As part of the continued refinement of the Cyclone Global Navigation Satellite System (CYGNSS) mission’s Level-1 calibration practices, this work reports on recent progress to compensate for thermal artifacts associated with reported effective isotropic radiated power (EIRP) tracked by the receivers’ zenith channels in the form of oscillations with a dominant 40–60-day principal component. The use of the uncorrected EIRPs in the calibration of normalized bistatic radar cross section (NBRCS) estimates is shown to lead to commensurate nongeophysical oscillations. A simple scheme for using in-orbit measurements to rederive prelaunch zenith channel LNA gain tables as a function of the relevant temperature is overviewed. These are subsequently used as part of the EIRP calibration process and to implicitly correct reported NBRCS estimates. The corrected NBRCS estimates are markedly more consistent. It is estimated that the revisions introduced as part of this work, and adopted for the mission’s latest v3.2 Level-1 data release, reduce the errors in the temperature dependence of the transmitter power estimates by an average of 88% while also reducing NBRCS relative spectral densities over the relevant frequency components by an average of 36%. The remaining energy is attributed to a combination of smaller residual, unaccounted for, calibration imperfections but more importantly to real geophysical phenomena causing oscillations over similar time scales whose minimization is not of interest.

We report a study of radio frequency interference (RFI) in the GPS L1 band (1575.42 ± 2 MHz) using spaceborne Global Navigation Satellite System Reflectometry (GNSS-R) data. An examination of the “noise” levels reported by both Spire, Inc. and CYGNSS (Cyclone Global Navigation Satellite System) standard Level 1 products is first presented that shows clear evidence of RFI contributions. Data from 498 acquisitions in CYGNSS’s special raw I/F (Intermediate Frequency) mode is then used to examine RFI source properties in greater detail. Both kurtosis and cross-frequency algorithms are applied with the raw I/F data to detect the presence of RFI. The results suggest that up to 25% of the acquisitions considered contain RFI, with both “narrow” and “wideband” sources of varying amplitudes observed. The results provide insights into RFI source properties in the GPS L1 band as well as methods by which the RFI can be detected.

The NASA Cyclone Global Navigation Satellite System (CYGNSS) mission generates several ocean surface wind speed data products. Level 2 products are sampled in time and space as the measurements are made by each individual satellite in the constellation. Due to the nature of GNSS-R measurements, the sampling is essentially a single pixel-wide swath along the specular point (SP) track on the Earth surface. However, the large number of satellites and their low orbit inclination angle result in dense spatial coverage and frequent revisit rates in the tropics when measurements are aggregated together. Standard Level 3 wind products combine measurements by the full constellation which are then gridded in space and averaged over one hour. A new Level 3 storm-centric gridded wind speed product is reported on a finer spatial grid every 6 hours to better resolve storm structure throughout its lifecycle. Spatial and temporal sampling properties, as well as wind speed uncertainties, are presented here for each of the CYGNSS ocean surface wind speed products.

Monitoring soil moisture enables detailed understandings of its role in the water cycle and how it is affected by climate change. Multiple airborne and spaceborne missions have been dedicated to estimating soil moisture, including Soil Moisture Active Passive (SMAP) [1] , Soil Moisture and Ocean Salinity (SMOS) [2] , and Airborne Microwave Observatory of Subcanopy and Subsurface (AirMOSS) [3] . Some other remote sensing missions have added soil moisture retrievals along with their primary goal. For instance, the primary objective of the Cyclone Global Navigation Satellite System (CYGNSS) mission [4] is wind speed retrieval over the ocean, but it is now also used for soil moisture retrieval, among other applications [5] , [6] . All these remote sensing systems and future systems need in situ soil moisture measurements to validate their products.

This work reports a case study of the use of synthetic cyclone models for the development, assessment and validation of global navigation satellite system reflectometry (GNSS-R) wind speed remote sensing algorithms using a cyclone global navigation satellite system (CYGNSS) data record extending from 1 August 2018 to 31 December 2022. Synthetic cyclone models are shown to be useful in assessing the high wind speed sensitivity of CYGNSS’s v1.0, v2.1, v3.0, v3.1, and future v3.2 normalized bistatic radar cross Section (NBRCS) products due to the extended matchup dataset of high wind speed information that is obtained. The models are also shown useful in investigating the impacts of specific error corrections terms and in the development of level-2 geophysical model functions (GMFs) for the retrieval of ocean surface winds.

