We demonstrate that spatial maps of significant wave height (SWH) with kilometric resolutions can be derived from the data acquired by the Ka-band radar interferometer (KaRIn) instrument onboard the surface water ocean topography (SWOT) mission by exploiting the measured interferometric decorrelation. We discuss the sensitivity to errors in the volumetric decorrelation estimates and show that a successful inversion of SWH, particularly in the outer part of KaRIn’s swath and for low values of SWH, requires factoring out all sources of decorrelation of instrumental origin to an exquisite precision. We then validate KaRIn’s SWH measurement against independent data, namely, GPS buoys, airborne LiDAR, Sentinel3, SWOT’s nadir altimeter, and the ECMWF global wave model. We show that biases between KaRIn and the other sensors are centimetric and that KaRIn is able to capture features in the 2-D SWH field of only a few kilometers. While KaRIn’s SWH measurement error is difficult to fully characterize due to the absence of 2-D ground-truth data valid at such fine spatial scales and spanning a wide range of sea states, we argue that the retrieved fields are dominated by signal rather than noise, except possibly in the last few kilometers of the swath at low SWH. We briefly discuss the implications in terms of advancing our understanding of the phenomena that shape the wave fields at small scales. The algorithm and calibration described in this article will be the basis for version D of the operational SWOT products.

The NASA/JPL Uninhabited Aerial Vehicle Synthetic Aperture Radar (UAVSAR) instrument has performed tomographic SAR experiments over a number of study areas, including Rabi Forest in Gabon in 2016 and Sierra National Forest in California, USA in 2021. Tomographic SAR, or TomoSAR, is a technique enabling 3-D radar imaging with diverse applications including mapping of vegetation structure. Convolutional neural networks (CNNs) have shown widespread potential for many image processing and computer vision tasks such as image segmentation, classification, and object recognition. By using 3-D CNNs rather than 2-D CNNs, the filters can be applied to all three dimensions of a forest volume imaged by TomoSAR. We have trained 3-D CNN-based deep learning models to estimate canopy height and canopy cover from fully polarimetric UAVSAR TomoSAR images using lidar data as training and validation. When applied to canopy height estimation in the Rabi Forest study area, a trained network had root mean square error (RMSE) of 3.6 m (11%) compared to the validation dataset. For canopy cover estimation in the Sierra National Forest study area, the RMSE was 12%. Further work can be done to optimize the network architecture, improve the output spatial resolution, and to check if these methods can be applied to other study areas or to other vegetation structure parameters such as above-ground biomass. The results show the strong potential of 3-D CNNs for mapping wall-to-wall vegetation structure from tomographic SAR imagery using lidar training data.

The Surface Water and Ocean Topography (SWOT) mission is a collaboration between NASA and CNES that measures water extent, surface heights, and river slopes for inland water bodies and sea surface height (SSH), wind speed, and significant wave height (SWH) over open ocean. SWOT was launched on Dec. 15, 2022 and is currently operational. In this paper we discuss the algorithm for retrieving ocean surface wind speed from backscatter measurements obtained from the SWOT Ka-band Radar Interferometer (KaRIn). We validate that algorithm by comparing the retrieved wind speed to collocated measurements from the ASCAT ocean wind scatterometer onboard ESA’s MetOP-B and -C satellites.

The Surface Water and Ocean Topography (SWOT) mission was recommended by the 2007 National Research Council Decadal Survey to expand on previous altimetry missions like TOPEX/Poseidon. Utilizing wide-swath altimetry technology, SWOT aims to achieve complete coverage of the world’s oceans and freshwater bodies through high-resolution elevation measurements. SWOT received approval for implementation in 2016, it was ultimately launched in December 2022, and it is currently delivering preliminary data to the public. The primary instrument in SWOT is the Ka-band Radar Interferometer (KaRIn) which utilizes JPL-developed radar interferometry technology to measure ocean and surface water levels with unprecedented accuracy. This paper focuses on the challenges in designing, testing, and finally commissioning in flight a complex instrument like KaRIn. We also present preliminary flight performance and compare it with ground measurements and simulations. Our analysis indicates that KaRIn meets or exceeds all its requirements, but it has also revealed several interesting and unexpected observations, offering just a glimpse of future scientific discoveries that KaRIn will enable.

