An Artificial-Neural-Network-Based Integrated Regional Model for Rain Retrieval Over Land and Ocean

An integrated regional model is proposed for rain-rate retrievals over land/ocean from the brightness temperature (Tb) values of the Tropical Rainfall Measuring Mission (TRMM) Microwave Imager (TMI). The polarization-corrected temperature calculated from the 85.5-GHz channels is also considered as one of the inputs along with the nine channel Tb values. This model is applicable over the region between and . For this purpose, an artificial neural network is utilized. The collocated precipitation radar (PR) near-surface rain rates as given by a 2A25 data product is considered as a target value. The methodology consists of the separation of land and ocean pixels, the separation of stratiform and convective pixels over land/ocean, and the selection of important features (inputs) for the multilayer perceptron network by the feature selection technique for each group. For the separation of land/ocean pixels, the Tb values of the 10.65-GHz vertical channel are utilized. The values are utilized to separate the stratiform and convective pixels both over land and ocean. The rain retrieval from the developed model is validated with TRMM PR. Overall result shows the better agreement of the model-retrieved rain rate with the PR observation compared to the TMI (2A12) rain rate particularly over land. The rain retrieved from the developed model is further validated with Doppler weather radar. A reasonably good agreement is observed between these two estimations.     

The beamfilling algorithm for retrieval of hydrometeor profileparameters from passive microwave measurements

The horizontal inhomogeneity of the atmosphere within a satellite microwave radiometer's field of view (FOV) has always been considered as a source of rainfall retrieval errors. The hydrometeor profile retrieval algorithm presented exploits it to obtain an approximation of a radiative transfer model, which allows relatively simple inversion. The atmosphere within the FOV is treated as a combination of horizontally homogeneous domains. Assuming that one of known “basic” hydrometeor profiles occurs in each domain, the inverse problem is reduced to a determination of “beamfilling coefficients.” The online procedure includes determination of beamfilling coefficients and a footprint-averaged hydrometeor profile as a linear combination of “basic” ones. Off-line procedures involve the selection of a minimum number of necessary “basic” brightness temperature vectors and correction of “basic” hydrometeor profiles to provide the best retrieval accuracy for a given cloud/radiative simulation. The performance of the algorithm is tested for both numerical simulations and TRMM/TMI data. Numerical simulation has allowed a comparison of the information content of radiometer measurements from SSM/I, TMI, and the future AMSR. The effectiveness of the algorithm is being tested for rain water integral and rain rate retrievals from TRMM TMI measurements     

Combining Artificial Neural Network Models, Geostatistics, and Passive Microwave Data for Snow Water Equivalent Retrieval and Mapping

A new modeling framework combining neural-network-based models, passive microwave data, and geostatistics is proposed for snow water equivalent (SWE) retrieval and mapping. Brightness temperature data from the seven-channel special sensor microwave/imager and the interpolated minimum temperature are the inputs of a multilayer feedforward neural network (MFF). Kriging with an external drift algorithm is applied to ground-based SWE data to produce gridded SWE data that are used as the target of the neural network. An optimal division of the sample of available pixels is achieved by a self-organizing feature map. Prediction error is used for model selection and is assessed by bootstrap. It is shown that a committee of a network containing neural networks with different architectures can provide consistent SWE retrievals. This modeling framework is applied for SWE retrieval and mapping over La Grande River basin in north eastern Quebec (Canada). The results are very promising for operational purposes particularly for SWE mapping during periods with no ground measurements and operational streamflow forecasting.     

