1. 1 RSP Measurements and Data Availability Brian Cairns

2. 2 RSP observations: Scan plane oriented along aircraft to allow for aggregation of multiple looks, with different view angles, of the same (within ~ 100 m) ground pixel. Measurement Approach Unvignetted views always ±50° with some variations depending on platform and mounting of instrument on aircraft (e.g. CRYSTAL-FACE +40/-80°) Spectral bands at 410, 470, 555, 670, 865, 960, 1590, 1880 and 2260 nm.

3. 3 How do you measure polarization? Use a retarder and a polarizer. V is negligible for solar illumination and only a polarizer is needed to measure I, Q and U. Measurement Approach Retarder,  Polarizer,  ElEl Detector, Stokes Vector Degree of Linear Polarization, Angle of Polarization

4. 4 Measurement Approach Q and U are calculated from the differences of measurements with different polarizer orientations. If the scene intensity changes (aircraft, or spacecraft moves) between these measurements “false” polarization will occur. This means that the accuracy of the measurements depends on the scene and has no guaranteed level of accuracy. Simultaneous MeasurementsSequential Measurements Polarization ImagesIntensity Images

5. 5 Measurement Approach This is apparent in airborne POLDER simulator data

6. 6 Measurement Approach The differences required to calculate Q and U are differences between orthogonal polarization states, so if we measure these orthogonal states such that they are looking at the same scene at the same time we can effectively eliminate “false” polarization. This can be done very simply using a Wollaston prism in the collimated beam of a relay telescope. Wollaston prism - splits beam into orthogonal polarizations ObjectiveCollimator Dichroics Field Stop

7. 7 Measurement Approach Using dichroic beam splitters you can make measurements for multiple spectral bands in a single telescope (3 in the case of RSP). Use one telescope for Q and one telescope for U. If we are measuring a total of 9 bands this means we need 3 telescopes for Q and 3 telescopes for U for a total of 6 telescopes. RSP telescopes each have three bands – 410, 555, 865 (A), 470, 670, 960 (B) and 1590, 1880 and 2260 nm (S).

8. 8 Measurement Approach We can make telescopes that will measure polarization accurately. Now we need to point them at things (the earth) without losing the polarimetric accuracy they provide. Crossed mirrors, if identical, introduce no polarization into the scene polarized radiance and allow the telescope fields of view to be scanned across the earth either across track like MODIS, or along track as is planned for APS. One polarization experiences an s, then a type p reflection, while the other experiences a p then an s type reflection. Polarization induced by scan mirror assembly of RSP was not measurable <<0.1%. Scanner uses matched mirrors illuminated at 45° with reflection planes at 90° to one another RSP mirror alignment

9. 9 Measurement Approach Since 2002 the RSP has had an onboard polarimetric calibrator to track relative gain of detectors. This approach has demonstrated accuracy of better then 0.1% using clouds to provide a bright weakly polarized source. Research Scanning Polarimeter aircraft instruments have been in operation for 12 years and the polarimetric calibration has been stable to the ~0.2% level throughout that period.

10. 10 Measurement Approach Uncertainties in RSP measurements are given by the following formulae where the uncertainty in radiometric calibration is   ~3%, depolarization is   ~0.1% and relative gain is   ~0.1%. The noise is almost always dominated by shot noise and for a reflectance R the shot noise is  n ~0.0003sqrt[cos(q s )R]/r sol

11. 11 Measurement Approach RSP has a 14 mrad IFOV so even at 16 km altitude the ground pixel size is only ~200 m. However it is possible to provide an estimate for the polarization error that is caused by sub-pixel heterogeneity and this is given by the expression where I1 and I2 are the intensities measured in paired telescopes for a given band and is the sub pixel variance in intensity in telescope 1 The IFOV matching can be measured using a collimator to scan within the IFOV and for the RSP all bands have matching of better than 0.5% (VNIR are < 0.2%).

12. 12 Lead scientists for this field experiment were Bill Smith Jr. and Tom Charlock. Chesapeake Lighthouse and Aircraft Measurementsfor Satellites ( July-August, 2001), CLAMSChesapeake Lighthouse and Aircraft Measurements for Satellites ( July-August, 2001), CLAMS Data Availability RSP on board the Cessna aircraft 210 aircraft, single engine so no distant off shore flights In situ and AATS data available. AATS data can be acquired from Phil Russell. In situ data requires contacting Peter Hobbs.