A spaceborne P-band signals of opportunity synthetic aperture radar concept is proposed for the remote sensing of terrestrial snow. We have completed a performance analysis assuming a formation flight of 3 to 5 SmallSats on one orbit plane. The spacing between the SmallSats is chosen so that their ground tracks will be separated by 50 to 100 m to allow the use of interferometric synthetic aperture radar processing technique to obtain a spatial resolution of a few hundred meters. A point system design has been completed to determine the antenna concept and to indicate the dependence of spatial resolution and signal to noise ratio on the number of receivers. The performance for range delay determination was analyzed to assess the impact of various error sources, including instrument receiver noise and ionospheric delay. The dominant error source is the ionospheric delay, which will be corrected using the split-spectrum algorithm. Our overall error budget analysis indicates that an accuracy of about 3 cm for the snow water equivalent in dry snow and 5 cm for the snow depth of wet snow can be achieved.

In this paper we discuss some observations of the Soil Moisture Active Passive (SMAP) mission's high-resolution synthetic aperture radar (SAR) for extreme winds and tropical cyclones. We find that the cross-polarized backscatter is far more sensitive to wind speed at extreme winds than the copolarized backscatter and it is essential to observations of extreme winds with L-band SAR. We introduce a cyclone wind speed retrieval algorithm and apply it to the limited SMAP SAR dataset of cyclones. We show that the SMAP SAR instrument is capable of detecting extreme winds up to the category 5 wind speed regime providing unique capabilities as compared to traditional scatterometer with C and Ku-band radars.

The paper discusses the hurricane intensity changes using machine learning technologies and data visualization.

The Soil Moisture Active Passive (SMAP) mission was launched January 31st, 2015. It is designed to measure the soil moisture over land using a combined active / passive L-band system. Due to the Aquarius mission, L-band model functions for ocean winds and salinity are already mature and may be directly applied to the SMAP mission. In contrast to Aquarius, the higher resolution and scanning geometry of SMAP allows for wide-swath ocean winds and salinities to be retrieved. We have found that the SMAP radiometer displays sensitivity all the way up to the most extreme wind speeds, possibly as high as 70 m/s, far beyond what is capable with typical C and Ku-band ocean wind scatterometers. In previous work we have validated the SMAP high wind speeds against Rapid Scatterometer (RapidScat) and Stepped Frequency Microwave Radiometer (SFMR). In this work we consider the size of the SMAP cyclones and compare them to the Automated Tropical Cyclone Forecasting (ATCF) system B-deck files and we update the SFMR analysis with an additional year of data to strengthen the previous conclusions.

DopplerScatt is a Ka-band pencil-beam Doppler scatterometer developed under NASA's ESTO Instrument Incubator Program (IIP) to serve as a demonstrator for future spaceborne instruments to measure ocean surface currents and winds simultaneously. We review the capabilities of the Doppler-Scatt instrument, and present results from multiple airborne campaigns. One of our primary results will be the development of Ka-band Geophysical Model Functions (GMFs) to translate backscatter and Doppler measurements into surface winds and currents.

The Soil Moisture Active Passive (SMAP) mission was launched on January 31, 2015. It is a combined L-band active/passive system envisioned for the measurement of soil moisture over land. In addition to the soil moisture measurement, the SMAP data readily permit retrievals of ocean surface winds and sea surface salinity. In the previous work, we have found that the SMAP radiometer displays sensitivity to ocean surface wind all the way up to the most extreme wind speeds, possibly as high as 70 m/s, far beyond what is capable with typical$C$- and$Ku$-bands ocean wind scatterometers. In this letter, we use the Rapid Scatterometer and stepped frequency microwave radiometer to further validate these SMAP radiometer-only high-wind speed retrievals. In addition, we consider the size of the retrieved high wind speeds, validating them with the Automated Tropical Cyclone Forecasting system B-deck files.

