Microwave emission and reflection from the wind-roughened seasurface at 6.7 and 18.6 GHz

Microwave radiometric observations were made with specially designed microwave radiometers at 6.7 and 18.6 GHz, and the results were compared with those of other investigators, over the frequency range of 1-40 GHz. Dependences of sea surface emission and reflection on wind speed, frequency, incidence angle, and polarization type are discussed in detail, following discussions of the reflective processes of sky radiation and error estimation in the retrieval of mainlobe-averaged brightness temperature. The wind speed sensitivity of brightness temperature, emissivity, and reflectivity is formulated with respect to frequency and incidence angle in each polarization. The brightness temperature, emissivity and reflectivity at arbitrary wind speed are derived employing this formulation. Based on the results obtained it is suggested that the 10-19-GHz band may be optimal for satellite microwave radiometer observations of sea-surface wind     

An updated analysis of the ocean surface wind direction signal in passive microwave brightness temperatures

We analyze the wind direction signal for vertically (v) and horizontally (h) polarized microwave radiation at 37 GHz, 19 GHz, and 11 GHz; and an Earth incidence angle of 53/spl deg/. We use brightness temperatures from SSM/I and TMI and wind vectors from buoys and the QUIKSCAT scatterometer. The wind vectors are space and time collocated with the radiometer measurements. Water vapor, cloud water and sea surface temperature are obtained from independent measurements and are uncorrelated with the wind direction. We find a wind direction signal that is noticeably smaller at low and moderate wind speeds than a previous analysis had indicated. We attribute the discrepancy to errors in the atmospheric parameters that were present in the data set of the earlier study. We show that the polarization combination 2v-h is almost insensitive to atmospheric changes and agrees with the earlier results. The strength of our new signals agrees well with JPL aircraft radiometer measurements. It is significantly smaller than the prediction of the two-scale sea surface emission model for low and intermediate wind speeds.     

A Simplified Microwave Model and Its Application to the Determination of Some Oceanic Environmental Parametric Values, Using Multichannel Microwave Radiometric Observations

This paper proposes a simplified microwave model, linerized by Taylor expansion and incorporating observed fundamental microwave emissive properties, and discusses its application along with the least square method for determining some oceanic environmental parametric values from multichannel observation data. As a case study, microwave radiometric observations were made in vertical and horizontal polarizations at 6.7 and 18.6 GHz at incidence angles from 200 to 700 at steps of 100, and sea-surface temperature, wind speed, and sky brightness temperatures at the above dual frequencies at zenith angles from 200 to 700 were determined from a set of observed data at each incidence angle. The algorithm-derived results were compared with observed ones and showed good agreements within 10 percent of errors at incidence angles from 200 to 600; however, no good agreement was attained at large incidence angles such as 70°, except sky brightness temperature at 6.7 GHz. This disagreement may be attributed mainly to the difficulty and incompleteness in corrections such as estimation of reflected sky radiation and antenna characteristics, both for retrieval of the antenna temperature averaged over the main lobe and observation errors. This fact, however, may also suggest that observations at such larger incidence angles are of less utility.     

Efficient Model-Based Estimation of Atmospheric Transmittance and Ocean Wind Vectors From WindSat Data

A new method for estimating the atmospheric transmittance and wind speed over the ocean from WindSat data is derived using a simplified model for the ocean surface reflectivity. The simplified reflectivity model is used to calculate both the surface emissivity and the reflection of downwelling atmospheric radiation. The wind-speed dependence of the surface reflectivity is parameterized using simple rational functions with coefficients determined from the WindSat data. Because the vertically polarized brightness temperature depends primarily on the atmospheric state, it is used to obtain an initial estimate of the atmospheric transmittance at each spatial location. These estimates are then combined with the horizontally polarized brightness temperature to estimate the wind speed at each location. The first wind-speed estimate is used to refine the estimate of the transmittance, and the process is repeated until the estimates converge, resulting in a simultaneous solution for the atmospheric transmittance and the wind speed. The results are illustrated for two WindSat data sets collected on September 12 and 14, 2003. We have also investigated two methods of estimating wind direction using WindSat measurements of the third and fourth Stokes parameters. The first method involves an algebraic solution for the wind direction from simultaneous measurements of the third and fourth Stokes parameters. The second method involves measurements of the third Stokes parameter from two look directions (fore and aft scan angles), made possible by the conical scanning geometry of WindSat. A comparison and evaluation of these methods is made using the same data sets.      