13. 13 Flew both RSP instruments onboard a Cessna 210 aircraft to look at smoke advecting north over the Mojave desert on 10/29/2003. No other data available Data Availability

14. 14 Arm Lidar Validation Experiment (ALIVE), September 2005.Arm Lidar Validation Experiment (ALIVE), September 2005. RSP on board the Sky Research J31 with the AATS-14 AATS data can be acquired from Phil Russell/ARM archive. Lead scientist for this field experiment was Beat Schmid. Data Availability

15. 15 500 feet over Mexico City near the approaches to the world’s busiest airport. AATS CAR RSP SSFR NavMet POS, March 2006.Megacity Initiative: Local and Global Research Observations, March 2006. Data Availability

16. 16, July 2003.Coastal Stratocumulus Imposed Perturbation Experiment, July 2003. RSP onboard the Cessna 310 aircraft - need two engines to fly off the coast. Data Availability

17. 17, July 2002.Cirrus Regional Study of Tropical Anvils and Cirrus Layers - Florida Area Cirrus Experiment, July 2002. RSP onboard the Scaled Composite Proteus aircraft - flies at up to 60 kft. Survived outside temperatures down to - 80°C. Data Availability

18. 18.Arctic Research of the Composition of the Troposphere from Aircraft and Satellites (ARCTAS), June and July 2008. RSP onboard the NASA LaRC B200 Collocated measurements with HSRL In Situ from NASA P3 Data Availability

19. 19 Routine AAF CLOWD (Clouds with Low OpticalWater Depths) Optical Radiative Observations (RACORO ), June 2009. RSP onboard the NASA LaRC B200 Collocated measurements with HSRL In Situ from CIRPAS Twin OtterIn Situ from CIRPAS Twin Otter Ground based SGP dataGround based SGP data Data Availability

20. 20 California Nexus of Air Quality and Climate Change (CALNEX) field experiment, May 2010. RSP onboard the NASA LaRC B200 Collocated measurements with HSRL In Situ primarily from NOAA P3In Situ primarily from NOAA P3 Data Availability

21. 21 Carbonaceous Aerosols and Radiative Effects Study (CARES), June 2010 RSP onboard the NASA LaRC B200 Collocated measurements with HSRL In Situ from DoE G1In Situ from DoE G1 Data Availability

22. 22 Caribbean Observations of Clouds of Aerosols (COCOA?), August 2010 RSP onboard the NASA LaRC B200 Collocated with HSRL Primarily CALIPSO validation, no in situ Data Availability

23. 23 RSP onboard the NASA LaRC B200 Collocated with HSRL Shipboard measurements, contact Yongxiang Hu for details Data Availability Chesapeake, September 2009

24. 24 Birmingham, September and October 2008 RSP onboard the NASA LaRC B200 Collocated with HSRL Support for EPA study using HSRL Data Availability

25. 25 Other Information Prior to ARCTAS the number of view angles in RSP data was 152. This meant that the operator had to modify the starting angle in order to view the polarimetric calibrator. So, calibration files are separate from data files. After ARCTAS the number of view angles in RSP data is 195. This means that the polarimetric calibrator is observed continuously and there is no difference between calibration and data files. We are in the process of doing a uniform reprocessing of all the data from ARCTAS forward. Once that is complete we will go back as far as CLAMS and reformat all the previous data into the same netCDF format as the more recent data.

26. 26 Other Information Look up tables that can be used to analyze the RSP data will be made available. These tables will be field experiment specific because of different aircraft operating altitudes. Existing vector radiation code is available to anyone who wants it. This code is being updated to also provide analytical Jacobians to allow iterative techniques to be more efficiently implemented. K-distributions are available and will be supplemented with a simple, but accurate parameterization of absorption for the 1590 and 2260 nm bands.

27. 27 Acknowledgements Funding: Funding for these activities was provided by the Glory Project managed by Bryan Fafaul and NASA Radiation Sciences Program managed by Hal Maring Many thanks to: Chris Hostetler, Rich Ferrare, Mike Obland, Ray Rogers, John Hair, NASA Langley Research Center, Hampton, VA, USA who have allowed us to work with them on the B200. Jens Redemann, Bay Area Environmental Research Institute, Sonoma, CA, USA; Phil Russell, NASA Ames Research Center; Anthony Clarke, Cam McNaughton, University of Hawaii, Honolulu, HI, USA; Yohei Shinozuka, NASA Postdoctoral Program; Bill Conant, University of Arizona; John Seinfeld, California Institute for Technology; Bill Smith and Tom Charlock NASA Langley Research Center and Beat Schmid, DoE ACF, who have helped with data and/or field experiment participation. Dick Chandos and Ed Russell built the RSP instruments.