PSR (Polarimetric Scanning Radiometer)
PSR is a NOAA-sponsored versatile airborne microwave imaging radiometer, a technology demonstration instrument, developed by the Georgia Institute of Technology (Georgia Tech) and NOAA/ETL (Environmental Technology Laboratory at Boulder, CO) for the purpose of obtaining polarimetric microwave emission imagery of the Earth's oceans, land, ice, clouds, and precipitation. The PSR was designed to provide several specific and unique observational capabilities from various aircraft platforms. 1) 2) 3) 4)
The PSR is the first airborne scanned polarimetric imaging radiometer suitable for post-launch satellite calibration and validation of a variety of future spaceborne passive microwave sensors. The capabilities of the PSR for airborne simulation are continuously being expanded through the development of new mission-specific scanheads to provide airborne post-launch simulation of a variety of existing and future U.S. sensors, including CMIS, ATMS, AMSU, SSMIS, WindSat, TMI, RAMEX, and GEM. 5)
The original design was based upon several observational objectives:
1) To provide fully polarimetric (four Stokes' parameters: Tv, Th, TU, and TV) imagery of upwelling thermal emissions at four of the most important microwave sensing frequencies (10.7, 18.7, 37.0, and 89.0 GHz), thus providing measurements from X-band to W-band
2) To provide the above measurements with absolute accuracy for all four Stokes' parameters of better than 1 K for Tv and Th, and 0.1 K for TU and TV
3) To provide radiometric imaging with both fore and aft look capability (rather than single swath observations)
4) To provide conical, cross-track, along-track, and spotlight mode scanning capabilities
5) To provide imaging resolutions appropriate for high resolution studies of precipitating and non-precipitating clouds, mesoscale ocean surface features, and satellite calibration/validation at Nyquist spatial sampling.
The original scanhead (denoted “PSR/D” for its use of digital correlators), developed in 1995-96, consisted of four polarimetric radiometers operating at 10.7, 18.7, 37.0, and 89.0 GHz. In order to efficiently utilize the scanhead faceplate area, the 10.7 and 37.0 GHz radiometers utilize a common dual-band antenna, while the 18.7 and 89.0 GHz receivers each utilize single-band antennas. The precise radiometric bands measured by the PSR are X- (10.6-10.8 GHz), Ku- (18.6-18.8 GHz), Ka- (36-38 GHz), and W-bands (86-92 GHz). These bands were selected to provide sensitivity to clouds, precipitation, and surface features over almost one decade of microwave bandwidth at octave intervals. The PSR data has been used to demonstrate the first-ever retrieval of ocean surface wind fields using conically-scanned polarimetric radiometer data. 6)
Tri-polarimetric detection (Tv, Th, TU) was accomplished in PSR/D using custom three-level (1.6-bit) digital correlators operating at 1 Gsample/s. The correlators allowed detection and cross-correlation of 500 MHz wide intermediate frequency bands at the Nyquist sampling rate. High-speed ECL logic was used to accomplish the digital detection. The IF (Intermediate Frequency) subband division technique was used to allow detection of the full IF bandwidths available from the various receivers (up to 2000 MHz). The digital detection hardware was backed up by redundant dual-polarization analog detection hardware.
The PSR/D scanhead contains an 80486 PC, an eight-channel digital correlator bank, and four total- power tri-polarimetric radiometers installed inside a 51 cm diameter and 51 cm long rotating drum. Radiometric data is processed using a 80486-based computer within the scanhead, then transmitted to an archival computer in the aircraft cabin via a 10-base 2 LAN link through the sliprings. Thus, all radiometric detection is accomplished inside the scanhead drum.
Calibration of all radiometers is performed in-flight using standard (unpolarized) hot and cold blackbody targets. The unpolarized load consists of an array of canted pyramidal iron-epoxy absorbing wedges organized in a two-faced L-shaped configuration.
The basic PSR instrument design concept consists of a set of polarimetric radiometers housed within a gimbal-mounted scanhead drum. The scanhead drum is rotatable by the gimbal positioner so that the radiometers can view any angle within 70º elevation of nadir at any azimuthal angle (a total of 1.32 π sr solid angle), as well as external hot and ambient calibration targets. The configuration thus supports conical, cross-track, along-track, fixed-angle stare, and spotlight scan modes (Ref. 5).