The Soil Moisture Active Passive (SMAP) mission was launched January 31st, 2015. It is designed to measure the soil moisture over land using a combined active / passive L-band system. Due to the Aquarius mission, L-band model functions for ocean winds and salinity are already mature and may be directly applied to the SMAP mission. In contrast to Aquarius, the higher resolution and scanning geometry of SMAP allows for wide-swath ocean winds and salinities to be retrieved. We have found that the SMAP radiometer displays sensitivity all the way up to the most extreme wind speeds, possibly as high as 70 m/s, far beyond what is capable with typical C and Ku-band ocean wind scatterometers. In this work we use the Rapid Scatterometer (RapidScat) and Stepped Frequency Microwave Radiometer (SFMR) to further validate these SMAP radiometer-only high-wind speed retrievals.

In this talk we introduce the combined active / passive (CAP) data product for the Soil Moisture Active Passive mission. We develop the algorithms for a radiometer-only salinity product, a radar-only vector wind product, and a combined active / passive vector wind and salinity product. We show the radiometer-only salinity product nears the Aquarius salinity accuracy requirements, that the radar-only vector wind product meets the QuikSCAT requirements, and that the combined active / passive salinity and vector wind product has performance better than both.

We investigated the use of L-band active and passive microwave data from the Soil Moisture Active Passive (SMAP) observatory for remote sensing of ocean surface winds during hurricanes. We analyzed the dependence of SMAP data on ocean surface wind speed and direction, and found excellent consistency with the geophysical model functions developed for the Aquarius L-band radar/radiometer although the spatial resolutions of SMAP and Aquarius are distinctly different. However the higher resolution data from SMAP allowed us to assess the sensitivity of L-band radiometer/radar signals to hurricane force winds. The matchup analysis with the data from typhoon Nangka confirms the feasibility of extrapolating the Aquarius model functions to very high winds. Therefore we applied the Aquarius model function to the retrieval of ocean winds for hurricanes for two options: 1) radiometer-only and 2) radar-only. Comparison of the SMAP winds with the RapidScat and National Center for Environmental Predictions (NCEP) wind was performed. We also compared the maximum wind speed in the SMAP products with the best track analysis and found a good agreement in general.

In this paper, we introduce the combined active/passive (CAP) data product for the Soil Moisture Active Passive mission. We develop the algorithms for a radiometer-only salinity product, a radar-only vector wind product, and a CAP vector wind and salinity product. We show that the performance of the radiometer-only salinity product nears but is still inferior to the Aquarius salinity accuracy performance when aggregated on a monthly timescale. Then, we show that the radar-only vector wind product has reasonable accuracy away from the nadir track while suffering from inadequate measurement geometry in the middle of the swath. Finally, we demonstrate that the CAP salinity and vector wind performance is superior to individual algorithms and provides wind vectors nearly as good as RapidScat for low-to-moderate winds and possibly superior to traditional scatterometers for wind speeds larger than 12.5 m/s.

The L-band passive microwave data from the Soil Moisture Active Passive (SMAP) observatory are investigated for remote sensing of ocean surface winds during severe storms. The surface winds of Joaquin derived from the real-time analysis of the Center for Advanced Data Assimilation and Predictability Techniques at Penn State support the linear extrapolation of the Aquarius and SMAP geophysical model functions (GMFs) to hurricane force winds. We apply the SMAP and Aquarius GMFs to the retrieval of ocean surface wind vectors from the SMAP radiometer data to take advantage of SMAP's two-look geometry. The SMAP radiometer winds are compared with the winds from other satellites and numerical weather models for validation. The root-mean-square difference (RMSD) with WindSat or Special Sensor Microwave Imager/Sounder is 1.7 m/s below 20-m/s wind speeds. The RMSD with the European Center for Medium-Range Weather Forecasts direction is 18° for wind speeds between 12 and 30 m/s. We find that the correlation is sufficiently high between the maximum wind speeds retrieved by SMAP with a 60-km resolution and the best track peak winds estimated by the National Hurricane Center and the Joint Typhoon Warning Center to allow them to be estimated by SMAP with a correlation coefficient of 0.8 and an underestimation by 8%-18% on average, which is likely due to the effects of spatial averaging. There is also a good agreement with the airborne Stepped-Frequency Radiometer wind speeds with an RMSD of 4.6 m/s for wind speeds in the range of 20-40 m/s.