This article presents a detailed examination of fluctuating input noise power levels on the analog-to-digital convertor (ADC) sampling hardware of the NASA Cyclone Global Navigation Satellite System (CYGNSS) instruments and the associated impacts on the level 1 normalized bistatic radar cross section (NBRCS) estimation performance. The impact of external noise variations on both the CYGNSS science and navigation channels is quantified with respect to how the fluctuating received power impacts the low-level ADC sampling distribution and subsequent NBRCS estimation. This work demonstrates that there are clear quantifiable geospatially dependent noise variations linked to Global Navigation Satellite System (GNSS) space-based augmentation systems [notably Japanese Quasi-Zenith Satellite System (QZSS) and U.S. Wide Area Augmentation System (WAAS)] and that these additional noise sources significantly alter the digital sampling distribution of the CYGNSS instruments, actively degrading the level 1 NBRCS estimation if not corrected. A derivation of the theoretical correction for both the science and navigation channel ADC sampling variations is presented which is later tuned based on empirical performance metrics in an effort to minimize the induced calibration errors. The impacts of the enhancements outlined in this work on CYGNSS level 1 calibration are evaluated using one year of observations before and after the digital sampling corrections, using model European Centre for Medium-Range Weather Forecasts (ECMWF) wind and Wavewatch III mean square slope (MSS) surface validation datasets.

Global Navigation Satellite System (GNSS) Reflectometry uses reflected GNSS signals for Earth remote sensing applications. Absolute calibration of a Delay Doppler Map (DDM) requires an accurate estimate of the effective isotropic radiated power (EIRP) of the GNSS transmitter, e.g., Global Positioning System (GPS). However, variable transmit power by numerous Block II Follow-on (IIF) and II Replenishment-Modernized (IIR-M) GPS space vehicles has been observed due to their flex power mode. Nonuniformity in the GPS antenna gain patterns further complicates EIRP estimation. A dynamic calibration approach is developed to address GPS EIRP variability. It uses measurements by the direct received GPS signal to estimate GPS EIRP in the specular reflected direction and then incorporates it into the calibration of normalized bistatic radar cross section (NBRCS). Error analyses using Monte Carlo simulations and a root sum of squares (RSS) approach show that the resulting error in NBRCS is about 0.32 dB. Dynamic EIRP calibration instantaneously detects and corrects for power fluctuations in the GPS transmitters and significantly reduces errors due to GPS antenna gain azimuthal asymmetry. It allows observations with the most variable Block IIF transmitters (approximately 37% of the GPS constellation) to be included in the standard data products and further improves the calibration quality of NBRCS and geophysical data products.

The v3 CYGNSS Level 1 calibration algorithm uses the direct signal received by the navigation channel to estimate the GPS EIRP for calibration of the NBRCS. Three sets of antenna patterns play an important role in this calibration algorithm, including the GPS transmit antenna and the CYGNSS zenith and nadir receive antennas. In this paper, we examine how the three antenna patterns are characterized on-orbit and how they are used by the CYGNSS Level 1b calibration algorithm. The impact of antenna pattern knowledge on calibration uncertainty is evaluated and discussed.

The science payload on each spacecraft in the CYGNSS constellation is a GNSS-R receiver and antennas designed to receive GPS L1 signals scattered from the Earth surface. The constellation was launched on 15 Dec 2016 and the GNSS-R instruments continue to operate successfully. A next generation version of the receivers and antennas is in development which adds significant measurement capabilities that are expected to enhance the resolution, precision and coverage of current CYGNSS science data products as well as enable some new ones.

The effective isotropic radiated power (EIRP) of the Global Positioning System (GPS) determines the power incident on the Earth surface. It is critical to the CYGNSS mission's Level 1 calibration of normalized bistatic radar cross section (NBRCS). This paper reports a dynamic EIRP calibration algorithm that uses the CYGNSS direct signal to correct the GPS EIRP in the direction to the specular reflection point. This approach can instantaneously detect any transmit power fluctuation in all GPS transmitters and any change of the receiver system gain, and then automatically applied to correct the science measurement. It also helps mitigate the error caused by the azimuthal asymmetry of the GPS antenna gain pattern without knowing the exact yaw attitude of the GPS satellite. The dynamic EIRP calibration algorithm brings back flagged observations from the Block IIF (~37% of the entire dataset) to the CYGNSS standard science data products. It will also help improve the accuracy of Level 2 wind speed retrieval.