The PSR scanhead (Figure 2) was designed for in-flight operation without the need for a radome (i.e., in direct contact with the aircraft slipstream), thus allowing precise calibration and imaging with no superimposed radome emission signatures. Moreover, the conical scan mode allows the entire modified Stokes' vector to be observed without polarization mixing.
The entire PSR assembly (including both scanhead and positioner) was designed for integration into several different aircraft. PSR instruments are flown since 1997 on two NASA aircraft:
• The DC-8 of NASA's DFRC (Dryden Flight Research Center)
• The Orion P-3B of NASA's WFF (Wallops Flight Facility)
After 2000, the PSR instrument assembly has also been flown on additional aircraft:
• Scaled Composites Proteus underbody pod
• WB-57F of NASA/JSC
• ER-2 of NASA/DFRC.
PSR scanning mechanism:
To provide the required imaging capabilities a nadir-viewing two-axis gimbal mount is used to scan the PSR radiometers in the airstream outside the aircraft. The gimbal mount utilizes a 77 cm diameter ring bearing to support a nominal aerodynamic wind load of up to 270 kg. Structural torsional deflections have been designed to be less than 0.01º to maintain positioning accuracy. Scanhead power is provided by azimuthal and elevation slip rings provide power to the scanhead. Sliprings are used to provide power to the scanhead and to facilitate digital and analog data communications via ethernet and NTSC video links. In order to minimize damage to the electronics caused by condensation, a dry gas system is used to purge the scanhead drum and sliprings. The weight of the positioner alone is 215 kg, yielding a total instrument mass (less computer control hardware) of 275 kg. An additional 90-180 kg (depending on the specific aircraft platform) is required for computer control hardware.
The scanning drum (mass = 77 kg) can be positioned to view any angle with 0.1º precision using a two-axis high-torque stepper motor positioner system driven by a dedicated 68000-based programmable controller.
The modularity of the PSR system allows for a straightforward replacement of the scanhead under field conditions with either of several new or planned scanheads. The replacement of a scanhead requires approximately two-hours of time and can be accomplished well within the duration of a typical field campaign. The logical, electrical, and mechanical interfaces to the new scanheads are all identical, thus minimizing design and fabrication costs as well as providing observational versatility.
The positioner and calibration load power requirements vary with the motor's dynamic loads and with ambient temperature, but the maximum required power (during warmup on ascent) is less than 1500 W at 120 VAC. The scanhead power consumption is an additional 240 W at 28 VDC.
Scan modes: All scan modes are software-selectable in flight, and include:
1) Conical, with either fixed-angle or stepped-angle cones and including both fore and aft views
2) Cross-track or along-track
3) Fixed-angle stare, including nadir and (during modest rolls) cold sky stare
Minimum scan times of ~10 seconds for a complete fore-and-aft viewing cone with 53.1º half-angle (or ~4 seconds without hot and cold calibration views) are attained in-flight with excellent mechanical stability. A network of IRIG-B clocks within the scanhead and controlling computers provides position/sample synchronization to better than 1 ms (millisecond). A two-axis gyroscope is mounted on the positioner to provide roll and pitch information at ~10 ms intervals. These data are later used to correct the observed imagery for minor aircraft attitudinal variations.
Calibration of all radiometers is performed in-flight using standard (unpolarized) hot and cold blackbody targets. The unpolarized load consists of an array of canted pyramidal iron-epoxy absorbing wedges organized in a two-faced L-shaped configuration (Figure 3). The absorbing pyramids overlie a thermally-conducting substrate of aluminum pyramids so that the physical temperature of the entire structure remains homogeneous to within 1-2 K. The microwave emission temperature of the structure is thus precisely calculable using measurements of the physical temperature of the array. The pyramids are canted at an angle of 45° so as to provide maximum absorption in the direction of PSR lens antennas. The dimensions of each pyramid were chosen so as to maximize absorption over the entire band of frequencies used by the PSR.
