Search

Radio-frequency interference (RFI) is an increasing problem particularly for Earth observation using microwave radiometry. RFI has been observed, for example, at L-band by the European Space Agency’s (ESA’s) soil moisture and ocean salinity (SMOS) Earth Explorer and by National Aeronautics and Space Administration’s (NASA’s) soil moisture active passive (SMAP) and Aquarius missions, as well as at C-band by Advanced Microwave Scanning Radiometer (AMSR)-E and AMSR-2, and at 10.7 and 18.7 GHz by AMSR-E, AMSR-2, WindSat, and GPM Microwave Imager (GMI). Therefore, systems dedicated to interference detection and removal of contaminated measurements are nowadays a must in order to improve radiometric accuracy and reduce the loss of spatial coverage caused by interference. In this work, the feasibility of using the empirical mode decomposition (EMD) technique for RFI mitigation is explored. The EMD, also known as Hilbert–Huang transform (HHT), is an algorithm that decomposes the signal into intrinsic mode functions (IMFs). The achieved performance is analyzed, and the opportunities and caveats that this type of methods present are described. EMD is found to be a practical RFI mitigation method, albeit presenting some limitations and considerable complexity. Nevertheless, in some conditions, EMD exhibits a better performance than other commonly used methods (such as frequency binning). In particular, it has been found that EMD performs well for RFI affecting the $< 25\%$ lower part of the intermediate frequency (IF) bandwidth. [ABSTRACT FROM AUTHOR]

The current state of microwave radiometry for remote sensing of the Earth is considered. There are currently some 30 satellite microwave radiometers operating and supplying data in the world, while Russia has only one microwave radiometer (MTVZA-GYa) with a 65° sounding angle. We propose to create a constellation of small spacecraft for global meteorological observations on the basis of the MIRS microwave radiometer, which is being developed for the Konvergentsiya space experiment on the Russian Segment of the International Space Station. Optimal parameters for satellite data of the microwave space system are a spatial resolution of 10–12 km and a temporal resolution of 3–6 h for analyzing current atmospheric processes, improving the quality of weather forecast, and predicting emergency situations. This problem can be solved if there are 4 to 8 simultaneous satellites with onboard radiometers in a single orbit of the Earth. [ABSTRACT FROM AUTHOR]

In this paper, we introduce a novel approach to enhance the spatial resolution of single-pass microwave data collected by mesoscale sensors. The proposed rationale is based on an $L^{p}$ -minimization approach with a variable $p$ exponent. The algorithm automatically adapts the $p$ exponent to the region of the image to be reconstructed. This approach allows taking benefit of the advantages of both the regularization in Hilbert ($p = 2$) and Banach ($1< p< 2$) spaces. Experiments are undertaken considering the microwave radiometer and refer to both actual and simulated data collected by the special sensor microwave imager (SSM/I). Results demonstrate the benefits of the proposed method in reconstructing abrupt discontinuities and smooth gradients with respect to conventional approaches in Hilbert or Banach spaces. [ABSTRACT FROM AUTHOR]

Deep learning artificial neural network techniques can be applied for on-orbit calibration of microwave and millimeter-wave radiometer spaceborne instruments, including those for small satellites. The noise-wave model has been employed for noise characterization and validation of the proposed deep learning calibration technique for a synthetically generated Dicke-switching radiometer. The developed deep learning neural network radiometer calibrator produces high accuracy estimates of antenna temperatures from the measurements of radiometer output voltage and thermistor readings. Tests with noise-free and noisy samples of the developed model have shown that the proposed calibration method does not add any significant noise to the radiometer calibration. The performance of the proposed method does not degrade with increased nonlinearity for a radiometer, while nonlinearity is a challenging issue for conventional calibration techniques. The deep learning calibration model learns the radiometer noise characteristics from radiometer prelaunch measurements during thermal vacuum chamber testing. The neural network calibrator proposed in this paper has self-learning capability during the on-orbit operation of a radiometer that can be used to improve the performance of on-orbit calibration. The proposed technique is demonstrated by comparing the residual uncertainty of the deep learning calibration with the theoretical value. No numerical study is presented to compare the performance with conventional calibration techniques. The new method may be solely applied to calibrate the radiometer or applied along with conventional calibration techniques. [ABSTRACT FROM AUTHOR]

