the Amundsen. Wave data were logged at 30minute intervals from a stabilized accelerometer. We collected continuous meteorological data at oneminute interval from a tower mounted on the bow of the Amundsen. An RMYoung model 5103 Wind Monitor provided the wind speed () data at 14 m above sea level (ASL) at accuracy of ±0.3 m/s or 1% of the reading. The Amundsen’s AXYS Automated Voluntary Observing Ship (AVOS) sensors provided water temperature at the sea surface with a resolution of 0.1°C and an accuracy of ±0.3°C.
We have used shipbased dualpolarized (H and V) passive microwave (Radiometrics, Boulder, CO, USA) radiometers operating at 37 and 89 GHz frequencies to acquire emissions at a fixed incidence angle of 53°, the same angle as the spaceborne SSM/I (Special Sensor Microwave/Imager) sensor and within 2° of the spaceborne AMSRE (Advanced Microwave Scanning Radiometer—Earth Observing System) sensor. We installed the system on the rail mount on the port side of the Amundsen at approximately 12 m ASL. To convert recorded voltages to , we followed the calibration method described by Farmer et al. [31]. The data were acquired for 24hour full cycle for all days and were averaged for every 30 minutes for the specified dates during the month of July 2008 at the locations shown in Figure 1. data were matched with buoy data over 2–7 hours duration for comparison in this study. This paper presents the relationships of and PSD and shipbased at meter scales (footprint at 89 GHz at 55°: 6.36 m^{2}; at 37 GHz: 6.62 m^{2}) at 0.81 cm (37 GHz) and 0.34 cm (89 GHz) wavelengths.
We have used a Cband fully polarimetric scatterometer (ProSensing Inc., Amherst, MA, USA) (Table 1) installed on the port side of the Amundsen at a height of 7.6 m ASL to collect backscatter signatures coincident to ocean buoy data over short (<0.5 hour) periods on 1, 9, and 26 of July. We collected backscatter and phase data at all four linear transmit/receive polarization combinations (HH, HV, VH, and VV) over the 20°–60° incidence angle range at 10° spacing and across a 60° azimuth centered perpendicular to the Amundsen. At 7.6 m height, and with the scatterometer positioned at a fixed azimuth and 45° incidence angle, its estimated footprint was 1.1 m^{2}. A single scan comprised several resolution cells over several meters by extending the azimuth to a 60° range, resulting in multiple statistically independent samples of the sea surface at each incidence angle. The use of a wide azimuth also increased the number of radar looks and reduced fading, though the resulting experimental dataset does not consider the wave direction relative to the scanning scatterometer. Isleifson et al. [32] and Geldsetzer et al. [33] have described the instrument calibration, noise levels, setup, and data limitations.
 
Note. NRCS: normalized radar crosssection; RF: radio frequency. 
We placed a limit of 3 dB of signaltonoise ratio on backscatter for analysis of scatterometer data, resulting in an upper incidence angle limit of 40° for scans taken during calm conditions. This limit extends to all samples to enable the intercomparison of all backscatter irrespective of sea state. The average covariance matrix, from which we derive HH, HV, and VV backscatter, as well as and , using backscatter amplitudes, represents the scan data. We obtained the following polarimetric coherence parameters: the copolarized correlation coefficient (); the crosspolarized correlation coefficient (); and the coherence of right and left circular transmit/receive polarizations () using the following formulae [34–36]: where is the complex scattering matrix and an asterisk (*) represents the complex conjugate. The brackets represent ensemble averages of the observed data. Polarimetric coherences and polarization ratios have demonstrated utility in reducing the ambiguities caused by the nonlinearity between system response and target properties [37]. Further details on the derivation of , a parameter shown to be sensitive to surface roughness, are available in [38, 39].
