SciELO - Scientific Electronic Library Online

 
vol.68 número2  suppl.TIIProcANALYSIS OF SPACE-TIME VARIABILITY OF THE PLATA RIVER PLUMEWIND DIRECTION SIGNAL IN POLARIZED MICROWAVE EMISSION OF SEA SURFACE UNDER VARIOUS INCIDENCE ANGLES índice de autoresíndice de materiabúsqueda de artículos
Home Pagelista alfabética de revistas  

Servicios Personalizados

Revista

Articulo

Indicadores

Links relacionados

Compartir


Gayana (Concepción)

versión impresa ISSN 0717-652Xversión On-line ISSN 0717-6538

Gayana (Concepc.) v.68 n.2 supl.TIIProc Concepción  2004

http://dx.doi.org/10.4067/S0717-65382004000300031 

  Gayana 68(2): 487-492, 2004

MULTI-ANGULAR MEASUREMENTS OF SEA-SURFACE POLARIZED MICROWAVE EMISSION CARRIED OUT DURING A SERIES OF GROUND-BASED EXPERIMENTS

 

Michael N. Pospelov & Alexey V. Kuzmin

Space Research Institute, 84/32 Profsoyuznaya St., Moscow 117997, Russia, Email: Michael.Pospelov@iki.rssi.ru


ABSTRACT

The results of sea-surface microwave brightness measurements carried out from a pier in 1999-2002 are presented. The measurements were performed by W-band microwave radiometer (wavelengths of 0.3 cm), and Ka-band radiometer-polarimeter (wavelength of 0.8 cm). All equipment was installed on a pier (200 meters long) at 3 meters over the surface; the sea depth at the site was approximately 8 meters. The microwave instruments were mounted on a rotating/scanning platform, which permitted measurements at angles from 10 up to 170 degrees off nadir over 300 degrees range of azimuthal angles. The measurements were carried out 24 hours a day under various meteorological conditions. Microwave brightness dependence on surface wind was studied over wide range of incidence angles. The modulus of brightness contrasts caused by the wind was shown to increase with wind speed increasing. Also, the brightness contrast at both polarizations changes its sign as the viewing angle increases. Under variable wind conditions, the delay of brightness variation relative to wind speed as large as 1-2 hours was found. This effect may be treated as an evidence of non-linear interaction of surface waves in coastal area.


INTRODUCTION

Global measurements of ocean surface winds are of vital importance for many oceanographic and atmospheric studies as well as for practical needs of navigation, weather forecasting, etc. Comparing with instruments on buoys and ships, only satellite-based sensors can provide global coverage in a reasonable time period. Additionally, sensors operating at microwave frequencies can measure the surface wind during nighttime and cloudy conditions, and therefore, greatly increasing the quantity of surface observations.

The paper presents the results of experimental studies of sea-surface thermal microwave emission carried out in the context of surface wind remote measurements. Recently, it has been shown that microwave radiometers are capable of measuring not only wind speed but wind direction as well. This ability comes out from the microwave brightness dependence on the azimuthal angle between the wind direction and the observation direction that first was discovered in aircraft experiments in late 1970s (Bespalova et al., 1981). Further experimental investigations have been demonstrated an importance of polarimetric measurements for wind vector retrieval (Bespalova et al., 1982; Dzura et al., 1992; Yueh et al., 1994). Measuring polarization parameters of microwave emission allowed surface wind vector retrieval even under unfavorable conditions (heavy clouds).

Until recently, experimental studies aimed at wind vector remote measurements have been carried out using aircraft radiometers and polarimeters (Kunkee and Gasiewski, 1994; Yueh et al., 1995; Kuzmin and Pospelov, 1999; Pospelov et al., 2000). The polarization parameters of sea-surface emission were measured at various view angles under very different weather conditions. These experimental results lend support to the validity of the proposed method of wind vector retrieval from passive polarimetric measurements. Last year, the first satellite polarimetric radiometer WindSat was launched with the object of global wind vector mapping using this method (Gaiser and St. Germain, 2000). However, one can hardly expect an one-to-one correspondence between the polarization parameters of microwave emission and near-surface wind to exist. The problem of waves and wind interaction under natural conditions is extremely complicated and calls for further experimental investigations. Aircraft measurements are very expensive and necessitate synchronous in situ measurements for validation. That is why we concentrate on platform measurements, which can provide a 247 regime with a reasonable support by accompanying meteorological and hydrological measurements.