Nonlinear statistical retrievals of ice content and rain rate frompassive microwave observations of a simulated convective storm

A numerical simulator for analysis of multispectral passive microwave mapping and retrieval is described. This simulator allows evaluation and optimization of satellite-based cloud and precipitation parameter retrieval algorithms. It contains three major components: the forward radiative transfer model, the sensor observation model, and the parameter retrieval algorithm. Simulated spaceborne observations of an oceanic tropical squall sampled at five stages in time are demonstrated for a simplified version of the proposed Earth Observation System (EOS) Multifrequency Imaging Microwave Radiometer (MIMR). The simulator uses a nonlinear statistical retrieval algorithm consisting of a Karhunen-Loeve (KL) transform, a projection operator, a nonlinear inverse mapping and a linear minimum mean-square error estimator. Retrievals of rain rate and integrated ice content are performed for each evolutionary frame at both full spatial resolution (1.5 km) and the degraded spatial resolution of a MIMR-class system. Results are presented for both KL-based and brightness temperature-based retrieval algorithms. It is found that the KL-based algorithm has a reduced complexity and performs better than the brightness temperature-based algorithm for degraded resolution imagery, especially for rain rate retrievals. In addition, rain rate retrievals are more affected by low image resolution than are integrated ice content retrievals. Retrieval accuracy of both rain and integrated ice is also found to depend on the evolutionary stage of the storm     

A soft computing approach for rainfall retrieval from the TRMM microwave imager

A neural network model for rainfall retrieval over ocean from remotely sensed microwave (MW) brightness temperature (BT) is proposed. BT data are obtained from the Tropical Rainfall Measuring Mission (TRMM) Microwave Imager (TMI). The BT values from different channels of TMI over the Pacific Ocean (163/spl deg/ to 177/spl deg/W and 18/spl deg/ to 34/spl deg/S) are the input features. The near-surface rainfall rate from the Precipitation Radar (PR) are considered as a target. The proposed model consists of a neural network with online feature selection (FS) and clustering techniques. A K-means clustering algorithm is applied to cluster the selected features. Different networks have been trained to give an instantaneous rainfall rate with all input features as well as with selected features obtained by applying the FS algorithm. It is found that the hybrid network utilizing FS and clustering techniques performs better. The developed network is also validated with two independent datasets on March 14, 2000 over the Atlantic Ocean having stratiform rain and on March 21, 2000 over the Pacific Ocean having both stratiform and convective rain. In both cases, the hybrid network performs well with correlation coefficient improving to 0.78 and 0.81, respectively, in contrast to 0.70 and 0.75 for the network with all features. The rainfall rate retrieved from the hybrid network is also compared with the TMI surface rain rate, and a correlation of 0.84 and 0.75 is found for the two events. The proposed hybrid model is validated with a Doppler Weather Radar, and correlation of 0.52 is observed.     

Characterizing the dependence of vegetation model parameters on crop structure, incidence angle, and polarization at L-band

To retrieve soil moisture over vegetation-covered areas from microwave radiometry, it is necessary to account for vegetation effects. At L-band, many retrieval approaches are based on a simple model that relies on two vegetation parameters: the optical depth (/spl tau/) and the single-scattering albedo (/spl omega/). When the retrievals are based on multiconfiguration measurements, it is necessary to take into account the dependence of /spl tau/ and /spl omega/ on the system configuration, in terms of incidence angle and polarization. In this paper, this dependence was investigated for several crop types (corn, soybean, wheat, grass, and alfalfa) based on L-band experimental datasets. The results should be useful for developing more accurate forward modeling and retrieval methods over mixed pixels including a variety of vegetation types.     

Sensitivity analyses of satellite rainfall retrieval and sampling error on flood prediction uncertainty

The Global Precipitation Measurement mission planned jointly by the United States, Japanese, and European space agencies envisions providing global rainfall products from a constellation of passive microwave (PM) satellite sensors at time scales ranging from 3-6 h. In this paper, a sensitivity analysis was carried out to understand the implication of satellite PM rainfall retrieval and sampling errors on flood prediction uncertainty for medium-sized (/spl sim/100 km/sup 2/) watersheds. The 3-h rainfall sampling gave comparable flood prediction uncertainties with respect to the hourly sampling, typically used in runoff modeling, for a major flood event in Northern Italy. The runoff prediction error, though, was magnified up to a factor of 3 when rainfall estimates were derived from 6-h PM sampling intervals. The systematic and random error components in PM retrieval are shown to interact with PM sampling introducing added uncertainty in runoff simulation. The temporal correlation in the PM retrieval error was found to have a negligible effect in runoff prediction. It is shown that merging rain retrievals from hourly infrared (IR) and PM observations generally reduces flood prediction uncertainty. The error reduction varied between 50% (0%) and 80% (50%) for the 6-h (3-h) PM sampling scenarios, depending on the relative magnitudes of PM and IR retrieval errors. Findings from this paper are potentially useful for the design, planning, and application assessment of satellite remote sensing in flood and flash flood forecasting.     