3. Results and Discussion
3.1. Ocean Surface Wave Conditions
The data primarily represent wave breaking and energy dissipation of waves depending on the prevailing wind conditions. From the measured buoy data, ranged between 12 and 214 cm (sea state on Beaufort scale 0–5) compared to those by Hwang et al. [40] who reported a range of 120–270 cm in the Atlantic Ocean (Figure 2). Observed wave numbers ranged between 0.083 and 0.729 rad/m, and the wind speeds ranged between 0.5 and 10.5 m/s compared to wind speed range of 7.0–14.5 m/s given by Hwang et al. [40]. The wave steepness, in this paper, ranged between 0.011 and 0.050, which corresponded with the wavelength ranging between 8.61 and 75.56 m. The wave periods of windgenerated waves addressed in this paper lie between 2.4 and 7.0 sec. We did not observe swells (wave period > 10 sec [41]) during the field program. The observed relationship between the nondimensional form of energy, (where is gravitational acceleration), and wave age, (where is wave phase speed), as initially proposed by Kitaigorodskii [42] and further developed by Donelan et al. [43], supports the windgenerated wave growth in the data (Figure 3). The RMS error in the fitted curve is 0.024; the coefficient of determination, ; and sum of residual errors, SSE = 0.052. The observed empirical relationship is
(a)
(b)
Table 2 shows wave roughness parameters and PSD, coincident to scatterometer acquisitions on July 1, 9, and 26, 2008, along with . While the scatterometer took eight minutes to complete one scan, wave height and PSD data were logged at 30minute intervals. Table 2 also provides a summary of independent scatterometer samples acquired at 20° and 40° incidence angles, the near and far ranges examined in this study. Figure 4 shows the wave spectra from the same dates relative to the analytical PiersonMoskowitz (PM), Bretschneider, and 3parameter Gamma wave spectra. These wave spectra considered both capillary and gravity wave (wavelength > 1.74 cm) development. The observed spectra is in agreement with the modeled spectra for the case on July 9, with peak frequencies of ~0.2 Hz. The observed spectra agree more closely with the PM model at wave frequency greater than 0.2 Hz for other cases. The cases also contain more observed wave energy in the low frequency part of the observed spectra, indicating the presence of longer waves. The low frequency (larger waves) part of the spectra is marginally modeled by the wave spectra models, likely because of insufficient number of statistical data points of long wavelength waves available to generate a spectrum that fits the analytical spectrum. Also, the data lacked the longer wavelength waves or swell observations. Therefore, we consider only the wave frequencies of 0.2 Hz and above in the present treatment. The high at low wind speeds indicates that the wave energy still retained in the ocean even after winds have subsided. This transition from calm to rough sea (and vice versa) depends on the rate of wave energy dissipation in the ocean (Figures 2(b) and 4). These observed wave conditions set the background for which we investigate the microwave signatures of ocean surface in the next section.
 
Note. UTC: Universal Time Coordinated. 
(a)
(b)
(c)
3.2. Comparison to Passive and Active Microwave Signatures
Table 3 shows the linear associations between geophysical variables (, PSD, water temperature, and ) and at 37 and 89 GHz as a correlation matrix. We observed a low probability value ( value ) for each combination of bolded variables given in Table 3, verifying statistical significance. Table 4 provides the regression coefficients for estimating the using and Figure 5 shows pertinent regression relationships. Wind speed does not correlate with ; however, we observe that at 37 GHz and 89 GHz shows significant correlation with the and the PSD of the waves and the correlation is stronger for Hpolarization than Vpolarization. Thus, the higher Hpolarization implies an indirect contribution of wind energy in inducing wave heights than Vpolarization. We did not observe any linear relationships between water surface temperature and . Besides this, several other external factors that alter the passive microwave emission of the ocean surface affect the .
 
Note: 89H: brightness temperature at 89 GHz Hpolarization (K). 89V: brightness temperature at 89 GHz Vpolarization (K). 37H: brightness temperature at 37 GHz Hpolarization (K). 37V: brightness temperature at 37 GHz Vpolarization (K). : significant wave height (cm). PSD: wave power spectral density (m^{2}/Hz). : water temperature (°C). : wind speed (m/s). Numerals in bold represent statistically significant correlation at the 95% confidence level. 

Although the depends on several other external factors (see Introduction), an increase in the surface area exposed to the sensor also contributes to a rise in . Higher waves have a larger surface area, and the Hpolarization especially detects the associated rise in . The passive microwave frequencies used in the study are sensitive to the individual contributions of smallscale features; for example, sea spray and whitecaps. However, at the observed Beaufort scale 0–5, sea sprays are not significant; thus, we assume the contribution of spray to modifying the as negligible. We observed an increase in Hpolarization at 37 and 89 GHz. This increase is consistent with the Hpolarization observations by previous researchers. increases with wind speed, foam, and whitecaps coverage at 8.36 and 19.34 GHz frequencies [9, 44] and between 13.4 and 37 GHz [45]. Other results illustrate a nonsignificant correlation between and wind speed at Vpolarizations of 6.6, 10.69, and 37 GHz frequencies [46]. In continuation of above studies, our results further confirm that the at Hpolarization is stronger than at Vpolarization at 37 and 89 GHz under rougher conditions.