EXPERIMENT DESCRIPTION

Microwave Remote Sensing Experiments were carried out on the Black Sea coast during summer months of 1999-2002 (MiRSEx'99 ­ MiRSEx'02). The experiments were performed at the South Department of the Shirshov Institute of Oceanology on the shore of Blue Bay near Gelendzhik, Russia.

All equipment was installed on a pier (200 meters long) at 3 meters above the surface; the sea depth at the site was approximately 8 meters. Remote sensing measurements were accompanied by concurrent measurements of meteorological and hydrological parameters. Wind speed and direction measurements were made by vane anemometer at a height of 7 meters. Air and water temperatures were registered at a height of 3 meters and at a depth of 1 meter respectively.

Microwave instruments were mounted on a rotating/scanning platform, which permitted measurements at incidence angles from 10 to 170 off nadir over 300 range of azimuthal angles. The platform was mounted on a 6-meter boom on the end of a pier, to reduce the disturbance of the wave pattern in antennas field of view by a pier construction and also to reduce a portion of the construction thermal emission reflected from waved sea surface. Vertical scanning was generally performed from 10 up to 170 off nadir at angular rate of 0.2 rpm, so at every scan sea-surface brightness temperature was measured at angles from near-nadir to horizon, and further the brightness temperature of atmosphere was measured at angles from horizon to near-zenith. The atmospheric part of the scan was used for microwave radiometers calibration using tipcurves as well as for calculating the atmosphere transparency at respective microwave bands to account for a portion of the atmosphere radiation reflected from a sea surface. The platform rotation was controlled by a computer, so any desirable algorithm of observations could be realised. Typical cycle of measurements endured about 30 minutes and consisted of 12 full vertical scans (upward and downward at 6 different azimuth angles stepped by 48 degrees) followed by a «black body» calibration. The measurements were carried out 24 hours a day.

The polarimetric measurements were performed with Ka-band radiometer-polarimeter. Combination of Dicke switching and Faraday rotation of polarization plane allowed measuring of the first three Stokes parameters of partially polarized microwave emission (Dzura et al., 1992). The polarimeter sensitivity was 0.15 K at 1 s integration time. Corrugated horn antenna had a half-power beamwidth of 9. Internal noise generator was used for monitoring gain factor stability. External calibration of V- and H-polarized channels was periodically performed using cold sky and microwave absorber at ambient temperature («black body») as two reference points; the black body temperature was measured by a thermistor. Special attention was paid on accurate measurements of the vertical view angle of the radiometer: built-in inclinometer provided accuracy within 10 angular minutes.

EXPERIMENTAL RESULTS

Analysis of the experimental data was performed for the cases when the wind blew from open sea, because in these cases wind field and waves were relatively uniform in the test area. Fig. 1 shows an example of Ka-band (37 GHz) brightness at vertical and horizontal polarization measured under different wind conditions. The data shown in the Fig. 1 were obtained by averaging over several successive scans, to reduce the fluctuations. View angles over 90 degrees pertain to the atmospheric portion of the scan. Since these measurements are related to two different days, the transparency of atmosphere was also different that is evident in the figure.


 
Figure 1. Angular dependence of microwave brightness at Ka-band (37 GHz) vertical (solid lines) and horizontal (dashed lines) polarization for wind speed 1.7 m/s (fine lines) and 6 m/s (bold lines). MiRSEx'99 Experiment, August 12 and 14, 1999.

To describe the microwave brightness dependence on the wind wave parameters more adequately, we will use hereafter the term «brightness contrast» for the difference between the measured brightness temperature and calculated brightness temperature of a flat water surface. The calculations were performed using Kirchhoff method for the same water salinity and sky brightness temperature as it was really observed at the same moment. In fact, the brightness contrast is due to wind waves only, if we neglect foam and spray effects that seems quite reasonable for wind speed below 8 m/s.

 In Fig. 2, the brightness contrasts at 94 GHz vertical polarization and 37 GHz vertical and horizontal polarizations are plotted versus observation angle for three mean wind speeds, from 1 up to 5.8 m/s. In this figure, the brightness temperature was averaged over all azimuth angles, so the azimuthal anisotropy of the microwave brightness is not accounted for. For the winds really observed during the MiRSEx'99 experiment, the azimuthal anisotropy did not exceed several tenths of a Kelvin.