The Cyclone Global Navigation Satellite System (CYGNSS) measures GPS signals specularly reflected from Earth's surface to remotely sense ocean surface roughness and wind speed. The mean square slope (mss) is a key physical parameter that relates the ocean surface properties (wave spectra) with the CYGNSS measurement of the normalized bistatic radar cross section (NBRCS). An approach to model the mss for validation with CYGNSS mss data was developed by adding the contribution of a high frequency tail to the IFREMER WAVEWATCH III (WW3) mss. It is demonstrated that the ratio of CYGNSS mss and the modified WW3 mss can be used to diagnose potential calibration errors that exist in the Level 1 calibration algorithm. This approach can help to improve CYGNSS data quality, including the Level 1 NBRCS and Level 2 ocean surface wind speed and roughness.

Shortly after its launch in December of 2016, the Cyclone Global Navigation Satellite System (CYGNSS) constellation began producing estimates of sea surface wind speeds. With its unique remote sensing method and use of eight satellites, the constellation is able to produce winds with shorter revisit times and in more intense weather than other sea surface wind missions. As a part of the CYGNSS calibration and validation (cal/val) process, these estimates are routinely compared to other estimates in order to understand their uncertainties as well as to diagnose and develop corrections for calibration issues. This paper presents trending of CYGNSS data, focusing primarily on comparisons to numerical weather forecast model analysis wind speed estimates and inter-comparison of simultaneous observations within the CYGNSS constellation.

The eight smallsat CYGNSS constellation was launched into low Earth orbit on 15 Dec 2016. Each satellite carries a four channel bistatic radar receiver which measures GPS signals scattered from the Earth surface, from which surface roughness and wind speed are determined over ocean and soil moisture and inland flooding are determined over land. The mission architecture and satellite design are presented and related to the resulting science data products and their applications. Examples are presented of early on orbit engineering commissioning, calibration and validation of the science data products, and some recent scientific results and applications.

The CYGNSS constellation was successfully launched on 15 Dec 2016 and has been operating continuously in science data-taking mode since March 2017. Updates will be presented on the mission status, calibration and validation activities for its science data products, and recent scientific applications of the measurements. Those applications include the use of ocean wind measurements to estimate air-sea latent heat flux, the assimilation of wind measurements made near tropical cyclones into hurricane numerical prediction models, the retrieval of soil moisture from the scattering measurements made over land, and the imaging of inland flooding using overland measurements.

Level 1 calibration of the Cyclone Global Navigation Satellite System (CYGNSS) measurements requires an accurate estimate of the effective isotropic radiated power (EIRP) of the GPS transmitter in the direction of specular reflection. Variable transmit power by numerous Block IIF and IIR-M GPS space vehicles was observed due to their flex power mode. GPS antenna gain patterns feature azimuthal asymmetry, further complicating EIRP estimation. As a result, all Block IIF observations by CYGNSS are flagged and not used to measure ocean winds. A new (version 3.0) Level 1 calibration algorithm is developed which uses measurements by the direct (zenith) antenna to estimate the specular GPS EIRP used to calculate the bistatic radar cross section (BRCS) of the ocean surface. Direct signal measurements are used to estimate GPS transmitter EIRP in the direction of the CYGNSS spacecraft. By applying corrections to the direct signal EIRP, it is possible to estimate the GPS EIRP in the direction of the specular reflection point. This real-time EIRP calibration algorithm instantaneously detects and corrects power fluctuations in all GPS block transmitters and significantly reduces errors due to the GPS antenna gain azimuthal asymmetry. It also allows observations with Block IIF transmitters (approximately 37% of the entire dataset) to be included in the standard data products.

The CYGNSS constellation of eight smallsats was successfully launched in low Earth orbit on December 15, 2016. Each satellite carries a four channel bistatic radar receiver which measures GPS signals scattered from the Earth surface, from which ocean surface wind speed is determined. The use of a constellation, and the way their orbits are configured relative to one another, enable critical sampling properties. In particular, short time scale physical processes like the rapid intensification phase of a tropical cyclone can be resolved. Results from the 2017 Atlantic hurricane season will be used to demonstrate this.