A nonlinear multispectral statistical CLEAN-based precipitationparameter-retrieval algorithm

An iterative algorithm incorporating CLEAN deconvolution concepts for precipitation parameter retrieval using passive microwave imagery is presented. The CLEAN algorithm was originally designed to deconvolve single-channel radio astronomy images. In order to use CLEAN to retrieve precipitation parameters from multispectral passive-microwave imagery, extensions of the algorithm to accommodate nonlinear, multispectral, and statistical data mere designed and implemented. The primary advantage of the nonlinear multispectral statistical (NMS) CLEAN retrieval algorithm relative to existing algorithms is the use of high-resolution (high-frequency) imagery to guide the retrievals of precipitation parameters from lower resolution (Low-frequency) imagery. The NMS-CLEAN retrieval algorithm was used to estimate rain rate (RR) and integrated ice content (IIC) using simulated imagery of oceanic convection as would be observed from six channels of the proposed Advanced Microwave-Scanning Radiometer. Both the accuracy and structural detail of the retrieved rain rate were improved relative to the retrievals from a single-step, nonlinear, statistical algorithm. Reduced error and improved spatial resolution of a more minor magnitude was also seen in the integrated ice-content retrievals. This study also showed that spatially-simple storm structures resulted in better performance of the NMS-CLEAN retrieval algorithm     

A physical model to determine snowfall over land by microwave radiometry

Falling snow is an important component of global precipitation in extratropical regions. This paper describes the methodology and results of physically based retrievals of snow falling over land surfaces. Because microwave brightness temperatures emitted by snow-covered surfaces are highly variable, precipitating snow above such surfaces is difficult to observe using window channels that occur at low frequencies (/spl nu/<100 GHz). Furthermore, at frequencies /spl nu//spl les/37 GHz, sensitivity to liquid hydrometeors is dominant. These problems are mitigated at high frequencies (/spl nu/>100 GHz) where water vapor screens the surface emission, and sensitivity to frozen hydrometeors is significant. However, the scattering effect of snowfall in the atmosphere at those higher frequencies is also impacted by water vapor in the upper atmosphere. The theory of scattering by randomly oriented dry snow particles at high microwave frequencies appears to be better described by regarding snow as a concatenation of "equivalent" ice spheres rather than as a sphere with the effective dielectric constant of an air-ice mixture. An equivalent sphere snow scattering model was validated against high-frequency attenuation measurements. Satellite-based high-frequency observations from an Advanced Microwave Sounding Unit (AMSU-B) instrument during the March 5-6, 2001 New England blizzard were used to retrieve snowfall over land. Vertical distributions of snow, temperature, and relative humidity profiles were derived from the Mesoscale Model (MM5) cloud model. Those data were applied and modified in a radiative transfer model that derived brightness temperatures consistent with the AMSU-B observations. The retrieved snowfall distribution was validated with radar reflectivity measurements obtained from a ground-based radar network.     

The determination of cloud altitudes using GOME reflectance spectra: multilayered cloud systems

This paper is devoted to the application of the Semi-Analytical Cloud Retrieval Algorithm (SACURA) to the cloud-top height determination using data from the Global Ozone Measurement Experiment (GOME) instrument onboard Earth Remote Sensing satellite (ERS-2). In particular, measurements of the top-of-atmosphere reflectance in the oxygen absorption A-band are used. The technique is based on the asymptotic radiative transfer theory as applied to studies of the oxygen absorption bands in reflected light. Our approach is valid for optically thick clouds with cloud optical thickness larger than approximately 5. The accuracy of the algorithm is checked against independent retrieval techniques for completely cloudy pixels. In particular, the close coincidence with data obtained from the Along Track Scanning Radiometer (ATSR-2) onboard ERS-2 is found. Some results of retrievals using these different instruments disagree (up to 2 km). This is explained by the strong horizontal inhomogeneity of clouds under investigation, which is not accounted by the SACURA or, possibly, by well-known problems of infrared techniques as applied to low-level clouds. The effective cloud geometrical thickness l is also retrieved. This parameter is defined as the geometrical thickness of a vertically homogeneous cloud, which allows for the minimization of differences between observed and calculated top-of-atmosphere reflectance spectra. For inhomogeneous clouds, the value of l differs from a real cloud geometrical thickness, but it gives us an indication of the possible existence of the multilayered cloud system in the field of view of the optical instrument.     