一维综合孔径微波辐射计海温反演精度研究

星载一维综合孔径微波辐射计(1D-ASMR)利用天线阵代替传统辐射计的大天线,具有高空间分辨率 和多入射角观测特点,是未来海温遥感的重要载荷。为评估1D-ASMR 亮温测量误差EMb 对海温反演精度的影响以 及为卫星系统的设计提供依据,文中构建了1D-ASMR 亮温仿真模型,采用多元线性回归算法(Multiple Linear Regression, MLR),定量研究了1D-ASMR 海面温度反演误差随入射角和EMb 的变化规律。其中,该1D-ASMR 工作频点为 6. 9 GHz、10. 65 GHz、18. 7 GHz、23. 8 GHz 和36. 5 GHz,极化方式为水平极化和垂直极化,入射角范围为0°~65°,EMb 范围为0~1 K。结果表明:海面温度反演误差随入射角的变化趋势受EMb 的影响,大入射角对应的反演误差低于小 入射角。采用控制变量法,分别研究不同频点EMb 对海面温度反演的影响,可得:在同一入射角下,反演误差会随着 各个频点EMb 的增大而增大,但当EMb 的大小超过0. 5 K 时,反演误差随着EMb 的增大而增大的趋势会明显减缓 且6. 9 GHz 的EMb 对海温反演误差的影响最大。     

An emissivity-based wind vector retrieval algorithm for the WindSat polarimetric radiometer

The Naval Research Laboratory WindSat polarimetric radiometer was launched on January 6, 2003 and is the first fully polarimetric radiometer to be flown in space. WindSat has three fully polarimetric channels at 10.7, 18.7, and 37.0 GHz and vertically and horizontally polarized channels at 6.8 and 23.8 GHz. A first-generation wind vector retrieval algorithm for the WindSat polarimetric radiometer is developed in this study. An atmospheric clearing algorithm is used to estimate the surface emissivity from the measured WindSat brightness temperature at each channel. A specular correction factor is introduced in the radiative transfer equation to account for excess reflected atmospheric brightness, compared to the specular assumption, as a function wind speed. An empirical geophysical model function relating the surface emissivity to the wind vector is derived using coincident QuikSCAT scatterometer wind vector measurements. The confidence in the derived harmonics for the polarimetric channels is high and should be considered suitable to validate analytical surface scattering models for polarized ocean surface emission. The performance of the retrieval algorithm is assessed with comparisons to Global Data Assimilation System (GDAS) wind vector outputs. The root mean square (RMS) uncertainty of the closest wind direction ambiguity is less than 20/spl deg/ for wind speeds greater than 6 m/s and less than 15/spl deg/ at 10 m/s and greater. The retrieval skill, the percentage of retrievals in which the first-rank solution is the closest to the GDAS reference, is 75% at 7 m/s and 85% or higher above 10 m/s. The wind speed is retrieved with an RMS uncertainty of 1.5 m/s.     

Multifrequency Microwave Radiometer Measurements of Soil Moisture

Ground-based microwave radiometer experiments were performed to investigate the effects of moisture, temperature, and roughness on microwave emission from bare soils. Measurements were made at frequencies of 0.6-0.9, 1.4, and 10.7 GHz using van-mounted radiometers to observe prepared soil sites in Kern County, CA. The sites were instrumented for monitoring soil characteristics and surface meteorological conditions. Brightness temperature variations of approximately 15 K at 1.4 GHz and 25 K at 10.7 GHz were observed as a result of diurnal changes in the soil temperature. Increasing the soil moisture content from 2 to 15 percent by volume resulted in brightness temperature decreases of approximately 70 K at 0.775 and 1.4 GHz, and 40 K at 10.7 GHz, depending, to a lesser extent, on polarization and viewing angle. The results show the significance of soil temperature in deriving soil moisture from microwave radiometer measurements. Comparisons of the microwave measurements with theoretical predictions using a smooth surface model show reasonable agreement and support previous results of this nature obtained with other soil types. Approximately equal sensitivity to soil moisture was observed at 0.775 and 1.4 GHz, although the sampling depth is greater at the lower frequency.     