Backscatter is compared to and considered in the context of surface roughness at the Cband scale. The interaction of incident radar waves with the capillary waves (millimeter level), whose restoring force is the surface tension of the water, is not accounted for, as the waves of this size are not detectable by the Waverider buoy. We compared the buoymeasured gravity waves, whose restoring force is gravity, with the backscatter. Kudryavtsev [47] has described the relationship between gravity (demonstrated as wave height) and capillary (demonstrated as Bragg scattering) waves. Figures 6(a) and 6(b) show the VV and HV backscatter at incidence angles 20°, 30°, and 40° for the three radar cases. The plots include backscatter intensities to provide insight into the relative contributions of gravity and capillary waves. We observed a decrease in backscatter with the increasing incidence angle, as expected, over the sampled range. The highest intensities correspond to the highest on July 1, though the weakest intensities correspond to the intermediate on July 9. Stronger HH (not shown), HV, and VV backscatter at all incidence angles for the lowest on July 26, relative to July 9, points to the likelihood of a stronger coupling of the radar waves to capillary waves, implying that higher capillary waves interaction with microwaves results in higher backscatter even at intermediate .
(a)
(b)
Figures 7(a) and 7(b) show the and for each of the analyzed incidence angles and roughness conditions. If we consider the roughness, interpreted using the HV and VV intensities in Figure 6, instead of the measured , the at 20° shows consistent behavior. The shows an inverse relation to the radarinterpreted roughness in a manner consistent with Bragg and Integral Equation Model (IEM) surface backscattering models [48]. The for given incidence angle and dielectric constant, according to Bragg and IEM theory, tends to zero (greater HH relative to VV) with increasing surface roughness provided that the surface roughness is greater than, or equal to the incident wavelength. The is independent of surface roughness for smooth surfaces with surface height very small relative to the incident wavelength [38]. The additional HH backscatter relative to VV is so strong for the roughest case on July 1 that the ratio at 20° becomes negative. For such a sea state, previous observations show greater HH backscatter sensitivity to the tilt modulation effect caused by larger waves on Bragg waves [23, 49], as well as wave breaking [50].
(a)
(b)
Our observations show correlation, which appears to be qualitatively proportionate, between and at 40° incidence angle; that is, and are much stronger on July 1 while July 9 and July 26 are similar. However, the results of Mouche et al. [24] show that one must consider azimuth angle (direction of waves relative to the radar look angle) at incidence angles larger than 25°. We have not considered it in the three cases evaluated in this study. In Figure 7(b), there is a significant increase in as a function of incidence angle and a small diversity between sea states. At 40° incidence angle, there is an increase in the ratio of 2 dB between the smooth and rough cases, and the ratio shows correlation with the . This effect is most likely because of the combined loss of HV backscatter with the incidence angle, which is also evident in Figure 6, coupled with an increased sensitivity of VV backscatter to the sea state. The HV intensity values are very low at incidence angles of 40° and above, in the context of utilizing from spaceborne SAR and one must consider it relative to the noise equivalent sigma zero of the utilized sensor at a particular incidence angle.
Figures 8(a)–8(c) provide polarimetric coherence parameters , , and at each of the incidence angles and roughness conditions. Figure 8(a) shows the , a measure of the correlation between complex HH and VH returns. There appears to be no relationship between and under the constraints imposed in this study. However, the in Figure 8(b) shows an inverse relationship between the magnitude of the coherence and at 20° incidence angle. This is consistent with a depolarization of incident polarized radar energy induced by the rougher ocean waves. Beyond 20°, there is no discernible relationship between and that one can deduce based on expected scattering mechanisms. Similar magnitudes of the correlation coefficient are evident for sea states on July 9 and 26. The rough sea state on July 1 (also seen at 20°) suggests that this parameter has utility for discriminating higher waves. One expects the parameter in Figure 8(c) to be decreasing with increasing surface roughness [34, 38, 51], though researchers developed it for terrestrial applications and did not verify for an ocean surface. At 20°, there is only a negligible decrease in associated with the highest case on July 1. The rougher sea state is linked with the highest at incidence angles greater than 20°. This is contrary to previous results over terrestrial surfaces. An important consideration, however, is the scale of physical roughness, which is much greater for the ocean waves under investigation in this study. Furthermore, in a study of Jet Propulsion Laboratory (JPL) polarimetric scatterometer observations off the California coast at 13.4 GHz, Lee et al. [52] found an azimuthal dependency on the , but did not specify the incidence angle. The study of by Keller et al. [22] (at Cband) mentioned above shows that one must consider the azimuth effect above 25° for this coherence. The Cband radar observations in this study, particularly , results in Figure 7(b), and the coherence results in Figures 8(a)–8(c), further support the need to consider ambiguity caused by the azimuth angle when considering the derivation of sea state from polarimetric radar data at incidence angles greater than 25°.