The Equivalent Isotropically Radiated Power (EIRP) by GPS satellites is needed to accurately calibrate the normalized bistatic radar cross section (NBRCS) measured by the Cyclone Global Navigation Satellite System (CYGNSS). EIRP is the product of GPS transmit power and antenna gain. To determine EIRP, we first estimate the GPS transmit power. A ground-based GPS constellation power monitor (GCPM) system has been built and calibrated to precisely measure GPS signals. The received power is repeatable and verified with German Aerospace Center (DLR)'s independent measurements. The estimated GPS transmit powers are validated with DLR's results and successfully applied to CYGNSS L1 calibration. GCPM measurements also demonstrate GPS antenna pattern asymmetries. Full GPS antenna patterns (over their terrestrial service volume) are estimated using the measurements made with the CYGNSS zenith antennas. CYGNSS zenith antenna measurements are able to sample the full transmit antenna pattern within a very short time.

The processing procedures used for the calibration and validation of Level 2 ocean surface wind speed data product for the CYGNSS mission will be presented in this work. The validation process is planned against a series of ground truth matchups which include buoy measurements, other existing satellite counterparts such as scatterometers, radiometers, altimeters and data from global forecast models.

An overview of the wind speed retrieval algorithm used to generate the first CYGNSS Level 2 wind speed products is presented. The algorithm uses two observables derived from Level 1b calibrated Delay/Doppler Maps, and constructs a geophysical model function which maps the observable value and its associated incidence angle into a wind speed value. The wind estimates from the two observables are also combined to form a best-weighted estimator, which represents our final retrieved wind speed. Here the major steps needed to implement the algorithm are reviewed, with a focus on some new aspects, such as the characterization of the dependence of the observables on incidence angle based on real data. An analysis of the algorithm performance is also presented.

The Tropical Rainfall Measuring Mission (TRMM) Microwave Imager (TMI) dataset released by the Precipitation Processing System (PPS) will be updated to a final version within the next year. These updates are based on increased knowledge in recent years of radiometer calibration and sensor performance issues. In particular, the Global Precipitation Measurement (GPM) Microwave Imager (GMI) is used as a model for many of the TMI version updates. This paper discusses four aspects of the TMI data product that will be improved: spacecraft attitude, calibration and quality control, along-scan bias corrections, and sensor pointing accuracy. These updates will be incorporated into the final TMI data version, improving the quality of the data product and ensuring accurate geophysical parameters can be derived from TMI.

The Cyclone Global Navigation Satellite System (CYGNSS) mission uses a constellation of 8 satellites to measure ocean winds in the tropics with high temporal resolution (~3 hr median, 7 hr mean revisit). Focusing on the measurement of winds in tropical cyclones, each satellite contains a Global Navigation Satellite System reflectometry (GNSS-R) receiver capable of measuring reflections from 4 Global Positioning System (GPS) satellite transmitters under all precipitating conditions over the full range of tropical cyclone winds. In order to provide wind estimates that are consistent from receiver to receiver, the data from each must be intercalibrated. This presentation describes the intercalibration method to be used for CYGNSS. Once the constellation is in operational mode (expected mid to late April 2017), on-orbit data will be used to provide intercalibration offsets between the receivers in order to provide consistent constellation wind estimates.

The Cyclone Global Navigation Satellite System (CYGNSS) uses the Global Positioning System (GPS) constellation (32 satellites) as the active source in a bi-static radar configuration, with CYGNSS acting as the passive radar receiver. A knowledge of Equivalent Isotropically Radiated Power (EIRP), based on transmit power and antenna pattern of GPS satellites, is of great importance in the accurate calibration of L1B data (bistatic radar cross section, BRCS) of the CYGNSS mission. However, the current knowledge of the EIRP of GPS satellites is limited. There exists an uncertainty of transmit power, and only 20 laboratory-measured antenna patterns have been published. Due to the azimuthal asymmetry of the patterns, the yaw attitude of GPS satellites may affect the EIRP. Therefore, a ground-based GPS constellation power monitor system has been built to accurately and precisely measure GPS signals in watts and, from that, estimate the transmit powers and antenna patterns of all GPS satellites. Measurement data without absolute calibration demonstrates that the GPS yaw attitude does affect the received power. A low noise amplifier (LNA) and calibration subsystem implemented on a PID controlled thermal plate is calibrated with a liquid nitrogen source, showing stable and reasonable results. With the absolute calibration of GPS signals. the retrieved GPS parameters will serve as inputs to the CYGNSS L1B calibration algorithm to improve the data accuracy.