Additional PSR gimbal positioners are being fabricated in order to allow more than one PSR scanhead to be operated from a single aircraft. Synchronized operation of two PSR systems is possible on the DC-8, P-3, WB-57F, or on the Proteus aircraft.
PSR scanhead versions:
• PRS/D: The first PSR scanhead is already described in the “background” above.
• PRS/A: PSR/D evolved into PSR/A during 1998 by replacement of the digital correlators with analog correlators. This replacement was performed as an experiment to study the two correlator types. In addition, several major improvements in the scanhead and data system have been implemented in preparation for the CAMEX-3 campaign, thus leading to a second (analog) iteration of the original scanhead (PSR/A).
- First, the 18 GHz feedhorn and receiver has been upgraded to cover two radiometric bands and now includes a fully-polarimetric channel at 18.6-21.7 GHz, and a dual-polarized water vapor channel at 21.3-21.6 GHz. The new K-band receiver will thus be sensitive to emissions from both the surface and integrated water vapor. The new receiver will also be a direct detecting type rather than superheterodyne.
- Second, each of the five radiometers will incorporate sub-interval calibration hardware to supplement the standard hot- and ambient-view calibration. The sub-interval calibration hardware will switch noise diodes into the RF inputs of the receivers at ~100 ms intervals, pulsing each diode both on and off so as to allow estimation of the radiometer gains and offsets at time intervals comparable to the drift time of receivers. In this manner, uncalibrated gain and offset drift caused by 1/f noise will be significantly reduced.
- Third, the digital correlators were replaced by analog adding correlators using quadrature hybrid arrays. The analog system allows measurement of all four Stokes' parameters for the 10 GHz and 18 GHz channels, and the first three Stokes' parameters for the 37 GHz and 89 GHz channel.
- Finally, using a new pyramidal calibration load the absolute accuracy of the PSR is expected to be ~1 K, thus providing excellent capabilities for absolute intercomparisons of measurements with radiative transfer model calculations.
• PRS/C: The original objective and implementation of this polarimetric radiometer was to provide airborne mapping capabilities in C-band, in preparation for SGP99. In preparation for SMEX02 an X-band radiometer was added to provide vertically and horizontally polarized measurements within four bands at 10.60-10.68, 10.68-10.70, 10.70-10.80, and 10.60-10.80 GHz. Fully polarimetric measurements are provided within 10.60-10.80 GHz. The combined dual-band system provides additional information on soil moisture, along with the capability to measure precipitation and the near-surface wind vector over water backgrounds. The X-band channels also provide additional RFI (Radio Frequency Interference) mitigation capability. 7)
Applications of PSR/CX include ocean surface emissivity studies, soil moisture mapping, sea ice mapping, and imaging of heavy precipitation. Recently, an interference detection and removal algorithm suitable for spaceborne application was demonstrated using PSR/C and PSR/CX data.
Legend to Figure 5: The primary lens/feedhorn antenna is located to the left of a co-boresighted video camera and longwave (10 µm) IR sensor.
• PRS/S: This instrument refers to a “submillimeter” wavelength scanhead with radiometers operating at the temperature and moisture sounding bands on AMSU/A and B. This scanhead contains several dual-orthogonal linearly polarized millimeter- and submillimeter wave radiometers with frequencies ranging from 18.7 to 500 GHz.
PSR/S was designed to study the polarization signatures of clouds and precipitation and provides dual-polarized sensitivity at all major window channels in this range. PSR/S will also be ideally suited for post-launch calibration of all JPSS (formerly NPOESS) microwave sensors (ATMS and CMIS) along with the ATMS on the NPOESS Preparatory Project (NPP).
In addition, the submillimeter-wave channels on PSR/S will be provide simulation capability for the Geosynchronous Microwave (GEM) Sounder, now under consideration for the NOAA GOES series of satellites. GEM has been proposed as a practical geosynchronous microwave sounder, providing sounding channels at 118, 183, 380, and 424 GHz with spatial resolution of ~15-20 km and rapid scanning and update capability. Here, the PSR/C and PSR/S scanheads - used in tandem - will provide important radiometric imagery necessary to determine the capabilities of a proposed microwave sensor to sound temperature and humidity from geosynchronous orbit. The combination of PSR/C and PSR/S scanheads will provide the first means of intercomparing the data available from the GEM-class sounder with data from well-understood systems such as AMSU and SSM/I (Ref. 5).