Recent evidence suggests that the tropics have expanded over the last few decades by a very rough 10per decade. Until now, understanding the mechanisms of that expansion has been confined to models and proxies because of the unavailability of systematic observations of the large-scale circulation. Scatterometer-derived ocean surface vector winds, provide for the first time, an accurate depiction of the large-scale circulation and allow the study of the Hadley cell evolution through analysis of its surface branch. In this study we determine the extent of the Hadley cell as defined by the subtropical zero-crossing of the zonally-averaged zonal wind component. We use scatterometer observations from a number of missions, covering ~13 years. Our analyses reveal seasonal and interannual variability, as well as a long-term trend for expansion of the Hadley cell width. More interestingly, our results show an apparent discontinuity in the signal when the data source changes from one observing system to another. This raises the question about the significance of the unresolved diurnal signal. Indeed, analyses of observations from tandem missions support this notion. Fortunately, the RapidScat mission makes it possible to resolve, for the first time, the details of the diurnal signal. Our preliminary analyses of the RapidScat observations show the presence of a clear semidiurnal signal in the width of the Hadley cell. This helps explain previously found discrepancies. More importantly, this points to a clear need to understand and resolve the diurnal signal before merging wind observations from different missions to form a consistent climate record.

In this paper, we introduce a reprocessing of the entire SeaWinds on QuikSCAT mission. The goal of the reprocessing is to create a climate data record suitable for climate studies and to incorporate recent algorithm improvements. Three different levels of QuikSCAT data are produced at the Jet Propulsion Laboratory: L1B, geolocated, calibrated, backscatter measurements in chronological order by acquisition time; L2A, backscatter measurements binned into a geographical grid; and L2B, gridded ocean surface wind vectors. This reprocessing only changes the L2A and L2B data; we have not changed the L1B processing at all. We introduce new algorithms used in the L1B to L2A processing and in the L2A to L2B processing. After introducing our new algorithms, we show the validation studies performed to date, which include comparisons to numerical weather products, comparisons to buoy data sets, comparisons to other remote sensing instruments, and spectral considerations.

It has been well known that a few areas on the Earth's surface have a relatively constant backscattering coefficient and can serve as radar calibration targets. Examples include the Amazon rain forest, the Greenland ice sheet, parts of Antarctica, etc. However, there has not been any extensive investigation and quantitative evaluation of these targets in terms of their time variation, isotropy, and spatial variation. Here, we have analyzed a consistent set of Ku-band radar measurements for more than ten years from the QuikSCAT mission. This valuable set of data provides an unprecedented opportunity for us to study the long-term variability of observed backscattering from the Earth's surface at Ku-band. In this paper, we performed a global survey of potential constant land targets and evaluated their variability. Quantitative measurements of temporal and spatial variabilities (homogeneity) and isotropy are used to identify the locations of the best natural calibration targets. We also discuss annual and long-term trends in the data and offer potential explanations for these trends. We examine small regions with the least overall variation. By concentrating on low-variation areas, we can identify useful calibration targets for future radar missions and for intercalibration between existing radars. At the same time, by focusing on regions with little spatial or temporal heterogeneity, we can analyze the temporal variation on diurnal, seasonal, and decadal scales in homogenous natural terrain types including rain forest, dry brushy areas, and ice sheets. We found that rain forest targets in the Amazon and Congo are very stable in time and homogeneous. However, they are subjected to diurnal difference. On the other hand, the Antarctica ice sheet is another good candidate for stable target, but it has seasonal variability. The Greenland ice sheet shows a significant trend in backscatter in recent years, and therefore, may not be a suitable calibration site anymore. Another location for a good stable target is a dry brushy area in the Sahara, which shows comparable stability and isotropy with those of the Amazon, Congo, and Antarctica.