(a)
(b)
(c)
4. Conclusions
We presented, in this paper, passive microwave emission and polarimetric backscatter data coincident to wind and ocean buoy data at unique spatial and temporal scales relative to airborne and satellite remote sensing studies. A statistical analysis of coincident geophysical variables showed that the is more sensitive to than the wind speed at both H and Vpolarizations and that it shows stronger correlation with at Hpolarization. The PSD and show significant correlation with the Hpolarization of at both of the passive microwave frequencies (better with 37 GHz because of longer wavelength) in this study. An assessment of Cband backscatter points to a dependency of HV and VV intensities on surface roughness, which is resonant to the radar frequency, but not captured by the Waverider buoy, as shown by greater backscatter intensities for the smoother (47 cm compared to 66 cm) at each of the observed incidence angles. and at 40° are positively associated with , and particularly shows an increase that is proportional to .
The polarimetric coherence parameter at 20° also shows an inverse relationship with because of an expected decorrelation of complex returns with greater surface roughness. The also holds potential for discriminating rough ( cm) from smooth ( cm) sea states at 30° and 40°, based on an increase in the magnitude of with increasing . We did not observe a decrease in with increasing roughness, as expected for all incidence angles based on previous observations made over terrestrial surfaces. While the results presented here point to the utility of Cband satellite radar for deriving ocean surface physical roughness information without the need for wind speed or azimuth angle information at incidence angles less than approximately 25°, importantly they provide a stimulus for further investigations including the azimuth angle for observations above approximately 25°. With an increasing spatial and temporal coverage of open water in the Arctic, it will be important to develop robust satellite retrieval algorithms which examine both gas and energy exchange across the AO interface and into the marginal ice zones (MIZ) of the Arctic.
Conflict of Interests
The authors declare that there is no conflict of interests regarding the publication of this paper.
Acknowledgments
The authors are thankful to the crew of Canadian Research Icebreaker Amundsen for their cooperation in the acquisition of data. Mukesh Gupta thanks Debbie Armstrong and Natalie Asselin of University of Manitoba for help during buoy deployments. Randall K. Scharien would like to acknowledge Torsten Geldsetzer at the University of Calgary and Jim Mead at ProSensing, Inc. for their discussions and insights on the polarimetric scatterometer; also Dr. Klaus Hochheim and Natalie Asselin for their assistance in the field; and Tim Papakyriakou and Matthew Asplin for making available the meteorological data for this project. The Canadian International Polar Year (IPY) program office; the Natural Sciences and Engineering Research Council (NSERC); the Canada Research Chairs (CRC) Program; Canada Foundation for Innovation (CFI); and numerous international partner organizations which provided funding for the CFL project are acknowledged. The Canadian Northern Studies Trust (CNST); Maritime Awards Society of Canada Graduate Scholarships; and Margaret P. Hess Graduate Scholarship awards provided support to Randall K. Scharien. The authors dedicate this paper to the memory of their colleague Dr. Klaus Hochheim who lost his life while conducting sea ice research in the Canadian Arctic in September 2013.