The brightness temperature data acquired during 2012–2015 from the scanning microwave radiometer (SMR), onboard the first Chinese ocean dynamic environment satellite—Haiyang-2A, were matched up with the WindSat Polarimetric Radiometer (WindSat) $0.25^{\circ }\,\,\times0.25^{\circ }$ gridded daily sea surface temperature (SST) data. Then, the artificial neural network (ANN) ensemble (ANNE) method implementing the Crogging technique was used to build the SMR SST retrieval algorithm. Different from a regular ANN, an ANNE combines the outputs of its ANN members to generate an algorithm. The developed ANNE algorithm for SMR SST was validated based on the SMR/WindSat data pairs that were not used in the tuning of the algorithm. The SST comparison shows the root mean square (rms) of 1.16 °C for the ANNE algorithm. We further validate the SMR SST products using the in situ measurements from the National Oceanic and Atmospheric Administration iQuam System. The rms of the ANNE algorithm in comparison with the global iQuam SSTs is 1.46 °C. All validations showed that ANNEs were more accurate than the other statistically based SST retrieval algorithms for SMR, and generally had much smaller uncertainties than regular ANNs. [ABSTRACT FROM AUTHOR]

Abstract: Accurate observations of liquid water content (LWC) in warm stratiform clouds are important for quantifying their radiative and hydrological effects and for studying aerosol‐cloud interactions. Retrieving LWC from radar reflectivity under drizzling or nondrizzling conditions has been investigated for several decades by the cloud remote sensing community. However, no physically plausible framework exists to address the biases introduced by drizzle on existing retrieval techniques. We present the modified Frisch retrieval (ModFrisch), which combines radar and microwave radiometer measurements to retrieve LWC in both nondrizzling and drizzling conditions. It is shown, using a 1‐D steady state microphysical model and a radar simulator, that the uncertainty of ModFrisch is up to four times smaller than the uncertainty of similar retrievals under drizzling conditions, enabling LWC profiling with an accuracy of 20%. The performance of the ModFrisch technique is evaluated using 1 year of observations. [ABSTRACT FROM AUTHOR]

The Global Change Observation Mission 1st-Water (GCOM-W1) satellite, which was launched by the Japan Aerospace Exploration Agency in May 2012, contains an Advanced Microwave Scanning Radiometer-2 (AMSR2). The AMSR2 provides multifrequency measurements of microwave energy (brightness temperature) emitted by the Earth's surface and atmosphere. However, each frequency field of view (FOV) differs in size because of its hardware design. For the retrieval of more accurate geophysical parameters from multifrequency brightness temperatures, the brightness temperatures should be modified to be the same as measured in the same FOV. The Backus–Gilbert (BG) method is one of the antenna pattern matching techniques used for this modification. We applied the BG method to the AMSR2 data to define a new data set of modified brightness temperatures, a level 1R (L1R) product that is freely and widely available. We optimized the implementation of the BG method to obtain the L1R product, with smoothing factors dynamically determined for all modified brightness temperatures. This paper describes the implementation method, including the criterion used to determine the smoothing factors and presents a quality assessment of the L1R product. [ABSTRACT FROM PUBLISHER]