This paper presents the procedure to perform cross-calibration of radar backscatter between the QuikSCAT and Oceansat-2 ocean wind scatterometers. Both QuikSCAT and Oceansat-2 are Ku-band dual pencil beam, rotating antenna scatterometers with similar design. There has been a joint effort by the Indian Space Research Organization, NASA, KNMI, and NOAA to perform calibration and validation of Oceansat-2 in order to extend the climate data record of ocean surface vector winds obtained by QuikSCAT. This has resulted in significant improvement in the quality of the normalized radar cross section (NRCS) data and the quality of the resultant winds produced using the Oceansat-2 NRCS measurements. An important aspect of this calibration is the reduction of the calibration bias between QuikSCAT and Oceansat-2. The nonspinning QuikSCAT scatterometer was repointed to achieve the same incidence angles for its two HH and VV polarized antenna beams as those utilized by Oceansat-2. The magnitudes of the NRCS (backscatter) measurements of the two scatterometers were then compared for two years in order to determine NRCS bias in decibels as a function of time. Biases for both antenna beams were computed. A wind speed/wind-relative azimuth angle histogram-matched method was applied to ocean data from the two scatterometers to determine the time series of the bias between the two. It has been determined that there was an ~0.5 dB drop in Oceansat-2 radar backscatter on August 20, 2010. As a result, we compute cross-calibration adjustments to apply to Oceansat-2 data before and after this distinct drop in backscatter.

We have developed a neural network technique for retrieving accurate 12.5-km resolution wind speeds from Ku-band scatterometer measurements in tropical cyclone conditions including typical rain events in such storms. The method was shown to retrieve accurate wind speeds up to 40 m/s when compared with aircraft reconnaissance data, including GPS dropwindsondes and Stepped-Frequency Microwave Radiometer surface wind speed measurements, and when compared to global best track maximum wind speeds. Wind directions were unchanged from the current (version 3) Jet Propulsion Laboratory (JPL) global wind vector product. The technique removes positive biases with respect to best track winds in the developing phase of tropical cyclones that occurred in the nominal (version 2) JPL QuikSCAT product. The new technique also reduces negative biases with respect to best track wind speeds that occurred in the nominal product (both versions 2 and 3) during the most extreme period of the lifetime of intense storms. The wind regime with the most notable improvement is 20-40 m/s (40-80 kn), with more modest improvement for higher winds and the improvement at lower winds comparable to that achieved previously by the version 3 JPL global rain-corrected product. The net effect of all the wind speed improvements is a much better measurement of storm intensity over time in the new product than what has been previously available. When compared with speed data from aircraft flights in Atlantic hurricanes, the new product exhibited a 1-2-m/s positive overall bias and a 3-m/s mean absolute error. The random error and systematic positive bias in the new scatterometer wind product is similar to that of the Hurricane Research Division H*WIND analyses when aircraft data are available for assimilation. This similarity may be explained by the fact that H*WIND data are used as ground truth to fit the coefficients used by the new technique to map radar measurements to wind speed. The fact that H*WIND was designed to match maximum winds while preserving radial symmetry may explain the overall positive biases that we observe in both H*WIND and the new scatterometer wind product which compared to aircraft reconnaissance data. The new scatterometer product could also be inheriting systematic biases in the presence of rain from H*WIND. Under the most extreme rain conditions, the radar signal from the surface can be lost. In such cases, the technique makes use of measurements in the 87.5-km region comprising the 7 $\times$ 7 neighboring cells around the target 12.5-km wind vector cell. In so doing, we sacrifice resolution in cases where the highest resolution region has no useful measurements. Even so, the most extreme rain conditions can result in reduced accuracy. The new technique has been used to retrieve wind fields for every tropical cyclone of tropical storm force or above that has been observed by QuikSCAT during the period of time from October 1999 to November 2009. The resulting data set has been made available online for use by the tropical cyclone research community.