 
 
 

Figure 2. Brightness contrast at 94 GHz vertical polarization (a), 37 GHz vertical (b) and horizontal (c) polarization, for averaged wind speed 1.0 m/s (solid line), 2.1 m/s (dotted line) and 5.8 m/s (dashed line).

Obviously, the modulus of brightness contrasts at both polarizations increases with wind speed increasing. Also, the brightness contrast at both polarizations changes its sign as the viewing angle increases. While at near-nadir and medium view angles the wind causes the brightness temperature to rise, at grazing angles it will result in the brightness temperature decreasing as large as 20 K for the wind speed 6.0 m/s.

Further, the brightness contrasts under variable wind conditions were investigated. As an example, we shall consider the data obtained on August 14, 1999. That day, the wind was directed from the open sea. Wind speed initially increased rapidly from 1.5 m/s up to 7 m/s, and then it gradually decreased down to values about 1 m/s.

In Fig. 3, the time dynamics of brightness contrast DTB at Ka-band (H-polarization) at various angles is compared to the wind speed. It is possible to note an interesting feature of the brightness contrast behaviour. While the wind speed decreased, the relevant decreasing of brightness contrast occurred with some delay. It is obvious from this plot that the maximum of brightness contrast is shifted approximately by one hour relative to wind speed maximum. For some angles, however, such a time delay is less.


 

Figure 3. Time history of Ka-band H-polarization brightness contrast under unstable wind speed: fine line - wind speed; bold lines and circles - brightness contrasts for observation angle (bottom to top) 20, 40, and 60 degrees; dashed lines and circles - brightness contrasts for observation angles (bottom to top) 70, 78 and 80 degrees. Reverse scale is related to the angles 70 to 80 degrees.

Such behaviour is even more evident in Fig. 4 where the brightness contrast is plotted vs. wind speed. Arrows in the plot indicate the beginning of the fragment shown in Fig. 3 (about 13:00 local time). Obvious hysteresis indicates the process of wave relaxation after wind speed decreasing. While the observation angle increases, the contribution from long waves (featuring long relaxation period) to brightness contrast increases as well.


 

Figure 4. Brightness contrasts at 37 GHz H-polarization vs. wind speed: observation angles (bottom to top) 80, 78, 20and 60 degrees. August 14, 1999.

DISCUSSION

Microwave brightness dependence on surface wind was studied over wide range of incidence angles. The modulus of brightness contrasts caused by the wind increases with wind speed increasing. Also, the brightness contrast at both polarizations changes its sign as the viewing angle increases. While at near-nadir and medium view angles the wind causes the brightness temperature to rise, at grazing angles it will result in the brightness temperature decreasing as large as 20 K for the wind speed 6.0 m/s. However, the dependence of brightness contrast on observation angle is quite different at vertical and horizontal polarization. At vertical polarization, the brightness contrast decreases smoothly as the angle increases, and it approaches zero at observation angles of 45-55 (that is the reason for the many of satellite radiometers to operate at the angle of 54). At horizontal polarization, the brightness contrast increases slightly as the observation angle increases up to 65-70, and then drops steeply to significant negative values at grazing angles. Therefor one may conclude the angles of 60-65 would be the best choice for the wind speed retrieval from microwave radiometry.

The task of wind speed retrieval is complicated by the fact that microwave brightness temperature dependence on wind speed is ambiguous, due to atmosphere stability (Pospelov, 1996) and wind history effects. Under variable wind conditions, the delay of the brightness response to changing wind may be as large as an hour or more. The detailed analysis of wave spectrum parameters (Kuzmin and Pospelov, 2004) permitted to relate the brightness delay observed in this experiment to short gravity-capillary surface waves, whereas longer gravity waves followed the wind practically without any delay. Since the relaxation period for short gravity-capillary waves can hardly exceed several minutes, the observed time lag may be associated with non-linear processes of the secondary ripples generation by inertial long waves.

ACKNOWLEDGMENTS

The authors should acknowledge the leading role in this study of Dr. Yuri Trokhimovski who passed away in 2002. He contributed a lot to the development of the microwave techniques for ocean and atmosphere remote sensing. A major part of the results presented in this paper was obtained by Yuri Trokhimovski in 2000-2001. Unfortunately, his premature death interrupted realisation of very promising ideas and plans.

