Frederick W. Chen

AIRS/AMSU/HSB precipitation estimates

Precipitation rates (mm per hour) with 15- and 50-km horizontal resolution are among the initial products of Atmospheric Infrared Sounder/Advanced Microwave Sounding Unit/Humidity Sounder for Brazil (AIRS/AMSU/HSB). They will help identify the meteorological state of the atmosphere and any AIRS soundings potentially contaminated by precipitation. These retrieval methods can also be applied to the AMSU 23-191-GHz data from operational weather satellites such as NOAA-15, -16, and -17. The global extension and calibration of these methods are subjects for future research. The precipitation-rate estimation method presented is based on the opaque-channel approach described by Staelin and Chen (2000), but it utilizes more channels (17) and training data and infers 54-GHz band radiance perturbations at 15-km resolution. The dynamic range now reaches 100 mm/h. The method utilizes neural networks trained using the National Weather Services Next Generation Weather Radar (NEXRAD) precipitation estimates for 38 coincident rainy orbits of NOAA-15 AMSU data obtained over the eastern United States and coastal waters during a full year. The rms discrepancies between AMSU and NEXRAD were evaluated for the following NEXRAD rain-rate categories: 32 mm/h. The rms discrepancies for the 3790 15-km pixels not used to train the estimator were 1.0, 2.0, 2.3, 2.7, 3.5, 6.9, 19.0, and 42.9 mm/h, respectively. The 50-km retrievals were computed by spatially filtering the 15-km retrievals. The rms discrepancies over the same categories for all 4709 50-km pixels flagged as potentially precipitating were 0.5, 0.9, 1.1, 1.8, 3.2, 6.6, 12.9, and 22.1 mm/h, respectively. Representative images of precipitation for tropical, mid-latitude, and snow conditions suggest the methods potential global applicability.

Precipitation observations near 54 and 183 GHz using the NOAA-15 satellite

Promising agreement over land and sea has been obtained between NEXRAD 3-GHz radar observations of precipitation rate and retrievals based on simultaneous passive observations at 50-191 GHz from the Advanced Microwave Sounding Unit (AMSU) on the NOAA-15 meteorological satellite. A neural network with three hidden nodes and one linear output node operated on 15 km resolution data at 183 ± 1 and 183 ± 7 GHz, plus the cosine of scan angle, to produce estimates that match well the morphology of NEXRAD hurricane and frontal precipitation data smoothed to 15-km resolution. A second neural network operated on the same three parameters used in the first network, but smoothed to 50-km resolution, plus spatially-filtered cold perturbations detected in three AMSU tropospheric temperature-sounding channels (channels 4-6), which also have 50-km resolution. Comparison with the same NEXRAD data smoothed to 50-km resolution yielded root mean square (rms) discrepancies for two frontal systems and two passes over Hurricane Georges of ∼1.1 mm/h, and ±1.4 dB for those precipitation events over 4 mm/h. Only 8.9% of the total AMSU-derived rainfall was in areas where AMSU saw more than 1- mm/h and NEXRAD saw less than 1-mm/h, and only 6.2% of the total NEXRAD-derived rainfall was in areas where NEXRAD saw more than 1-mm/h and AMSU saw less than 1-mm/h.

NPOESS Aircraft Sounder Testbed-Microwave (NAST-M): instrument description and initial flight results

The National Polar-Orbiting Operational Environmental Satellite System (NPOESS) Aircraft Sounder Testbed (NAST) has been developed and deployed on the NASA ER-2 high-altitude aircraft. The testbed consists of two co-located cross-track scanning instruments: a Fourier transform interferometer spectrometer (NAST-I) with spectral coverage of 3.7-15.5 /spl mu/m and a passive microwave spectrometer (NAST-M) with 17 channels near the oxygen absorption lines at 50-57 GHz and 118.75 GHz. The testbed provides the first coregistered imagery from high-resolution microwave and infrared sounders and will provide new data that will help (1) validate meteorological satellite environmental data record feasibility, (2) define future satellite instrument specifications, and (3) demonstrate operational issues in ground validation, data calibration, and retrievals of meteorological parameters. To help validate the performance and potential of NAST-M, imagery was collected from more than 20 overpasses of hurricanes Bonnie and Earl during the Convection and Moisture Experiment (CAMEX-3), Florida, boreal summer 1998. The warm core and convection morphology of Hurricane Bonnie (August, 1998) is clearly revealed both by aircraft-based microwave brightness temperature imagery and temperature retrievals within the eye. Radiance comparisons with the Advanced Microwave Sounding Unit on the NOAA-15 satellite and radiosonde observations yield root mean-squared agreements of approximately 1 K or less.