Variation of the Microwave Brightness Temperature of Sea Surfaces Covered with Mineral And Monomolecular Oil Films

Airborne microwave radiometer measurements over mineral and monomolecular oil films and adjacent clean sea surfaces are reported. An artificial crude-oil spill experiment in the New York Bight area showed a brightness temperature increase of the sea surface at 1.43 GHz as expected from a multilayered system with different dielectric constants. However, a monomolecular surface-film experiment with oleyl alcohol conducted in the North Sea during MARSEN in 1979 showed a strong brightness temperature depression at 1.43 GHz and no change in brightness temperature at 2.65 GHz. It is postulated that the monomolecular layer, because of its physical and chemical properties, polarized the underlying water molecules so strongly that the emissivity decreased from 0.31 to 0.016. It is estimated that the effective dielectric constant changed from 90 to 5.2 × 104. Because these phenomena occurred at 1.43 GHz it may be concluded that this frequency is very close to the center of a new anomalous dispersion region resulting from a restructuring of the water layer below the surface film.     

Classification of Baltic Sea ice types by airborne multifrequencymicrowave radiometer

An airborne multifrequency radiometer (24, 34, 48, and 94 GHz, vertical polarization) was used to investigate the behavior of the brightness temperature of different sea ice types in the Gulf of Bothnia (Baltic Sea). The measurements and the main results of the analysis are presented. The measurements were made in dry and wet conditions (air temperature above and below 0°C). The angle of incidence was 45° in all measurements. The following topics are evaluated: a) frequency dependency of the brightness temperature of different ice types, b) the capability of the multifrequency radiometer to classify ice types for winter navigation purposes, and c) the optimum measurement frequencies for mapping sea ice. The weather conditions had a significant impact on the radiometric signatures of some ice types (snow-covered compact pack ice and frost-covered new ice); the impact was the highest at 94 GHz. In all cases the overall classification accuracy was around 90% (the kappa coefficient was from 0.86 to 0.96) when the optimum channel combination (24/34 GHz and 94 GHz) was used     

The RADTRAN microwave surface emission models

The AFGL RADTRAN microwave atmospheric transmission/brightness temperature computer code has been enhanced by a surface emissivity modeling subpackage. This enhancement will provide realistic calculations of frequency-dependent polarized surface emissivity to support simulations of Earth-viewing microwave sensing systems. These models are appropriate for the frequency range 1-40 GHz. Two distinct modeling approaches have been applied: one based on wave theory for random discrete scatterers, and one based on radiative transfer theory for continuous random media. The former approach is used to model the ocean surface, various forms of sea ice, and snow over land, whereas the latter is used to treat soils and vegetation. Surface emissivity values obtained from these models as functions of polarization, frequency, and look angle are illustrated     

Polarimetric brightness temperatures of sea surfaces measured withaircraft K- and Ka-band radiometers

Dual-frequency (19 and 37 GHz), multi-incidence measurements of the Stokes parameters of sea surface microwave emission are reported. A series of aircraft polarimetric radiometer flights were carried out over the National Data Buoy Center (NDBC) moored buoys deployed off the northern California coast in July and August 1994. Measured radiometric temperatures showed a few Kelvin azimuth modulations in all Stokes parameters with respect to the wind direction. Wind directional signals observed in the 37-GHz channel were similar to those in the 19-GHz channel. This indicates that the wind direction signals in sea surface brightness temperatures have a weak frequency dependence in the range of 19-37 GHz. Harmonic coefficients of the wind direction signals were derived from experimental data versus incidence angle. It was found that the first harmonic coefficients, which are caused by the up and downwind asymmetric surface features, had a small increasing trend with the incidence angle. In contrast, the second harmonic coefficients, caused by the up and crosswind asymmetry, showed significant variations in T _v and U data, with a sign change when the incidence angle increased from 45° to 65°. Besides the first three Stokes parameters, the fourth Stokes parameter, V, which had never been measured before for sea surfaces, was measured using our 19-GHz channel. The Stokes parameter V. Has an odd symmetry just like that of the third Stokes parameter U, and increases with increasing incidence angles. In summary, sea surface features created by surface winds are anisotropic in azimuth direction and modulate all Stokes parameters of sea surface microwave brightness temperatures by as large as a few Kelvin in the range of incidence angles from 45° to 65° applicable to spaceborne observations     