Application of neural networks to AVHRR cloud segmentation

The application of neural networks to cloud screening of AVHRR data over the ocean is investigated. Two approaches are considered, interactive cloud screening and automated cloud screening. In interactive cloud screening a neural network is trained on a set of data points which are interactively selected from the image to be screened. Because the data variability is limited within a single image, a very simple neural network topology is sufficient to generate an effective cloud screen. Consequently, network training is very quick and only a few training samples are required. In automated cloud screening, where a general network is designed to handle all images, the data variability can be significant and the resulting neural network topology is more complex. The latitudinal, seasonal and spatial dependence of cloud screening large AVHRR data sets is studied using an extensive data set spanning 7 years. A neural network and associated feature set are designed to cloud screen this data set. The sensitivity of the thermal infrared bands to high atmospheric water vapor concentration was found to limit the accuracy of cloud screening methods which rely solely on data from these channels. These limitations are removed when the visible channel data is used in combination with the thermal infrared data. A post processing algorithm is developed to improve the cloud screening results of the network in the presence of high atmospheric water vapor concentration. Post processing also is effective in identifying pixels contaminated by subpixel clouds and/or amplifier hysteresis effects at cloud-ocean boundaries. The neural network, when combined with the post processing algorithm, produces accurate cloud screens for the large, regionally distributed AVHRR data set     

Impact of Uncertainty in the Drop Size Distribution on Oceanic Rainfall Retrievals From Passive Microwave Observations

The variability of the drop size distribution (DSD) is one of the factors that must be considered in understanding the uncertainties in the retrieval of oceanic precipitation from passive microwave observations. Here, we have used observations from the Precipitation Radar on the Tropical Rainfall Measuring Mission spacecraft to infer the relationship between the DSD and the rain rate and the variability in this relationship. The impact on passive microwave rain rate retrievals varies with frequency and rain rate. The total uncertainty for a given pixel can be slightly larger than 10% at the low end (ca. 10 GHz) of frequencies commonly used for this purpose and smaller at higher frequencies (up to 37 GHz). Since the error is not totally random, averaging many pixels, as in a monthly rainfall total, should roughly halve this uncertainty. The uncertainty may be lower at rain rates less than about 30 mm/h, but the lack of sensitivity of the surface reference technique to low rain rates makes it impossible to tell from the present data set.     

AIRS/Vis near IR instrument

A component of the Atmospheric Infrared Sounder (AIRS) instrument system is the AIRS/Visible Near InfraRed (Vis/NIR) instrument. With a nadir ground resolution of 2.28 km and four channels, the Vis/NIR instrument provides diagnostic support to the infrared retrievals from the AIRS instrument and several research products, including surface solar flux studies. The AIRS Vis/NIR is composed of three narrowband (channel 1: 0.40-0.44 /spl mu/m; channel 2: 0.58-0.68 /spl mu/m, and channel 3: 0.71-0.92 /spl mu/m) and one broadband (channel 4: 0.49-0.94 /spl mu/m) channel, each a linear detector array of nine pixels. It is calibrated onboard with three tungsten lamps. Vicarious calibrations using ground targets of known reflectance and a cross-calibration with the Moderate Resolution Imaging Spectroradiometer (MODIS) augment the onboard calibration. One of AIRS Vis/NIR's principal supporting functions is the detection of low clouds to flag these conditions for atmospheric temperature retrievals. Once clouds are detected, a cloud height index is obtained based on the ratio (channel 2 - channel 3)/channel 1 that is sensitive to the partitioning of water vapor absorption above and below clouds. The determination of the surface solar radiation flux is principally based on channel 4 broadband measurements and the well-established relationship between top-of-the atmosphere (broadband) radiance and the surface irradiance.     