Vicarious cold calibration (VCC) for spaceborne microwave radiometers is analyzed and modified for application to conical scanning microwave imagers at frequencies from 6 to 90 GHz. The details of the algorithm are modified to account for additional frequencies and polarizations that were not included in the development of the original algorithm. The modified algorithm is shown to produce a more stable cold reference brightness temperature (TB) than the original algorithm. An analysis is performed of this updated algorithm to show the global regions that contribute to the derivation of the cold reference TB and to show which geophysical parameters contribute to the coldest TBs. The analysis suggests that water vapor variability has the largest impact on the TBs in the VCC algorithm. The modified VCC algorithm is applied to microwave imager data and is used as an intercalibration method. It is shown to agree well with other intercalibration methods, demonstrating that it is a valid and accurate method for calibration of microwave imagers.

GeoSTAR-III, a 180 GHz prototype for the Precipitation and All-weather Temperature and Humidity Sounder (PATH), is the culmination of a decade of technology development funding. The interferometric radiometer comprises 144 receivers operating from 165-183 GHz and utilizes a 192×192 input, ASIC based mixed signal correlator. The demonstration of this instrument raises the technology readiness of the radiometer subsystem to level 6 (TRL 6) and the correlator subsystem to TRL 5. We demonstrate the full functionality of this system with observations of the Sun and Moon as well as nearby thermally emissive objects. This represents the final milestones in the development effort of pre-mission technologies for this decadal survey mission.

The Earth-scene brightness temperature (TB) seen by satellite radiometers is generally cold over the ocean due to low emissivity, warm over land with high emissivity, and with a large dynamic range dependent on frequency and polarization. Calibration at either cold or warm end cannot fully characterize the calibration dependence on the full TB dynamic range. We have developed calibration methods using both warm and cold reference TB tie points and applied them to the Global Precipitation Measurement (GPM) mission particularly for the GPM Microwave Imager (GMI). The two-end method characterizes the GMI calibration dependence on TB and enables the development of TB-dependent intercalibration correction tables for GPM.

Intercalibrating multiple radiometers using data from collocated oceanic fields of view has been a critical method for assessing in-flight instrument performance and reconciling constellation data. In this letter, we present a simple and effective improved method for radiometer intercalibration and apply it to the Global Precipitation Measurement (GPM) Microwave Imager (GMI), the core radiometer of the GPM constellation. By optimizing the collocation criteria, this method is able to produce improved scan-dependent calibration as indicated by the ability to characterize small-scale anomalies and edge-of-scan biases within GMI observations. The accurate characterization of scan-dependent calibration also allows further examination of the intercalibration dependence on brightness temperature (TB). An effective novel method using the cumulative distribution of TBs is used to examine the impact of TB on scan-dependent calibration. It is shown that intercalibration is dependent on TB; however, the GMI magnetic anomalies are additive signals independent of TB and therefore can be corrected by simply subtracting it from the data.

We develop methods of using boreal, temperate, and tropical forests as vicarious calibration sites for spaceborne microwave radiometers. The extended sites and improved calibration techniques enable examining warm-scene performances as a complement to cold scenes over the ocean, providing information about scan-dependent biases, detecting and correcting possible calibration errors, and intercalibrating different radiometers. Specifically, this paper shows results as follows. Warm-scene calibration is expanded beyond previously utilized sites in the Amazon rainforest to include inland boreal and temperate forests, as well as forests in coastal and island regions. These additional sites increase the available warm-scene sample size by a factor of 30 compared with using Amazon rainforests alone, allowing more accurate and statistically robust calibration. In addition, their widespread near-global distribution enables near-continuous monitoring of radiometer performance, e.g., with a temporal resolution of almost once per day. Accurate warm-end vicarious calibration enables or enhances several new capabilities for radiometer characterization. Scan-dependent calibration biases that are present at the high end of a radiometer's dynamic range can, for the first time, be reliably detected and characterized. Using this method, we are able to detect the along-scan magnetic interference and edge-of-scan biases that are present in Global Precipitation Measurement Microwave Imager (GMI) observations at warm brightness temperatures (TBs). Our techniques are able to detect and characterize very small calibration anomalies, e.g., a scan-dependent error in GMI with a peak-to-peak amplitude of 0.1 K. In general, scan-dependent performance and antenna pattern correction can be improved with better constraints at both the high and low ends of on-Earth observed TBs. Regional dependence and seasonal changes in the brightness and predictability of the forested calibration sites are detected and characterized. Calibration exhibits a dependence on latitude, most significantly at water vapor channels, where tropical regions are different from middle and high latitudinal regions. Although the root cause is still under investigation, its presence suggests that using limited sites could result in undesirable latitude-dependent calibration biases. The distribution of calibration sites shifts seasonally, which results from green-up of the vegetation canopy in summer and defoliation in winter, and from snow contamination.