The study was supported by RFBR (projects 00-05-64508 and 01-02-16538) and INTAS (projects 97-10569 and 03-51-4789).

 

REFERENCES

Bespalova, E.A., V.M. Veselov, A.A. Glotov, Y.A. Militskiy, V.G. Mirovskiy, I.V. Pokrovskaya, A.E. Popov, M.D. Raev, E.A. Sharkov, & V.S. Etkin, 1981, Sea-ripple anisotropy estimates from variations in polarized thermal emission of the sea, Oceanology, 21, 213-215 [Translated from Russian: Doklady Akademii Nauk SSSR, 246, No.6, 1482-1485, 1979].         [ Links ] [1]

Bespalova, E.A., V.M. Veselov, V.E. Gershenzon, Y.A. Militskiy, V.G. Mirovskiy, I.V. Pokrovskaya, M.D. Raev, A.G. Semin, N.K. Smirnov, V.A. Skachkov, Y.G. Trokhimovski, Y.B. Hapin, V.N. Chistyakov, E.A. Sharkov, & V.S. Etkin, 1982, Surface wind velocity determination from the measurements of the polarization anisotropy of microwave emission and backscatter, Soviet Journal of Remote Sensing, No.1, 121-131 [Translated from Russian: Issledovania Zemli iz Kosmosa, No.1, 87-94, 1982].         [ Links ] [2]

Dzura, M.S., V.S. Etkin, A.S. Khrupin, M.N. Pospelov, & M.D. Raev, 1992, Radiometers-polarimeters: Principles of design and applications for sea surface microwave emission polarimetry, International Geoscience and Remote Sensing Symposium (IGARSS'92) Proceedings, Houston, USA, 1432-1434.         [ Links ] [3]

Gaiser, P.W., & K.M. St. Germain, 2000, Spaceborne polarimetric microwave radiometry and the Coriolis WindSat system, IEEE Aerospace Conference Proceedings, 5, 159-164.         [ Links ] [4]

Kunkee, D.B. & A.J. Gasiewski, 1994, Airborne passive polarimetric measurements of sea surface anisotropy at 92 GHz, International Geoscience and Remote Sensing Symposium (IGARSS'94) Proceedings, Pasadena, USA, 2413-2415.         [ Links ] [5]

Kuzmin, A.V. & M.N. Pospelov, 1999, Measurements of sea surface temperature and wind vector by nadir airborne microwave instruments in joint United States/Russia internal waves remote sensing experiment JUSREX'92, IEEE Transactions on Geoscience and Remote Sensing, 37, No.4, 1907-1915.         [ Links ] [6]

Kuzmin, A.V. & M.N. Pospelov, 2004, Retrieval of gravity-capillary spectrum parameters by means of microwave radiometric techniques, IEEE Transactions on Geoscience and Remote Sensing, in press.         [ Links ] [7]

Pospelov, M.N., 1996, Surface wind speed retrieval using passive microwave polarimetry: the dependence on atmospheric stability, IEEE Transactions on Geoscience and Remote Sensing, 34, No.5, 1166-1171.         [ Links ] [8]

Pospelov, M.N., A.V. Kuzmin, & Y.G. Trokhimovski, 2000, Ocean surface temperature and wind vector measurements by airborne radiometers, The Fifth Pacific Ocean Remote Sensing Conference (PORSEC 2000) Proceedings, Goa, India, 1, 177-181.         [ Links ] [9]

Yueh, S.H., R. Kwok, F.K. Li, S.V. Nghiem, W.J. Wilson, & J.A. Kong, 1994, Polarimetric Passive Remote Sensing of Ocean Wind Vectors, Radio Science, 29, No.4, 799-814.         [ Links ] [10]

Yueh, S.H., W.J. Wilson, F.K. Li, S.V. Nghiem, & W.B. Ricketts, 1995, Polarimetric measurements of sea surface brightness temperature using an aircraft K-band radiometer, IEEE Transactions on Geoscience and Remote Sensing, 33, No.1, 85-92.         [ Links ] [11]

 

Creative Commons License Todo el contenido de esta revista, excepto dónde está identificado, está bajo una Licencia Creative Commons