A neural-network technique for the retrieval of atmospheric temperature and moisture profiles from high spectral resolution sounding data

A novel statistical method for the retrieval of atmospheric temperature and moisture profiles has been developed and evaluated with simulated clear-air and observed partially cloudy sounding data from the Atmospheric InfraRed Sounder (AIRS) and the Advanced Microwave Sounding Unit (AMSU). The algorithm is implemented in two stages. First, a projected principal components (PPC) transform is used to reduce the dimensionality of and optimally extract geophysical profile information from the cloud-cleared infrared radiance data. Second, a multilayer feedforward neural network (NN) is used to estimate the desired geophysical parameters from the PPCs. For the first time, NN temperature and moisture retrievals are presented using actual microwave and hyperspectral infrared observations of cloudy atmospheres, over both ocean and land (with variable terrain elevation), and at all sensor scan angles. The performance of the NN retrieval method (henceforth referred to as the PPC/NN method) was evaluated using global Earth Observing System Aqua orbits colocated with European Center for Medium-range Weather Forecasting fields for seven days throughout 2002 and 2003. Over 350,000 partially cloudy footprints were used in the study, and retrieval performance was compared with the AIRS Science Team Level-2 retrieval algorithm (version 3). Performance compares favorably with that obtained with simulated clear-air observations from the NOAA88b radiosonde set of approximately 7500 profiles. The PPC/NN method requires significantly less computation than traditional variational retrieval methods, while achieving comparable performance.     

Wind-Vector Retrievals Under Rain With Passive Satellite Microwave Radiometers

We have developed algorithms that retrieve ocean-surface wind speed and direction under rain using brightness-temperature (TB) measurements from passive satellite microwave radiometers. For accurate radiometer retrievals of wind speeds in the rain, it is essential to use TB signals at different frequencies, whose spectral signature makes it possible to find channel combinations that are sufficiently sensitive to wind speed but little or not sensitive to rain. The wind-speed retrieval accuracy of an algorithm that utilizes C-band frequencies and is trained for tropical cyclones ranges from 2.0 m/s in light rain to 4.0 m/s in heavy rain. We have also trained and tested global algorithms that are less accurate in tropical storms but can be applied under all conditions. The wind-direction retrieval accuracy degrades from about 10$^{circ}$ in light rain to 30$^{circ}$ at the onset of heavy rain. We compare the performance of wind-vector retrievals under rain from microwave radiometers with those from scatterometers and discuss advantages and shortcomings of both instruments. We have also analyzed the wind-induced sea-surface emissivity, including its wind-direction dependence for wind speeds up to 45 m/s.      

云雷达联合微波辐射计的水凝物识别研究

为降低飞机飞行时的积冰风险,利用模糊逻辑算法和神经网络算法对云雷达联合微波辐射计探测数据进行水凝物分类研究.研究结果表明,两种算法均能有效地完成水凝物分类:模糊逻辑算法对退极化比极为敏感;神经网络算法可以区分水凝物是否存在和进行小粒径粒子分类,但对大粒径非球形粒子不敏感.本研究可从水凝物类型的角度对飞机积冰提供预警服务.     

A nonlinear optimization algorithm for WindSat wind vector retrievals

WindSat is a space-based polarimetric microwave radiometer designed to demonstrate the capability to measure the ocean surface wind vector using a radiometer. We describe a nonlinear iterative algorithm for simultaneous retrieval of sea surface temperature, columnar water vapor, columnar cloud liquid water, and the ocean surface wind vector from WindSat measurements. The algorithm uses a physically based forward model function for the WindSat brightness temperatures. Empirical corrections to the physically based model are discussed. We present evaluations of initial retrieval performance using a six-month dataset of WindSat measurements and collocated data from other satellites and a numerical weather model. We focus primarily on the application to wind vector retrievals.     