• PSR/L: Remote sensing of soil moisture can be accomplished under most vegetation conditions using L-band (~1.4 GHz) microwave radiometry. Such data provides good penetration of crop and vegetation canopy as well as measurements of soil moisture at soil depths of up to ~10 cm. A new conically-scanning PSR scanhead operating at L-band (PSR/L) promises to provide more accurate maps of sea surface salinity than currently available. The high resolution of the PSR/L airborne sensor will be particularly valuable for coastal and estuarian salinity mapping and sea ice melt research.
Configuration valid from 1997 to present, operating in 10.7, 18.7, 21.5, 37.0 and 89 GHz. Use of digital correlators in 1997 (PSR/D).
All antennas are of the lens/corrugated feedhorn type and are dual orthogonal-linear polarized with grooved rexolite lenses.
Antenna beam widths
8º (for both X- and K-bands), and 2.3º (for both Ka- and W-bands)
95-97% range, all on-axis cross-polarization isolations exceed -27 dB
Receiver types in PSR/D
Dual-channel superheterodyne with a common local oscillator driving each of two identical mixers
Tv, Th, TU, and TV were measured using custom three-level (1.6 bit) digital correlators operating at 1 Gsample/s. Current capability (PSR/A) is for Tv, Th, TU and TV at 10.7 and 18.7 GHz, Tv, Th, and TU at 37 and 89 GHz, and Tv, and Th at 21.5 GHz, all using analog correlation.
The correlators allowed detection and cross-correlation of 500 MHz wide intermediate frequency bands at the Nyquist sampling rate
IF subband division used
To allow detection of the full IF bandwidths available from the various receivers (up to 2000 MHz)
An 80486 PC, an 8-channel digital correlator bank and 4 total-power tri-polarimetric radiometers installed inside a 51 cm diameter rotating scanhead drum.
Additional detectors in scanhead
Includes a NSTC CCD color video camera oriented in the direction of the antenna main beams.
Scanning mechanism and instrument calibration
A nadir-viewing 2-axis gimbal mount is used to scan the radiometers in the airstream outside the aircraft. External hot and ambient calibration targets are viewed periodically.
No use of radome. Torsions are <0.01º for positioning accuracy and to reduce aerodynamics
Total instrument mass
About 500 kg, ( the positioner and scanhead are 380 kg, an additional 90-160 kg are used for instrument electronics equipment, depending on aircraft type)
Total instrument power
400 W to 1800 W depending on outside air temperature
Configuration valid from June, 1999
Channel definition (GHz)
5.80-6.20 (v, h), with 10º antenna beamwidth
PSR/S configuration. The PSR/S scanhead is implemented for first use in 2001, and permits imaging over a wide range of microwave sounding and imaging bands.
Channel definition (GHz)
18.7-18.8 (v, h, U, V), with 7º beamwidth
All versions of polarimetric scanning radiometers are a compatible series of scanheads, channels, polarizations, and beamwidth. The following PSR versions (PSR/L, PSR/E, and PSR/R) are in the planning stage with PSR/L funded for design in 2001.
Configuration valid for 2002
Channel definition (GHz)
1.400 - 1.406 (v, h), with 10º beamwidth
Configuration planned for 2003/4
Channel definition (GHz)
149-151 (v, h) with 1º beamwidth
Configuration planned for 2003/4
Channel definition (GHz)
35.0 Doppler Radar (VV, HH) with 2º beamwidth
Some early campaigns of PSR instrument support:
The various PRS instruments have been used in support of the following campaigns:
1) Labrador Sea Deep Convection Experiment (March, 1997)
The primary goal of the P-3 Labrador Sea flights was to collect data to verify the utility of passive ocean wind vector sensing in high seas, with secondary goals being to better characterize the thermal emission and backscattering signatures of the wind-driven ocean surface. The PSR was flown under a variety of meteorological conditions in coordinated patterns over both ocean buoys along the eastern U.S. coast and an instrumented research vessel (the RV Knorr) within the Labrador Sea. The flights resulted in the acquisition by the Ocean Winds Imaging (OWI) science team of the first high-resolution polarimetric conically-scanned imagery of the ocean in a broad set of microwave bands, and the first combined joint high-resolution passive and active imagery of the ocean surface. The term high-resolution refers here to a factor of ten beyond that available from spaceborne instruments, typically between 300 m to 3 km spot sizes, and Nyquist sampling.