References
 D. G. Barber and J. M. Hanesiak, “Meteorological forcing of sea ice concentrations in the southern Beaufort Sea over the period 1979 to 2000,” Journal of Geophysical Research C, vol. 109, no. 6, Article ID C06014, 2004. View at: Publisher Site  Google Scholar
 J. V. Lukovich and D. G. Barber, “On sea ice concentration anomaly coherence in the southern Beaufort Sea,” Geophysical Research Letters, vol. 32, no. 10, Article ID L10705, 2005. View at: Publisher Site  Google Scholar
 E. C. Carmack and R. W. Macdonald, “Oceanography of the Canadian shelf of the Beaufort Sea: a setting for marine life,” Arctic, vol. 55, no. 1, pp. 29–45, 2002. View at: Google Scholar
 W. O. Smith Jr. and D. G. Barber, Eds., Polynyas: Windows to the World, Elsevier, Amsterdam, The Netherlands, 2007.
 H. He and D. Chen, “Effects of surface wave breaking on the oceanic boundary layer,” Geophysical Research Letters, vol. 38, no. 7, Article ID L07604, 2011. View at: Publisher Site  Google Scholar
 I. N. Esau, “Amplification of turbulent exchange over wide Arctic leads: largeeddy simulation study,” Journal of Geophysical Research D, vol. 112, no. 8, Article ID D08109, 2007. View at: Publisher Site  Google Scholar
 K. St. Germain, G. A. Poe, and P. W. Gaiser, “Polarimetric emission model of the sea at microwave frequencies and comparison with measurements,” Progress in Electromagnetics Research, vol. 37, pp. 1–30, 2002. View at: Publisher Site  Google Scholar
 S. H. Yueh, “Modeling of wind direction signals in polarimetric sea surface brightness temperatures,” IEEE Transactions on Geoscience and Remote Sensing, vol. 35, no. 6, pp. 1400–1418, 1997. View at: Publisher Site  Google Scholar
 J. P. Hollinger, “Passive microwave measurements of sea surface roughness,” IEEE Transactions on Geoscience Electronics, vol. 9, no. 3, pp. 165–169, 1971. View at: Google Scholar
 F. T. Ulaby, R. K. Moore, and A. K. Fung, Microwave Remote Sensing: Active and Passive—Volume 3: Scattering and Emission Theory, Advanced Systems and Applications, Artech House, Dedham, Mass, USA, 1986.
 Y. A. Kravtsov and A. N. Churyumov, “Contribution of steep irregularities to the radiobrightness temperature of the ocean,” Radiophysics and Quantum Electronics, vol. 43, no. 3, pp. 196–201, 2000. View at: Publisher Site  Google Scholar
 E. C. Monahan and I. G. O'Muircheartaigh, “Whitecaps and the passive remote sensing of the ocean surface,” International Journal of Remote Sensing, vol. 7, no. 5, pp. 627–642, 1986. View at: Google Scholar
 A. E. Strong, “Mapping seasurface roughness using microwave radiometry,” Journal of Geophysical Research, vol. 76, no. 36, pp. 8641–8648, 1971. View at: Publisher Site  Google Scholar
 W. Nordberg, J. Conaway, D. B. Ross, and T. Wilheit, “Measurements of microwave emission from a foamcovered, winddriven sea,” Journal of Atmospheric Sciences, vol. 28, no. 3, pp. 429–435, 1971. View at: Publisher Site  Google Scholar
 O. M. Phillips, “Radar returns from the sea surface—Bragg scattering and breaking waves,” Journal of Physical Oceanography, vol. 18, no. 8, pp. 1065–1074, 1988. View at: Publisher Site  Google Scholar
 D. Vandemark, B. Chapron, J. Sun, G. H. Crescenti, and H. C. Graber, “Ocean wave slope observations using radar backscatter and laser altimeters,” Journal of Physical Oceanography, vol. 34, no. 12, pp. 2825–2842, 2004. View at: Google Scholar
 H. Hersbach, “Comparison of CBand scatterometer CMOD5.N equivalent neutral winds with ECMWF,” Journal of Atmospheric and Oceanic Technology, vol. 27, no. 4, pp. 721–736, 2010. View at: Publisher Site  Google Scholar
 H. Hersbach, A. Stoffelen, and S. de Haan, “An improved Cband scatterometer ocean geophysical model function: CMOD5,” Journal of Geophysical Research C, vol. 112, no. 3, Article ID C03006, 2007. View at: Publisher Site  Google Scholar