Proposed approaches to the NRC Earth Science Decadal Survey's Precipitation, All-Weather Temperature, and Humidity (PATH) mission involving synthetic aperture arrays require massively parallel, high speed correlators implemented on a geostationary satellite platform. We present testing methodology for a coarse digital correlator chip using a low-power ASIC architecture. The chip was designed in the Electrical Engineering and Computer Science Department of the University of Michigan. These tests precede the integration of the chip into a Geostationary Synthetic Thinned Aperture Array (GeoSTAR) instrument prototype in development at NASA's Jet Propulsion Laboratory.

A constellation of disparate radiometers is inherent to the Global Precipitation Measurement (GPM) mission concept. The task of the Intersatellite Calibration Working group is to generate adjustments to make the measurements of all these radiometers physically consistent. A key role of the GPM Microwave Imager (GMI) on the GPM Core satellite is to serve as a transfer standard among the constellation radiometers. The TRMM Microwave Imager (TMI) has served this role during the development phase and for interim corrections early in the GPM mission. The stability of GMI appears to be very good and a physically based calibration has been generated that appears to be accurate at the 1K level or better.

A calibration issue with the Global Precipitation Measurement (GPM) Microwave Imager (GMI) has been identified and resolved using an extended version of the vicarious warm calibration method in combination with the vicarious cold calibration method. With this extended warm-scene method, we found along-scan anomalies in GMI. The cause is determined to be an error in the existing along-scan correction. The original correction was developed using cold scene data alone, which cannot properly correct biases for warm scenes. A new correction, accounting for both cold and warm scenes, is derived and applied to the GMI data. It properly corrects for the along-scan anomalies across the full dynamic range of on-Earth TBs.

The effect of inter-calibration on a Level 3 rainfall product for the Global Precipitation Measurement (GPM) mission is examined using two spaceborne microwave radiometers that are currently used to derive rain measurements, the Tropical Rainfall Measuring Mission Microwave Imager (TMI) and the Special Sensor Microwave/Imager (SSM/I). It is found that inter-calibrating the microwave radiometer brightness temperatures from the two instruments improves the agreement of the derived rain accumulations between the two radiometers. The average difference between TMI and F13 derived rain accumulations is 0.60 mm/day before inter-calibration is applied. This difference decreases to 0.08 mm/day when F13 is inter-calibrated to TMI.

The electromagnetic spectrum is a valued shared resource. Its scientific use allows us to learn about our universe, measure and monitor our planet, and communicate scientific data. The use of the spectrum is managed by national, regional, and global regulatory frameworks. There are increasing demands for new or extended allocations because of vast technological advances in the past few years. Understanding spectrum management is important in the successful planning and execution of missions and instruments, as well as in determining the potential source of radio frequency interference in existing data and instruments, and in working to ameliorate its impact. This paper provides a summary of this framework for radio scientists and engineers.

Vicarious cold calibration in the frequency range of 85-92 GHz is analyzed. Vicarious cold calibration cannot be applied at these frequencies as easily as at lower frequencies due to greater sensitivity to water vapor and hydrometeor scattering. The effects of that sensitivity are mitigated by selective filtering of the high-frequency brightness temperatures (TBs) to remove those data where large amounts of water vapor and/or hydrometeor scattering are present. Potential filtering algorithms are presented, and the performance of each with respect to vicarious cold calibration TB stability is characterized. A scattering-based precipitation filter that utilizes a combination of both the lower frequencies from 19 to 37 GHz and the frequencies from 85 to 92 GHz is shown to be the most effective and easily implemented filter. For horizontal polarization, the theoretical minimum TB at the higher frequencies occurs at an unphysically high sea surface temperature (SST), which makes the vicarious cold statistic more sensitive to the population of actual SST values as well as the higher amounts of water vapor associated with warm SSTs. The statistic is stabilized in this case by considering the difference between observed and simulated vicarious cold TBs. Intercalibration between two radiometers using the vicarious cold calibration double difference method at high frequencies is shown to be greatly improved when using the precipitation filter.