WindSat has been providing accurate ocean vector winds (OVWs) for over a decade. WindSat makes polarimetric brightness temperature measurements of Earth looking forwards and backwards. However, because the overlap of these two swaths is relatively narrow, the benefit of two-look polarimetric (2LP) retrieval accuracy has not been utilized. This paper derives OVW from WindSat 2LP measurements using a radiative transfer model and maximum-likelihood estimation. The purpose of this paper is a comparison of WindSat 2LP wind direction accuracy with WindSat one-look, QuikSCAT, and Advanced Scatterometer (ASCAT) wind directions. Retrievals are compared with anemometer measurements on collocated moored buoys. Statistics are examined for both the first-ranked wind direction ambiguity and the selected ambiguity after median filtering initialized with a numerical weather prediction wind field. For winds above 8 m/s, WindSat 2LP retrievals have the most accurate first-ranked direction compared with all other sensors. For winds of 6–9 m/s, the standard deviation relative to buoys is 17°, and for 9–20 m/s, it is less than 10°. One-look standard deviations are nearly twice as large. At low winds, QuikSCAT provides the most accurate wind directions, for first-ranked and selected ambiguity. Thus, scatterometer and radiometer OVW measurements provide complementary capabilities. The accuracy of 2LP OVW is particularly relevant now that new internal calibration technology allows for a 360° conical scan of earth observations. Moreover, new low-cost designs would make it possible to affordably deploy a constellation of OVW sensors capable of providing accurate winds under a wide range of conditions, described herein. [ABSTRACT FROM PUBLISHER]

Considered is the correspondence between the identification of sea ice properties using the scattering index (SI) computed from the results of measurements accomplished with MTVZA-GYa and AMSU microwave radiometers that register heat radiation at the different angles of incidence. In the framework of the model of the homogeneous ice layer located on the strongly absorbing surface, the difference is specified in the value of the thickness of thin and young ice using the measurements of these radiometers. It is revealed that the minimum values of SI computed from the MTVZA-GYa radiometer data are highly sensitive to the scattering inhomogeneities in the uppermost ice cover layer. Corroborated is the considerable daily variability of SI of the Arctic and Antarctic ice cover that can be an effect of the similar variability of internal stresses and concentration of ice. [ABSTRACT FROM AUTHOR]

A method is applied to maintain the long-term calibration of a microwave radiometer through intersatellite calibration and is used to mitigate an observed calibration drift of the Advanced Microwave Radiometer (AMR) on Jason-2/Ocean Surface Topography Mission. The AMR provides a correction for the wet tropospheric path delay (PD) of the radar altimeter signal, and it is critical that any drift in the radiometer be estimated and removed to enable studies of global mean sea-level variability. The intersatellite calibration method transfers the long-term calibration from other satellite microwave radiometers using a transfer function to map the other sensor's brightness temperature (TB) observations to those of the AMR. Intersensor mapping functions are derived separately for ocean observations and observations over the Amazon rainforest. This provides a warm and cold TB calibration reference to enable the distinction between long-term gain and offset drifts. A database of co-incident observations is generated between the AMR and conically scanning microwave sensors, namely, AMSR-E, TMI, and SSMIS. Monthly averaged differences are found between the AMR and the AMR equivalent TBs computed from the reference sensors. The apparent change in the AMR calibration determined from the three reference sensors is intercompared between the sensors and compared to that determined using natural on-Earth references. It is found that apparent trends in the AMR TBs between the reference sensors and the natural on-Earth references agree within a month to better than 0.4 K. The AMR 18.7- and 23.8-GHz channels are found to be stable to 0.5 K over the first three years of the mission, and the calibration 34.0-GHz channel is found to drift downward by approximately 6 K. In all channels, the calibration change is determined to be a series of offset jumps (independent of TB). These calibration changes in each AMR channel are estimated and removed using the comparisons to the reference sensors. The uncertainty in the PD long-term stability after recalibration is estimated to be less than 0.5 mm/year from July 2008 to August 2011. [ABSTRACT FROM PUBLISHER]

This paper describes a study of compensation for the physical temperature variation in microwave radiometers. The compensation method is derived by using a multivariable least squares estimation with a radiometer model with temperature-dependent parameters for three system blocks. The feasibility of the method is demonstrated with several experiments with a Ka-band total power radiometer and a temperature control system developed to test the method in various thermal conditions. The results show that the compensation method can be applied to calibrate microwave radiometers without switching systems and is expected to compensate internal temperature gradients and temporal variation of physical temperature. [ABSTRACT FROM PUBLISHER]