 A. S. Komarov, V. Zabeline, and D. G. Barber, “Ocean surface wind speed retrieval from Cband SAR images without input of wind direction,” IEEE Transactions on Geoscience and Remote Sensing, vol. 52, no. 2, pp. 980–990, 2014. View at: Publisher Site  Google Scholar
 A. Stoffelen and D. Anderson, “Scatterometer data interpretation: measurement space and inversion,” Journal of Atmospheric and Oceanic Technology, vol. 14, no. 6, pp. 1298–1313, 1997. View at: Google Scholar
 W. L. Jones and L. C. Schroeder, “Radar backscatter from the ocean: dependence on surface friction velocity,” BoundaryLayer Meteorology, vol. 13, no. 1–4, pp. 133–149, 1978. View at: Publisher Site  Google Scholar
 W. C. Keller, V. Wismann, and W. Alpers, “Towerbased measurements of the ocean C band radar backscattering cross section,” Journal of Geophysical Research, vol. 94, no. 1, pp. 924–930, 1989. View at: Google Scholar
 D. R. Thompson, T. M. Elfouhaily, and B. Chapron, “Polarization ratio for microwave backscattering from the ocean surface at low to moderate incidence angles,” in Proceedings of the IEEE International Geoscience and Remote Sensing Symposium (IGARSS '98), pp. 1671–1673, Seattle, Wash, USA, July 1998. View at: Google Scholar
 A. A. Mouche, D. Hauser, J. Daloze, and C. Guérin, “Dualpolarization measurements at Cband over the ocean: results from airborne radar observations and comparison with ENVISAT ASAR data,” IEEE Transactions on Geoscience and Remote Sensing, vol. 43, no. 4, pp. 753–769, 2005. View at: Publisher Site  Google Scholar
 V. Kudryavtsev, D. Hauser, G. Caudal, and B. Chapron, “A semiempirical model of the normalized radar crosssection of the sea surface 1. Background model,” Journal of Geophysical Research C, vol. 108, no. 3, pp. FET 21–FET 224, 2003. View at: Publisher Site  Google Scholar
 V. Kudryavtsev, D. Hauser, G. Caudal, and B. Chapron, “A semiempirical model of the normalized radar cross section of the sea surface, 2. Radar modulation transfer function,” Journal of Geophysical Research C, vol. 108, no. 3, pp. FET 31–FET 316, 2003. View at: Google Scholar
 O. M. Phillips, “Spectral and statistical properties of the equilibrium range in the windgenerated gravity waves,” Journal of Fluid Mechanics, vol. 156, pp. 505–531, 1985. View at: Google Scholar
 G. R. Valenzuela, “Theories for the interaction of electromagnetic and oceanic waves—a review,” BoundaryLayer Meteorology, vol. 13, no. 1–4, pp. 61–85, 1978. View at: Publisher Site  Google Scholar
 D. G. Barber, M. G. Asplin, Y. Gratton et al., “The international polar year (IPY) circumpolar flaw lead (CFL) system study: overview and the physical system,” AtmosphereOcean, vol. 48, no. 4, pp. 225–243, 2010. View at: Publisher Site  Google Scholar
 K. Hunkins, “Waves on the Arctic Ocean,” Journal of Geophysical Research, vol. 67, no. 6, pp. 2477–2489, 1962. View at: Publisher Site  Google Scholar
 L. D. Farmer, D. T. Eppler, and A. W. Lohanick, “Converting digital passive microwave radiances to Kelvin units of brightness temperatures,” Technical Note 427, Naval Ocean Research and Development Activity (NORDA), Bay St. Louis, Miss, USA, 1990. View at: Google Scholar
 D. Isleifson, B. Hwang, D. G. Barber, R. K. Scharien, and L. Shafai, “Cband polarimetric backscattering signatures of newly formed sea ice during fall freezeup,” IEEE Transactions on Geoscience and Remote Sensing, vol. 48, no. 8, pp. 3256–3267, 2010. View at: Publisher Site  Google Scholar
 T. Geldsetzer, J. B. Mead, J. J. Yackel, R. K. Scharien, and S. E. L. Howell, “Surfacebased polarimetric Cband scatterometer for field measurements of sea ice,” IEEE Transactions on Geoscience and Remote Sensing, vol. 45, no. 11, pp. 3405–3416, 2007. View at: Publisher Site  Google Scholar
 M. Gupta, R. K. Scharien, and D. G. Barber, “Cband polarimetric coherences and ratios for discriminating sea ice roughness,” International Journal of Oceanography, vol. 2013, Article ID 567182, 13 pages, 2013. View at: Publisher Site  Google Scholar