Retrieval of Snow and Rain From Combined X- and W-Band Airborne Radar Measurements

Two independent airborne dual-wavelength techniques, based on nadir measurements of radar reflectivity factors and Doppler velocities, respectively, are investigated with respect to their capability of estimating microphysical properties of hydrometeors. The data used to investigate the methods are taken from the ER-2 Doppler radar (X-band) and cloud radar system (W-band) airborne Doppler radars during the Cirrus Regional Study of Tropical Anvils and Cirrus Layers-Florida Area Cirrus Experiment campaign in 2002. Validity is assessed by the degree to which the methods produce consistent retrievals of the microphysics. For deriving snow parameters, the reflectivity-based technique has a clear advantage over the Doppler-velocity-based approach because of the large dynamic range in the dual- frequency ratio (DFR) with respect to the median diameter D₀ and the fact that the difference in mean Doppler velocity at the two frequencies, i.e., the differential Doppler velocity (DDV), in snow is small relative to the measurement errors and is often not uniquely related to D₀. The DFR and DDV can also be used to independently derive D₀ in rain. At W-band, the DFR-based algorithms are highly sensitive to attenuation from rain, cloud water, and water vapor. Thus, the retrieval algorithms depend on various assumptions regarding these components, whereas the DDV-based approach is unaffected by attenuation. In view of the difficulties and ambiguities associated with the attenuation correction at W-band, the DDV approach in rain is more straightforward and potentially more accurate than the DFR method.     

Numerical investigation of intense rainfall effects on coherent and incoherent slant-path propagation at K-band and above

A model investigation is carried out to analyze the impact of intense rainfall on slant-path microwave propagation, using a rainfall microphysical model. The effects are evaluated both for path attenuation, undergone by coherent radiation, and for multiple scattering phenomena, originating incoherent radiation along the path. Atmospheric spatial inhomogeneity is taken into account. The EM propagation model is formulated by means of the radiative transfer theory. The propagation model is applied both to simplified rain slabs and to vertically and horizontally inhomogeneous raining cloud structures in order to compare the impact of atmospheric models on coherent and incoherent propagation. Beacon frequencies between 20 and 50 GHz, elevation angles between 20/spl deg/ and 40/spl deg/ and surface rain rates from 1 to 100 mm/h are considered. Appropriate sensitivity analysis parameters are defined to present and discuss the numerical results. Our main conclusion is that the impact of the convective rainfall structure can be significant both in determining total attenuation and quantifying the contribution of multiple scattering to the received power. For intense rainfall, the use of a rain slab model can both overestimate coherent attenuation and underestimate incoherent intensity. The analysis of realistic raining clouds structures reveals the significance of modeling the volumetric albedo of precipitating ice, particularly at V-band. Total path attenuation can strongly depend on the pointing direction of the receiving antenna due to the intrinsic variability of the precipitating cloud composition along the slant path. Coupling cloud-resolving models with radiative transfer schemes may be foreseen as a new approach to develop statistical prediction methods at Ka-band and above in a way analogous to that pursued by using weather-radar volume data.     

Passive microwave remote sensing of snow constrained byhydrological simulations

This paper describes a snow parameter retrieval algorithm from passive microwave remote sensing measurements. The three components of the retrieval algorithm include a dense media radiative transfer (DMRT) model, which is based on the quasicrystalline approximation (QCA) with the sticky particle assumption, a physically-based snow hydrology model (SHM) that incorporates meteorological and topographical data, and a neural network (NN) for computational efficient inversions. The DMRT model relates physical snow parameters to brightness temperatures. The SHM simulates the mass and heat balance and provides initial guesses for the neural network. The NN is used to speed up the inversion of parameters. The retrieval algorithm can provide speedy parameter retrievals for desired temporal and spatial resolutions, Four channels of brightness temperature measurements: 19V, 19H, 37V, and 37H are used. The algorithm was applied to stations in the northern hemisphere. Two sets of results are shown. For these cases, the authors use ground-truth precipitation data, and estimates of snow water equivalent (SWE) from SHM give good results. For the second set, a weather forecast model is used to provide precipitation inputs for SHM. Additional constraints in grain size and density are used. They show that inversion results compare favorably with ground truth observations