We exhibit a technique for improving wind accuracy in Ku-band ocean wind scatterometers in the presence of rain. The technique is autonomous in that it only makes use of measurements made by the scatterometer itself, so that no colocation of an external data set (e.g., rain radiometers) is required to perform the correction. The only inputs to the technique are the normalized radar cross-section measurements for each wind vector cell, the cross-track distance of the cell as a proxy for measurement geometry, and the nominal retrieved wind vector for the cell without rain correction. This last input is used to avoid modifying winds not contaminated by rain. The technique was applied to QuikSCAT data for the month of January 2008, resulting in a marked improvement to rainy data. For data that were determined to be rain contaminated by the Jet Propulsion Laboratory rain flag, the rms speed error with respect to National Data Buoy Center buoy winds improved from 8.9 to 3.5 m/s for colocations within 25 km. The rms speed error in rain also improved when compared with the European Centre Medium-Range Weather Forecast winds from 7 to 3 m/s. Data that were not flagged as rain contaminated were not significantly changed, despite the fact that the technique does not make use of the rain flag. The technique was able to distinguish between rain-contaminated wind cells and rain-free wind cells and to substantially improve the wind speed accuracy of the former using QuikSCAT data alone without recourse to any external information about the extent of the rain.

We describe a method for retrieving winds from colocated Ku- and C-band ocean wind scatterometers. The method utilizes an artificial neural network technique to optimize the weighting of the information from the two frequencies and to use the extra degrees of freedom to account for rain contamination in the measurements. A high-fidelity scatterometer simulation is used to evaluate the efficacy of the technique for retrieving hurricane force winds in the presence of heavy precipitation. Realistic hurricane wind and precipitation fields were simulated for three Atlantic hurricanes, Katrina and Rita in 2005 and Helene in 2006, using the Weather Research and Forecasting model. These fields were then input into a radar simulation previously used to evaluate the Extreme Ocean Vector Wind Mission dual-frequency scatterometer mission concept. The simulation produced high-resolution dual-frequency normalized radar cross-section (NRCS) measurements. The simulated NRCS measurements were binned into 5 x 5 km wind cells. Wind speeds in each cell were estimated using an artificial neural network technique. The method was shown to retrieve accurate winds up to 50 m/s even in intense rain.

The Cassini RADAR is a multimode instrument used to map the surface of Titan, the atmosphere of Saturn, the Saturn ring system, and to explore the properties of the icy satellites. Four different active mode bandwidths and a passive radiometer mode provide a wide range of flexibility in taking measurements. The scatterometer mode is used for real aperture imaging of Titan, high-altitude (around 20 000 km) synthetic aperture imaging of Titan and Iapetus, and long range (up to 700 000 km) detection of disk integrated albedos for satellites in the Saturn system. Two SAR modes are used for high- and medium-resolution (300-1000 m) imaging of Titan's surface during close flybys. A high-bandwidth altimeter mode is used for topographic profiling in selected areas with a range resolution of about 35 m. The passive radiometer mode is used to map emission from Titan, from Saturn's atmosphere, from the rings, and from the icy satellites. Repeated scans with differing polarizations using both active and passive data provide data that can usefully constrain models of surface composition and structure. The radar and radiometer receivers show very good stability, and calibration observations have provided an absolute calibration good to about 1.3 dB. Relative uncertainties within a pass and between passes can be even smaller. Data are currently being processed and delivered to the planetary data system at quarterly intervals one year after being acquired.

The SeaWinds scatterometer onboard QuikSCAT covers approximately 90% of the global ocean under clear and cloudy condition in 24 h, and the standard data product has 25-km spatial resolution. Such spatial resolution is not sufficient to resolve small-scale processes, especially in coastal oceans. Based on range-compressed normalized backscatter and a modified wind retrieval algorithm, a coastal wind dataset at 12.5-km resolution was produced. Even with larger error, the high-resolution winds, in medium to high strength, would still be useful over coastal ocean. Using measurements from moored buoys from the National Buoy Data Center, the high-resolution QuikSCAT wind data are found to have similar accuracy as standard data in the open ocean. The accuracy of both high- and standard-resolution winds, particularly in wind directions, is found to degrade near shore. The increase in error is likely caused by the inadequacy of the geophysical model function/ambiguity removal scheme in addressing coastal conditions and light winds situations. The modified algorithm helps to bring the directional accuracy of the high-resolution winds to the accuracy of the standard-resolution winds in near-shore regions, particularly in the nadir and far zones across the satellite track.