 J.S. Lee, D. L. Schuler, and T. L. Ainsworth, “Polarimetric SAR data compensation for terrain azimuth slope variation,” IEEE Transactions on Geoscience and Remote Sensing, vol. 38, no. 5, pp. 2153–2163, 2000. View at: Publisher Site  Google Scholar
 D. L. Schuler, J.S. Lee, D. Kasilingam, and G. Nesti, “Surface roughness and slope measurements using polarimetric SAR data,” IEEE Transactions on Geoscience and Remote Sensing, vol. 40, no. 3, pp. 687–698, 2002. View at: Publisher Site  Google Scholar
 H. J. Eom and W.M. Boerner, “Statistical properties of the phase difference between two orthogonally polarized SAR signals,” IEEE Transactions on Geoscience and Remote Sensing, vol. 29, no. 1, pp. 182–184, 1991. View at: Publisher Site  Google Scholar
 I. Hajnsek, E. Pottier, and S. R. Cloude, “Inversion of surface parameters from polarimetric SAR,” IEEE Transactions on Geoscience and Remote Sensing, vol. 41, no. 4, pp. 727–744, 2003. View at: Publisher Site  Google Scholar
 F. Mattia, T. L. Toan, J.C. Souyris et al., “The effect of surface roughness on multifrequency polarimetric sar data,” IEEE Transactions on Geoscience and Remote Sensing, vol. 35, no. 4, pp. 954–966, 1997. View at: Publisher Site  Google Scholar
 P. A. Hwang, M. A. Sletten, and J. V. Toporkov, “Analysis of radar sea return for breaking wave investigation,” Journal of Geophysical Research C, vol. 113, no. 2, Article ID C02003, 2008. View at: Publisher Site  Google Scholar
 N. C. Wells, The Atmosphere and Ocean, John Wiley & Sons, Oxford, UK, 2012.
 S. A. Kitaigorodskii, “Applications of the theory of similarity to the analysis of windgenerated wave motion as a stochastic process,” Bulletin of the Academy of Sciences, USSR Geophysics Series, vol. 1, pp. 105–117, 1962. View at: Google Scholar
 M. Donelan, M. Skafel, H. Graber, P. Liu, D. Schwab, and S. Venkatesh, “On the growth rate of windgenerated waves,” AtmosphereOcean, vol. 30, no. 3, pp. 457–478, 1992. View at: Google Scholar
 J. P. Hollinger, “Passive microwave measurements of the sea surface,” Journal of Geophysical Research, vol. 75, no. 27, pp. 5209–5213, 1970. View at: Google Scholar
 A. Stogryn, “The emissivity of sea foam at microwave frequencies,” Journal of Geophysical Research, vol. 77, no. 9, pp. 1658–1666, 1972. View at: Publisher Site  Google Scholar
 P. C. Pandey and R. K. Kakar, “An empirical microwave emissivity model for a foamcovered sea,” IEEE Journal of Oceanic Engineering, vol. 7, no. 3, pp. 135–140, 1982. View at: Google Scholar
 V. N. Kudryavtsev, “Physical model of the spectrum of capillarygravity waves,” Physical Oceanography, vol. 8, no. 2, pp. 69–80, 1997. View at: Google Scholar
 A. K. Fung and K. S. Chen, Microwave Scattering and Emission Models for Users, Artech House, Norwood, Mass, USA, 2010.
 D. B. Trizna and D. J. Carlson, “Studies of dual polarized low grazing angle radar sea scatter in nearshore regions,” IEEE Transactions on Geoscience and Remote Sensing, vol. 34, no. 3, pp. 747–757, 1996. View at: Publisher Site  Google Scholar
 H. Johnsen, G. Engen, and G. Guitton, “Seasurface polarization ratio from envisat ASAR AP data,” IEEE Transactions on Geoscience and Remote Sensing, vol. 46, no. 11, pp. 3637–3646, 2008. View at: Publisher Site  Google Scholar
 R. Touzi, W. M. Boerner, J. S. Lee, and E. Lueneburg, “A review of polarimetry in the context of synthetic aperture radar: concepts and information extraction,” Canadian Journal of Remote Sensing, vol. 30, no. 3, pp. 380–407, 2004. View at: Google Scholar
 J.S. Lee, S. H. Yueh, and D. L. Schuler, “Polarimetric analysis of scatterometer data for ocean surface wind measurement,” in Proceedings of the IEEE International Geoscience and Remote Sensing Symposium (IGARSS '04), vol. 1, pp. 20–24, Anchorage, Alaska, USA, September 2004. View at: Publisher Site  Google Scholar
Copyright
Copyright © 2014 Mukesh Gupta et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.