Multiple independent intercalibration techniques are used to derive calibration adjustments for the development of a fundamental climate data record of physically consistent brightness temperature data from the series of six special sensor microwave/imagers (SSM/Is). The techniques include direct polar matchups, double differencing against model simulations from reanalysis profile data, double differencing against matchups with the Tropical Rainfall Measuring Mission Microwave Imager, vicarious cold calibration, and an Amazon warm calibration. Multiple realizations of three of the five techniques have been applied using different reanalysis data and retrieval techniques to account for Earth incidence angle-dependent differences between sensors. Excellent agreement has been achieved between each of the techniques with typical spread within 0.5 K at the cold end, with slightly higher spread when the warm end estimate is included. A strategy for estimating mean intercalibration values is described with justification for the use of a simple offset based on error characteristics. Intercalibration offsets are smaller for the more recent SSM/I (<; 1 K for F14 and F15 compared with F13) and slightly larger for the older satellites (<; 2 K for F08, F10, and F11 when compared to F13).

A method for estimating the pitch and roll errors of a satellite with an onboard conical scanning microwave radiometer is described. The method makes use of the vicarious cold calibration algorithm which derives a stable cold brightness temperature (TB) over ocean. This cold TB is sensitive to the Earth Incidence Angle (EIA) of the radiometer. Given no pitch or roll errors, the EIA can be modeled as a function of the Earth radius and altitude of the satellite. Deviation from this EIA can then be used to estimate the pitch and roll errors. The pitch/roll algorithm is applied to the current spaceborne microwave radiometer WindSat to show its performance, and the results are compared to the derived pitch and roll of WindSat that are found using a different attitude analysis method.

The robustness of the double difference method used for inter-calibration of microwave radiometers in the Global Precipitation Measurement (GPM) mission is analyzed. The double difference provides a way to compare two different radiometers and is more accurate than just a direct comparison. This is due to the double difference being able to remove geophysical variability from a radiometer's data, as well as frequency and incidence angle dissimilarity between radiometers that would otherwise get included in a direct comparison. These variations are removed by incorporating radiometer modeled brightness temperatures into the inter-calibration process using a vicarious calibration technique. This analysis shows how well the modeled radiometer data is able to remove these variations as well as how well the double difference method performs compared to a direct radiometer comparison.

The Global Precipitation Measurment (GPM) mission requires a high degree of consistency among the microwave radiometers in the constellation which, in turn, demands a standard against which all the sensors can be compared. Ultimately this standard will be the GPM Microwave Imager, but for the present the TRMM Microwave Imager (TMI) fills this need. Since its calibration leaves much to be desired, a refinement using Windsat has been developed. This article defines the Consensus Calibration 1.1 which is applied to the TMI. In turn the TMI serves as a transfer standard to other satellite radiometers.

The vicarious warm and cold calibration techniques are combined to provide an end-to-end two point calibration method for spaceborne microwave radiometers. The method uses stable external calibration sources to permit an end-to-end calibration of the complete radiometer, including its primary antenna. Both gain and offset corrections to the radiometer calibration can be computed since vicarious reference points at both the cold and warm ends of the measurement range are available. The method is demonstrated using the WindSat radiometer. Calibration errors are found which vary with azimuthal scan position in a manner that suggests that the cause is beam spoiling from on-board spacecraft obstructions. The impact on gain and offset calibration errors of the on-board obstructions can be determined from the vicarious calibration. This information is used to characterize the beam spoiling-specifically to determine the decrease in the antenna's beam efficiency and the mean brightness temperature entering the far sidelobes of the antenna, both as functions of azimuthal scan position. With this characterization available, a calibration correction algorithm can be constructed that is based on the root cause of the problem.

The Radio frequency Interference Survey of Earth (RISE) is a new type of instrument used to survey and characterize the presence of Radio Frequency Interference (RFI) that can affect microwave radiometers. It consists of a combined microwave radiometer and kurtosis spectrometer with broad frequency coverage and high temporal and spectral resolution. A K-Band airborne version has been built and flown across southeast Michigan. A kurtosis detector is included in RISE to reliably detect the presence of RFI, even at very low levels, and to aid in its characterization. A radiometer is included to measure the impact of the RFI on observed brightness temperature.

The vicarious cold calibration statistic has been analyzed to determine its stability. Modeled top of atmosphere brightness temperatures representing microwave radiometer observations are computed for the month of July 2005 using a radiative transfer model. The vicarious cold calibration algorithm is applied to the population of brightness temperatures along with two other statistics for comparison. The stability is assessed by perturbing the sea surface temperatures and atmospheric water vapor in the model to simulate a global warming event. The results show that the vicarious cold calibration statistic is the most stable since it has the least variation for the simulated warming event.