Neural Network Estimation of Atmospheric Profiles Using AIRS/IASI/AMSU Data in the Presence of Clouds

A novel statistical method for the retrieval of atmospheric temperature and water vapor profiles has been developed and evaluated with sounding data from the Atmospheric InfraRed Sounder (AIRS) and the Advanced Microwave Sounding Unit (AMSU) on the NASA Aqua satellite and the Infrared Atmospheric Sounding Interferometer (IASI) and AMSU on the EUMETSAT MetOp-A satellite. The present work focuses on the cloud impact on the AIRS and IASI radiances and explores the use of the stochastic cloud clearing methodology together with neural network estimation. A stand-alone statistical algorithm will be presented that operates directly on cloud-impacted AIRS/AMSU and IASI/AMSU data, with no need for a physical cloud clearing process. The performance of this method was evaluated using global (ascending and descending) EOS-Aqua orbits collocated with ECMWF fields for a variety of days throughout 2003, 2004, 2005, and 2006. Over 1,000,000 fields of regard (3×3 arrays of footprints) over ocean and land were used in the study. The method requires significantly less computation than traditional variational retrieval methods, while achieving comparable performance. Retrieval accuracy will be evaluated using ECMWF atmospheric fields as ground truth. The accuracy of the neural network retrieval method will be compared to the accuracy of the AIRS Level 2 (Version 5) retrieval method.

Archiving and distribution of 2-D geophysical data using image formats with lossless compression

Certain types of two-dimensional (2-D) numerical remote sensing data can be losslessly and compactly compressed for archiving and distribution using standardized image formats. One common method for archiving and distributing data involves compressing data files using file compression utilities such as gzip and bzip2, which are widely available on UNIX and Linux operating systems. GZIP-compressed files and bzip2-compressed files must first be uncompressed before they can be read by a scientific application (e.g., MATLAB, IDL). Data stored using an image format, on the other hand, can be read directly by a scientific application supporting that format and, therefore, can be stored in compressed form, saving disk space. Moreover, wide use of image formats by data providers and wide support by scientific applications can reduce the need for providers of geophysical data to develop and maintain software customized for each type of dataset and reduce the need for users to develop and maintain or download and install such software. This letter demonstrates the utility of standardized image formats for losslessly compressing, archiving, and distributing 2-D geophysical data by comparing them with the traditional file compression utilities gzip and bzip2 on several types of remote sensing data. The formats studied include TIFF, PNG, lossless JPEG, JPEG-LS, and JPEG2000. PNG and TIFF are widely supported. JPEG2000 and JPEG-LS could become widely supported in the future. It is demonstrated that when the appropriate image format is selected, the compression ratios can be comparable to or better than those resulting from the use of file compression utilities. In particular, PNG, JPEG-LS, and JPEG2000 show promise for the types of data studied.

Neural Network Characterization of Geophysical Processes With Circular Dependencies

This paper describes a method for training neural networks to learn circular dependencies. Variables with circular structure (e.g., time of day, day of year, and Earth location) appear in many different contexts within geoscience and remote sensing. Some common representations of circular variables (e.g., time of day in hours) can introduce discontinuities or topological distortions in estimation problems. They do not necessarily prevent a neural network from learning a relationship with circular dependencies. However, using topologically appropriate representations of circular variables can reduce the complexity necessary for a neural network to accurately learn such a relationship despite possibly increasing the number of inputs thereby reducing training times. In this paper, neural networks are trained to learn fictitious geophysical functions of time of day, of geolocation, and of time of year. In all three examples, using topologically appropriate representations of time and geolocation instead of conventional representations as inputs significantly reduced rms errors. Neural networks are also trained to learn the variations of air temperature observations with time of day and day of year, and one-month averages of sea surface temperature with geolocation. The improvement achieved by using topologically appropriate representations was limited by a natural random behavior in the data. However, there is a small but significant improvement in estimating the sea surface temperatures. Using the topologically appropriate representations of circular data when training the neural networks could be important in global Earth science remote sensing contexts, where significant diurnal, seasonal, or geographical variations exist. The studies presented also suggest the development of a more general framework for training neural networks that considers the topology of variables.

Combined Infrared and Microwave Retrievals of Atmospheric Profiles in the Presence of Clouds using Nonlinear Stochastic Methods: The SCENE Algorithm

A nonlinear stochastic method for the retrieval of atmospheric temperature and moisture profiles has been devel- oped and evaluated with sounding data from the Atmospheric InfraRed Sounder (AIRS) and the Advanced Microwave Sound- ing Unit (AMSU), and is presently being adapted for use with the NPOESS Cross-track Infrared Microwave Sounding Suite (CrIMSS) consisting of the Cross-track Infrared Sounder (CrIS) and the Advanced Technology Microwave Sounder (ATMS). The algorithm is implemented in three stages, motivating the name, SCENE (Stochastic Cloud clearing (1), followed by Eigenvector radiance compression and denoising, followed by Neural network Estimation). First, the infrared radiance perturbations due to clouds are estimated and corrected by combined processing of the infrared and microwave data. Second, a Projected Principal Components (PPC) transform (2) is used to reduce the dimen- sionality of and optimally extract geophysical profile information from the cloud-cleared infrared radiance data. Third, an artificial feedforward neural network is used to estimate the desired geophysical parameters from the projected principal components. This paper has two major components. First, details of the SCENE algorithm are discussed, including both the architectural implementation and parameter selection and optimization. Sec- ond, the performance of the SCENE algorithm is compared with that of the AIRS Level 2 algorithm (version 4.0.9) (3) currently being used for the Aqua mission. The performance of the SCENE algorithm was evaluated using global, ascending EOS-Aqua orbits colocated with ECMWF fore- casts (generated every three hours on a 0.5-degree lat/lon grid) for a variety of days throughout 2002 and 2003. Over 300,000 fields of regard (3x3 arrays of footprints) over ocean were used in the study. The RMS temperature and moisture profile retrieval errors for the SCENE algorithm were compared to those of the AIRS Level 2 algorithm, and the performance of the SCENE algorithm exceeded that of the AIRS Level 2 algorithm throughout most of the troposphere. The SCENE algorithm requires significantly less computation than traditional variational retrieval methods while achieving comparable performance, thus the algorithm is particularly suitable for quick-look retrieval generation for post- launch CrIMSS performance validation.