The use of QuikSCAT data for wind retrievls of tropical cyclones is described. The evidence of QuikSCAT /spl sigma//sub 0/ dependence on wind direction for >30 m/s wind speeds is presented. The QuikSCAT /spl sigma//sub 0/s show a peak-to-peak wind direction modulation of /spl sim/1 dB at 35 m/s wind speed, and the amplitude of modulation decreases with increasing wind speed. A correction of the QSCAT1 model function for above 23 m/s wind speed is proposed. We explored two microwave radiative transfer models to correct the effects of rain for wind retrievals. Both radiative transfer models have been used to retrieve the ocean wind vectors from the collocated QuikSCAT and SSM/I rain rate data for several tropical cyclones. The resulting wind speed estimates of these tropical cyclones show improved agreement with the wind fields derived from the best track analysis for up to about 15 mm/h SSM/I rain rate. A comparative analysis of maximum wind speed estimates suggests that other rain parameters likely have to be considered for further improvements.

The use of QuikSCAT data for wind retrievals of tropical cyclones is described. The evidence of QuikSCAT /spl sigma//sub 0/ dependence on wind direction for >30-m/s wind speeds is presented. The QuikSCAT /spl sigma//sub 0/s show a peak-to-peak wind direction modulation of /spl sim/1 dB at 35-m/s wind speed, and the amplitude of modulation decreases with increasing wind speed. The decreasing directional sensitivity to wind speed agrees well with the trend of QSCAT1 model function at near 20 m/s. A correction of the QSCAT1 model function for above 23-m/s wind speed is proposed. We explored two microwave radiative transfer models to correct the attenuation and scattering effects of rain for wind retrievals. One is derived from the collocated QuikSCAT and Special Sensor Microwave/Imager (SSM/I) dataset, and the other one is a published parametric model developed for rain radars. These two radiative transfer models account for the effects of volume scattering, scattering from rain-roughened surfaces and rain attenuation. The models suggest that the /spl sigma//sub 0/s of wind-roughened sea surfaces for 40-50-m/s winds are comparable to the /spl sigma//sub 0/s of rain contributions for up to about 10-15 mm/h. Both radiative transfer models have been used to retrieve the ocean wind vectors from the collocated QuikSCAT and SSM/I rain rate data for several tropical cyclones. The resulting wind speed estimates of these tropical cyclones show improved agreement with the wind fields derived from the best track analysis and Holland's model for up to about 15-mm/h SSM/I rain rate. A comparative analysis of maximum wind speed estimates suggests that other rain parameters likely have to be considered for further improvements.

The SeaWinds scatterometer was developed by NASA JPL, Pasadena, CA, to measure the speed and direction of ocean surface winds. It was then launched onboard the QuikSCAT spacecraft. The accuracy of the majority of the swath and the size of the swath are such that the SeaWinds on QuikSCAT Mission (QSCAT) meets its science requirements despite shortcomings at certain cross-track positions. Nonetheless, it is desirable to modify the baseline processing in order to improve the quality of the less accurate portions of the swath, in particular near the far swath and nadir. Two disparate problems have been identified for these regions. At far swath, ambiguity removal skill is degraded due to the absence of inner beam measurements, limited azimuth diversity and boundary effects. Near nadir, due to nonoptimal measurement geometry, (measurement azimuths approximately 180/spl deg/ apart) there is a marked decrease in directional accuracy even when ambiguity removal works correctly. Two algorithms have been developed: direction interval retrieval (DIR) to address the nadir performance issue and thresholded nudging (TN) to improve ambiguity removal at far swath. The authors illustrate the impact of the two techniques by exhibiting prelaunch simulation results and postlaunch statistical performance metrics with respect to ECMWF wind fields and buoy data.

The accuracy of Ku-band ocean wind scatterometers (i.e., NSCAT and SeaWinds) is impacted to varying degrees by rain. In order to determine how to best flag rain-contaminated wind vector cells and ultimately to calibrate out the effects of rain as much as possible, we must understand the impact of rain on the backscatter measurements that are used to retrieve wind vectors. This study uses collocated SSM/I rain rate measurements, NCEP wind fields, and SeaWinds on QuikSCAT backscatter measurements to empirically fit a simple theoretical model of the effect of rain on /spl sigma//sub 0/, and to check the validity of that model. The chief findings of the study are (1) horizontal polarization measurements are more sensitive to rain than vertical polarization, (2) sensitivity to rain varies dramatically with wind speed, and (3) the additional backscatter due to rain overshadows the rain-related attenuation.