2) HOWEX (Hurricane Winds Experiment) in Sept.-Oct. 1997
3) CAMEX-3 (Third Convection And Moisture Experiment) took place from Aug. 6 - Sept. 23, 1998 at Patrick AFB, Florida to study Hurricanes Bonnie, Danielle, Earl and Georges.
4) SGP99 (Southern great Plains Experiment). SGP99 occurred over Oklahoma in July of 1999 to study soil moisture imaging at C-band. 8)
5) Meltpond 2000 experiment, June-July 2000. Meltpond was flown out of Thule, Greenland to study high-resolution imaging of arctic sea ice. Both SGP99 and Meltpond 2000 were organized in preparation for the NASA Aqua AMSR mission.
6) SMEX02 (Soil Moisture Experiment 2002). PRS/C measurements were taken in the period 25 June through 12 July 2002 in the regional and Walnut Creek Watershed areas in Iowa, USA. 9)
7) SMEX03 (Soil Moisture Experiment 2003). SMEX03 was designed to provide data to further develop and validate algorithms to accurately retrieve soil moisture from current satellite radiometers such as AMSR-E (flown on Aqua) as well as future microwave sensors such as the NPOESS Conical Microwave Imager and Sounder (CMIS). SMEX03 included several study regions (Oklahoma, Alabama, and Georgia) to provide a diversity of soil types, moisture levels, and vegetation cover. SMEX03 was conducted during the summer of 2003 (June-July) using the PSR/CX version. 10)
8) CLPX02-03 (Cold Land Processes Experiments-2003). Multispectral polarimetric microwave brightness temperature maps of snowpack in the Colorado Rocky Mountains were obtained using the NOAA Polarimetric Scanning Radiometer (PSR) during three Cold Land Processes Experiments (CLPX) in February 2002, February 2003, and March 2003. The CLPX program is sponsored by the NASA Terrestrial Hydrology and the Earth Observing System Programs to study means of remote sensing of properties of the terrestrial cryosphere. Cold land regions play an important role in the Earth's hydrologic cycle and have a significant impact on global weather and climate through their modulation of the Earth's radiation budget. 11)
9) SMEX04 (Soil Moisture Experiment 2004). The objective of SMEX04 is to study linkages between soil moisture and the atmospheric system - in particular, to understand the processes related to the NAME (North American Monsoon Experiment) phenomenon. The PSR/CX instrument was flown on a P-3B aircraft as part of SMEX04. SMEX04 consisted of about 84 flightlines at high altitude resulting in 21 mapping domains flown during August, 2004 (11 over Arizona and 10 over Sonora). 12)
10) SMEX05/POLEX (Soil Moisture Experiment 2005/Polarimetry Land Experiment). SMEX05/POLEX took place in Ames, Iowa in the period June 13 toJuly 4, 2005. The experiments were intended to address algorithm development and validation related to all of the current and scheduled soil moisture satellite systems. Specific objectives were: 13) 14)
- Exploration of unique polarimetric information from satellite instruments such as Coriolis/WindSat and NPOESS/CMIS for soil moisture with supporting aircraft instrumentation
- Diurnal effects associated with soil, vegetation and atmosphere at the 6 am/6 pm observing times of HYDROS, SMOS, CMIS, and WindSat
- Enhancement of Aqua AMSR-E soil moisture validation
- Statistics and mitigation of RFI for CMIS risk reduction.