Neural network retrieval of atmospheric temperature and moisture profiles from AIRS/AMSU data in the presence of clouds

A nonlinear stochastic method for the retrieval of atmospheric temperature and moisture profiles has been developed and evaluated with sounding data from the Atmospheric InfraRed Sounder (AIRS) and the Advanced Microwave Sounding Unit (AMSU), and is presently being adapted for use with the NPOESS Cross-track Infrared Microwave Sounding Suite (CrIMSS) consisting of the hyperspectral Cross-track Infrared Sounder (CrIS) and the Advanced Technology Microwave Sounder (ATMS). The algorithm is implemented in three stages, motivating the name, SCENE (Stochastic Cloud clearing,1 followed by Eigenvector radiance compression and denoising, followed by Neural network Estimation). First, the infrared radiance perturbations due to clouds are estimated and corrected by combined processing of the infrared and microwave data. Second, a Projected Principal Components (PPC) transform2 is used to reduce the dimensionality of and optimally extract geophysical profile information from the cloud-cleared infrared radiance data. Third, an artificial feedforward neural network is used to estimate the desired geophysical parameters from the projected principal components. This paper has two major components. First, details of the SCENE algorithm are discussed, including both the architectural implementation and parameter selection and optimization. Second, the performance of the SCENE algorithm is compared with that of the AIRS Level 2 algorithm (version 4.0.9) 3 currently being used for the Aqua mission. The stochastic cloud-clearing algorithm estimates infrared radiances that would be observed in the absence of clouds. This algorithm examines 3×3 sets of nine AIRS fields of view, selects the clearest ones, and then in a series of simple linear and non-linear operations on both the infrared and microwave channels estimates a single cloud-cleared infrared spectrum for the 3×3 set. The algorithm is both trained and tested using global numerical weather analyses within 60 degrees of the equator. The analyses were generated by the European Center for Medium-range Weather Forecasting (ECMWF), and were converted to radiances using the SARTA v1.04 radiative transfer package. The PPC compression technique was used to reduce the infrared radiance dimensionality by a factor of 100, while retaining over 99.99% of the radiance variance that is correlated to the geophysical profiles. A feedforward neural network (NN) with a single hidden layer of approximately 3000 degrees of freedom was then used to estimate the atmospheric moisture and temperature profiles at approximately 60 levels from the surface to 20 km. The performance of the SCENE algorithm was evaluated using global, ascending EOS-Aqua orbits colocated with ECMWF forecasts (generated every three hours on a 0.5-degree lat/lon grid) for a variety of days throughout 2002 and 2003. Over 300,000 fields of regard (3×3 arrays of footprints) over ocean were used in the study. The RMS temperature and moisture profile retrieval errors for the SCENE algorithm were compared to those of the AIRS Level 2 algorithm, and the performance of the SCENE algorithm exceeded that of the AIRS Level 2 algorithm throughout most of the troposphere. The SCENE algorithm requires significantly less computation than traditional variational retrieval methods while achieving comparable performance, thus the algorithm is particularly suitable for quick-look retrieval generation for post-launch CrIMSS performance validation.

Combined Microwave and Hyperspectral Infrared Retrievals of Atmospheric Profiles in the Presence of Clouds Using Nonlinear Stochastic Methods

A novel retrieval technique using stochastic cloud clearing and neural network estimation has been developed and evaluated. Retrieval RMS accuracies and quality control estimates are similar to physical, iterated methods while requiring substantially less computation.

On-orbit radiometric validation and field-of-view calibration of spaceborne microwave sounding instruments

This paper outlines two calibration/validation efforts planned for current and future spaceborne microwave sounding instruments. First, the NPOESS Aircraft Sounder Testbed-Microwave (NAST-M) airborne sensor is used to directly validate the microwave radiometers (AMSU and MHS) on several operational satellites. Comparison results for underflights of the Aqua, NOAA, and MetOp-A satellites are shown. Second, a potential approach is presented for on-orbit field-of-view (FOV) calibration of the Advanced Technology Microwave Sounder (ATMS, to be launched in 2010). A constrained deconvolution technique is used to estimate spurious sidelobes in the ATMS antenna patterns from radiometric data collected while the sensor fields of view are scanned across the Earthpsilas limb.