Microwave emission of vegetation: sensitivity to leafcharacteristics

The effects of leaf characteristics on the microwave emission of land surfaces are analyzed. In order to simulate these effects, a radiative transfer model is presented. The medium consists of a vegetated layer containing randomly oriented leaves, modeled as elliptic-shaped scatterers, over the ground surface. Radiative transfer equations are solved with a discrete-ordinate-eigenanalysis method. The calculation of the phase matrix of the elliptic scatterers is based on the generalized Rayleigh-Gans approximation, which increases the frequency range of the modeling. The sensitivity of brightness temperature and polarization ratio to leaf characteristics, volume fraction, gravimetric moisture, size, shape, and inclination distribution is investigated at C-, and X-band. The behavior of the simulated emission of a soybean canopy versus frequency and incidence angle is studied for different soil moisture levels. Up to 10 GHz the microwave emission appears to contain significant information on underlying soil moisture     

Effects of foam on ocean surface microwave emission inferred from radiometric observations of reproducible breaking waves

WindSat, the first satellite polarimetric microwave radiometer, and the NPOESS Conical Microwave Imager/Sounder both have as a key objective the retrieval of the ocean surface wind vector from radiometric brightness temperatures. Available observations and models to date show that the wind direction signal is only 1-3 K peak-to-peak at 19 and 37 GHz, much smaller than the wind speed signal. In order to obtain sufficient accuracy for reliable wind direction retrieval, uncertainties in geophysical modeling of the sea surface emission on the order of 0.2 K need to be removed. The surface roughness spectrum has been addressed by many studies, but the azimuthal signature of the microwave emission from breaking waves and foam has not been adequately addressed. Recently, a number of experiments have been conducted to quantify the increase in sea surface microwave emission due to foam. Measurements from the Floating Instrumentation Platform indicated that the increase in ocean surface emission due to breaking waves may depend on the incidence and azimuth angles of observation. The need to quantify this dependence motivated systematic measurement of the microwave emission from reproducible breaking waves as a function of incidence and azimuth angles. A number of empirical parameterizations of whitecap coverage with wind speed were used to estimate the increase in brightness temperatures measured by a satellite microwave radiometer due to wave breaking in the field of view. These results provide the first empirically based parameterization with wind speed of the effect of breaking waves and foam on satellite brightness temperatures at 10.8, 19, and 37 GHz.     

Comparison of WindSat brightness temperatures with two-scale model predictions

Predictions of the polarized microwave brightness temperatures over the ocean are made using a two-scale surface bidirectional reflectance model combined with an atmospheric radiative transfer model. The reflected atmospheric radiation is found to contribute significantly to the magnitude and directional dependence of the brightness temperatures. The predicted brightness temperatures are also sensitive to the form of the shortwave spectrum. Calculations are made using a new physically based model for the wave spectrum, and preliminary comparisons are made with WindSat observations at 10.7, 18.7, and 37 GHz, for wind speeds ranging from 0-20 m/s and for vertically integrated atmospheric water vapor concentrations from 0-70 mm. Predictions of the mean (azimuthally averaged) brightness temperatures for vertical and horizontal polarization agree quite well with WindSat observations over this range of wind speeds and water vapor concentrations. The predicted azimuthal variations of the third and fourth Stokes parameters also agree fairly well with the observations, except for the fourth Stokes parameter at 37 GHz. Further adjustments of the wave spectrum are expected to improve the agreement.     

基于支持向量机的粗糙海面风速及海表盐度反演研究

将支持向量机(Support Vector Machine, SVM)回归技术应用到海况参数(如海表盐度、海面风速等)反演研究.利用双尺度模型(Two-Scale Model, TSM)作为前向电磁算法, 数值模拟不同雷达参数下风驱粗糙海面微波后向散射系数, 经过敏感性分析, 选取L波段(1.4 GHz)、C波段(6.8 GHz)及其合适的入射角作为雷达参数, 并设计多种反演方案, 分别以单频率双极化双角度、双频率双极化双角度及双极化后向散射系数的比值作为SVM的训练样本数据信息, 经过适当的训练, 利用SVM回归技术对海洋表面风速和盐度进行了反演研究.研究结果表明, 针对于海面风速的反演, C波段的反演精度最高, 针对于海表盐度的反演, L波段同极化散射系数比值作为SVM输入的反演精度较高.最后, 检验了SVM反演方法的抗噪声性能, 表明文中提出的SVM方法能较好地应用于实际海况参数反演问题.     