The QuikSCAT radar measurements of several tropical cyclones in 1999 have been studied to develop the geophysical model function (GMF) of Ku-band radar /spl sigma//sub 0/ values (normalized radar cross section) for extreme high wind conditions. To account for the effects of precipitation, the authors analyze the co-located rain rates from the Special Sensor Microwave/Imager (SSM/I) and propose the rain rate as a parameter of the GMF. The analysis indicates the deficiency of the NSCAT2 GMF developed for the NASA scatterometer, which overestimates the ocean /spl sigma//sub 0/ for tropical cyclones and ignores the influence of rain. It is suggested that the QuikSCAT /spl sigma//sub 0/ is sensitive to the wind speed of up to about 40-50 m s/sup -1/. The authors introduce modifications to the NSCAT2 GMF and apply the modified GMF to the QuikSCAT observations of Hurricane Floyd. The QuikSCAT wind estimates for Hurricane Floyd in 1999 was improved with the maximum wind speed reaching above 60 m s/sup -1/. The authors perform an error analysis by comparing the QuikSCAT winds with the analyses fields from the National Oceanic and Atmospheric Administration (NOAA) Hurricane Research Division (HRD). The reasonable agreement between the improved QuikSCAT winds and the HRD analyses supports the applications of scatterometer wind retrievals for hurricanes.

Spaceborne wind scatterometers provide useful measurements of ocean surface winds and are important to climatological studies and operational weather forecasting. Past and currently planned scatterometers use measurements of the copolarized backscatter cross-section at different azimuth angles to infer ocean surface wind speed and direction. Although successful, current scatterometer designs have limitations such as degraded wind performance in the near-nadir and outer regions of the measurement swath and a reliance on external wind information for vector ambiguity removal. Theoretical studies of scattering from the wind-induced ocean surface indicate that polarimetric measurements provide orthogonal and complementary directional information to aid the wind retrieval process. In this paper, potential benefits of making polarimetric backscatter measurements to improve wind retrieval performance are addressed. To investigate the performance of a polarimetric scatterometer, a modified version of the SeaWinds end-to-end simulator at the Jet Propulsion Laboratory (JPL), Pasadena, CA, is employed. To model the effect of realistic measurement errors, expressions for polarimetric measurement variance and bias are derived. It is shown that a polarimetric scatterometer can be realized with straightforward and inexpensive modifications to a current scanning pencil-beam scatterometer system such as SeaWinds. Simulation results show that such a system ran improve wind performance in the nadir region and eliminate the reliance on external wind information.

This paper introduces a general structure that is capable of approximating input-output maps of nonlinear discrete-time systems. The structure is comprised of two stages, a dynamical stage followed by a memoryless nonlinear stage. A theorem is presented which gives a simple necessary and sufficient condition for a large set of structures of this form to be capable of modeling a wide class of nonlinear discrete time systems. In particular, we introduce the concept of a "complete memory". A structure with a complete memory dynamical stage and a sufficiently powerful memoryless stage is shown to be capable of approximating arbitrarily wide class of continuous, causal, time invariant, approximately-finite-memory mappings between discrete-time signal spaces. Furthermore, we show that any bounded-input bounded output, time-invariant, causal memory structure has such an approximation capability if and only if it is a complete memory. Several examples of linear and nonlinear complete memories are presented. The proposed complete memory structure provides a template for designing a wide variety of artificial neural networks for nonlinear spatiotemporal processing.

Based on the habituation mechanism found in biological neural systems, novel dynamic neural networks are proposed for recognizing temporal patterns. The specific task considered in this paper is the classification of whale songs from passive sonar data, but the networks are also readily applicable to other temporal pattern recognition problems. The fact that the networks designed operate dynamically is important, because it makes the goal of real time data analysis possible.