An algorithm to detect radio-frequency interference in microwave radiometer brightness temperatures is developed and applied to K-band observations from AMSR-E, WindSat and SSM/I. This algorithm uses the monthly peak difference between co-polar brightness temperatures at 22 and 19 GHz to find RFI. Data from July 2005, July 2008, and January 2008 are shown. Less K-band RFI is seen in SSM/I data than in WindSat or AMSR-E data, likely due to differences in K-band center frequency and spatial resolution. A significant source of RFI is present in the 2008 and 2009 AMSR-E and WindSat data that was not present in 2005. This is likely due to transmissions from the DirecTV 10 satellite, which was launched in July of 2007. This RFI source is seen in reflection off of the Earth's surface. This reflection is strongest over ocean, but is also seen over snow where the diffuse component of the reflection creates a relatively wide swath of RFI.

The WindSat polarimetric microwave radiometer measures top-of-atmosphere brightness temperature, useful for retrieving surface wind vector over the ocean. This procedure was previously documented in low to moderate wind and light precipitation. An atmospheric clearing algorithm designed to remove the emissive and absorptive effects of stronger precipitation and extract the emissivity of the wind-driven ocean surface worked well in moderate rain but had limited success with strong rainfall and high winds. This paper presents results using an improved forward model including Mie scattering from rain. We consider three 2005 WindSat hurricane overpasses for proper atmospheric conditions (Dennis, Katrina and Rita). The improved atmospheric clearing algorithm extracts ocean surface emissivity in and near the hurricane rain bands and eyewall. The emissivity is compared to NOAA H*Wind analysis of the near-surface wind field. Results show a monotonic dependence of emissivity on wind speed up to category 3 hurricane-force winds.

The vicarious cold and warm calibration methods of Ruf, and Brown and Ruf, have been used to assess the calibration of the WindSat radiometer as well as the biases of the TMI, SSM/I F13 and SSM/I F14 radiometers relative to WindSat. WindSat biases computed as a function of scan position are consistent with a roll offset of the instrument of 0.2° and a pitch offset of 0.15°, and biases in the 6 GHz and 22 GHz channels of up to 8 Kelvins are consistent with an obstruction in the feed-horn edge of scan fields of view. Beam fractions and effective brightness temperatures of the obstructions are estimated for each affected channels using both the vicarious cold and warm calibration observations to produce an end-to-end WindSat calibration. Computed biases of TMI, SSM/I F13 and SSM/I F14 relative to WindSat agree well with independent estimates.

The vicarious cold calibration method of Ruf has been used to assess the calibration of the TMI, WindSat, SSM/I F13 and SSM/I F14 microwave radiometers using data from the GPM Inter-Calibration Working Group. Significant scan position dependent biases are seen for TMI (as large as 1 K) and for WindSat (as large as 5 K)-scan position dependent biases in SSM/I data were removed prior to processing. These biases are thought to be due to obstructions in the edge of scan field of view from the given instrument and its spacecraft. WindSat vertically polarized data also show a linear decrease in vicarious cold calibration brightness temperatures with scan position. SSM/I F13 and F14 vicarious cold brightness temperatures differ by an amount consistent with a ~0.2deg offset in their relative Earth incidence angles.

An intercomparison of microwave multiple scattering radiative transfer codes used in generating databases for satellite rainfall retrieval algorithms has been carried out to ensure that differences obtained from retrieval techniques do not originate from the underlying radiative transfer code employed for the forward modeling. A set of profiles containing liquid water and ice contents of cloud and rain water as well as snow, graupel and pristine ice were distributed to the participants together with a black box routine providing Mie single scattering, atmospheric background absorption and surface emissivity. Simulations were to be carried out for nadir and off-nadir (53.1/spl deg/) observation angles at frequencies between 10 and 85 GHz. Among the radiative transfer models were two-stream, multiple stream and Monte Carlo models. The results showed that there were two major sources of differences between the codes. 1) If surface reflection/emission was considered isotropic, simulated brightness temperatures were significantly higher than for specular reflection and this effect was most pronounced at nadir observation and over ocean-type surfaces. 2) Flux-type models including delta-scaling could partially compensate for the errors introduced by the two-stream approximation. Largest discrepancies occurred at high frequencies where atmospheric scattering is most pronounced and at nadir observation. If the same surface boundary conditions, the same multiple-stream resolution and the same scaling procedures are used, the models were very close to each other with discrepancies below 1 K.