Radiometric observations of sea temperature at 2.65 GHz over the Chesapeake Bay

The present work describes the various corrections necessary in order to deduce ocean surface temperature fromS-band microwave radiometer measurements and applies these results to a series of data obtained with a high absolute accuracy radiometer. Measurements made with a 2.65 GHz radiometer from an aircraft flown over the Chesapeake Bay area are presented and compared in detail with accurately obtained sea truth data. For the calm sea, it was found that the observed brightness temperature agreed well with that calculated from the known sea surface and atmospheric properties over a fairly wide range of surface salinity values (0.2 per mille to 25 per mille). For cases where the surface wind speeds are of the order of 7 to 15 knots, an excess brightness temperature was observed which is attributable to surface roughness and microscale surface disturbances. The excess brightness temperature dependence on wind speed was found to correlate to a certain extent with the rms wave slope dependence on wind speed.     

Calculations of the Microwave Brightness Temperature of Rough Soil Surfaces: Bare Field

A model for simulating the remotely sensed microwave brightness temperatures of soils with rough surfaces is developed. The surface emissivity of the soil media is calculated from one minus its reflectivity, which is obtained by the integration of the bistatic scattering coefficients for rough soil surfaces. The soil brightness temperature is obtained from the product of the surface emissivity and the effective soil temperature, which is calculated with measured soil moisture profiles and soil temperature profiles at various soil depths. The roughness of a soil surface is characterized by two parameters, the surface height standard deviation a and its horizontal correlation length l. The model calculations are compared to the measured angular variations of the polarized brightness temperatures at both L-band (1.4 GHz) and C-band (5 GHz) frequencies. A nonlinear least squares fitting method is used to match the model calculations with the data, and the best fit results produce the parameter values of a and l that best characterize the surface roughness. The effect of rough surface shadowing is also incorporated into the model by introducing a shadowing function S(?), which represents the probability that a point on a rough surface is not shadowed by other parts of the surface. The model results for horizontal polarization are in excellent agreement with the data, both qualitatively and quantitatively. For vertical polarization, some discrepancies exist between the calculations and data. Possible causes of the discrepancy are discussed.     

Effects of Vegetation Cover on the Microwave Radiometric Sensitivity to Soil Moisture

The reduction in sensitivity of the microwave brightness temperature to soil moisture content due to vegetation cover is analyzed using airborne observations made at 1.4 and 5 GHz. The data were acquired during six flights in 1978 over a test site near Colby, Kansas. The test site consisted of bare soil, wheat stubble, and fully mature corn fields. The results for corn indicate that the radiometric sensitivity to soil moisture S decreases in magnitude with increasing frequency and with increasing angle of incidence (relative to nadir).The sensitivity reduction factor, defined in terms of the radiometric sensitivities for bare soil and canopy-covered conditions Y=1 - Scan/ Ss was found to be equal to 0.65 for normal incidence at 1.4 GHz, and increases to 0.89 at 5 GHz. These results confirm previous conclusions that the presence of vegetation cover may pose a serious problem for soil moisture detection with passive microwave sensors.     

Microwave Emission from Row Crops

In order to examine the emission properties of vegetation without taking into consideration the effects of variations in the soil background, strips of metal screening were used to cover the soil surface between adjacent rows of plants. Temporal measurements were made at 2.7 and 5.1 GHz for soybean, wheat, and corn canopies. Several special experiments were conducted to evaluate the sensitivity of brightness temperature to look direction (relative to row direction), polarization configuration, and incidence angle, and to evaluate the emission contributions of defoliated stalks. In general, the results show that the canopy is highly anisotropic, the emission exhibits a strong dependence on polarization and look direction, and the scattering albedo is typically less than 0.1. Canopy transmissivity was estimated from the radiometric observations and then related empirically to the canopy's integrated water content. Using this relation in a zero-order radiative transfer model led to good agreement between the experimental observations and the model predictions.