Of these objectives, polarimetric microwave studies was the primary driver for experiment design. SMEX05/POLEX is the first campaign designed to study the unique and unexplored information that can be extracted for land applications using fully polarimetric observations. The APMIR (Airborne Polarimetric Microwave Imaging Radiometer) instrument of NRL, an aircraft simulator of WindSat and CMIS, was also available for SMEX05/POLEX to facilitate replicate observations of a range of landscape features.
1) J. R. Piepmeier, A. J. Gasiewski, “High-Resolution Passive Microwave Polarimetric Mapping of Ocean Surface Wind Vector Fields,” IEEE Transactions on Geoscience and Remote Sensing, Vol. 39, No 3, March 2001, pp. 606-622
2) J. R. Piepmeier, A. J. Gasiewski, M. Klein, V. Bohm, R.C. Lum, “Ocean Surface Wind Direction Measurement by Scanning Microwave Polarimetric Radiometry,” Proceedings of the International Geoscience and Remote Sensing Symposium, Seattle, WA, July 6-10, 1998
3) A. J. Gasiewski, J. R. Piepmeier, M. Klein, and V. Boehm, “”Remote Sensing of Ocean Surface Wind Vectors by Passive Microwave Polarimetry - PSR Airborne Validation Study,” Report to the NPOESS Integrated Program Office, Silver Spring MD, December 20, 1999
4) Information provided by Albin. J. Gasiewski of NOAA/ETL, Boulder, CO, USA
6) J. R. Piepmeier, A. J. Gasiewski, ”Polarimetric Scanning Radiometer for Airborne Microwave Imaging Studies,” Proceedings of IGARSS 1996 (International Geoscience and Remote Sensing Symposium), Lincoln, NE, USA, May 27-31, 1996, pp. 1688-1691
7) A. J. Gasiewski, M. Klein, A. Yevgrafov, V. Leuski, “Interference Mitigation in Passive Microwave Radiometry,” Proceedings of the IGARSS 2002, June 24-28, 2002, Toronto, Canada
8) T. J. Jackson, A. .J. Gasiewski, A, Oldak, M. Klein, E. G. Njoku, A. Yevgrafov, S. Christiani, R. Bindlish, “Soil moisture retrieval using the C-band polarimetric scanning radiometer during the Southern Great Plains 1999 Experiment,” IEEE Transactions on Geoscience and Remote Sensing, Vol. 40, No 10, Oct. 2002, pp. 2151- 2161
9) “SMEX02 Aircraft Polarimetric Scanning Radiometer (PSR) Tb Data,” http://www.nsidc.org/data/docs/daac/nsidc0205_smex_psr.gd.html
10) T. J. Jackson, R. Bindlish, A. J. Gasiewski, B. Stankov, M. Klein, E. G. Njoku, D. Bosch, T. L. Coleman, C. Laymon, P. Starks, “Polarimetric Scanning Radiometer C- and X-Band Microwave Observations During SMEX03,” IEEE Transaction on Geoscience and Remote Sensing, Vol. 43, No 11, Nov. 2005, pp. 2418-2430,, URL: http://www.ghcc.msfc.nasa.gov/surface_hydrology/publications/tgars_smex03_psr_jackson.pdf
11) B. B. Stankov, A. J. Gasiewski, M. Klein, V. Leuski, B. L. Weber, V. Irisov, D. Cline, A. Yevgrafov, “Airborne Measurement of Snow Cover Properties using the Polarimetric Scanning Radiometer during the Cold Land Processes Experiments (CLPX02-03), Proceedings of IGARSS 2003, Toulouse, France, July 21-25, 2003
12) R. Bindlish, T. J. Jackson, A. J. Gasiewski, B. B. Stankov, M. Klein, “Analysis of PSR Microwave Observations during SMEX04,” AGU Spring Meeting 2005, New Orleans, LA, USA, May 23-27, 2005
13) “About the SMEX05 Polarimetry Land Experiment (POLEX),” URL: http://nsidc.org/data/amsr_validation/soil_moisture/smex05/about_smex05.html
The information compiled and edited in this article was provided by Herbert J. Kramer from his documentation of: "Observation of the Earth and Its Environment: Survey of Missions and Sensors” (Springer Verlag) as well as many other sources after the publication of the 4th edition in 2002. - Comments and corrections to this article are always welcome for further updates.