MZA Associates Corporation Optical turbulence profiling at White Sands Missile Range North Oscura Peak Dr. Matthew R. Whiteley, Rebecca L. Beauchamp, and Jonathan Diemunsch (MZA) Dr. Richard J. Bagnell Air Force Research Laboratory, Starfire Optical Range November 2, 2006 MZA Associates Corporation 6651 Centerville Business Pkwy Ste B Dayton, OH 45459-2678 CLEARED FOR PUBLIC RELEASE Voice: (937) 432-6560 x237 AFRL/DEO-PA, 31 OCT 06 1 MRW – 11/2/2006 Introduction MZA Associates Corporation z AFRL Starfire Optical Range owns/operates a device known as the “Differential-Tilt Turbulence Profiler” z Development timeline 2000: AFRL/DE and AFOSR developed original concept, theory, initial design 2001-2003: AFRL/DE and MZA integrated the laser & optical system, developed data acquisition software 2002-2005: AFRL/DE conducted simulations, developed processing algorithms, initial testing and data analysis 2006: Full-scale system test at North Oscura Peak (NOP) z Testing timeline 2003: NOP to Bug (~1 km), NOP to Beck (~10 km) Initial data, but issue with source assembly 2004: SOR to 2-mile (~3 km) Revised source assembly Reasonable test data and profiling results for 3 km path 2005: NOP to Salinas (~50 km) Poor SNR limited measurements, profiling questionable 2006: NOP to Beck (~10 km) Good SNR, reasonable profiles, large data volume, other diagnostics 2 MRW – 11/2/2006 Differential-Tilt Turbulence Profiler MZA Associates Corporation atmospheric turbulence Telescope Telescope and source and source assembly assembly data data acquisition acquisition 0 normalized path position 1 3 MRW – 11/2/2006 Estimation of Coherence Diameter from MZA Associates Corporation Profiler Centroid Data z Can use centroid (tilt) data directly from profiler units with standard techniques for r0 estimation Estimation from tilt-variance: Differential-tilt variance for 3 aperture pairs on each unit z Ap-1 / Ap-2 with s/D = 1.5875 Æ z Ap-3 / Ap-4 with s/D = 6.9850 Æ z Ap-5 / Ap-6 with s/D = 1.5875 Æ 4 MRW – 11/2/2006 Profiler Theory of Operation MZA Associates Corporation z Profiler theory published in SPIE proceedings Whiteley, M. R., Washburn, D. C., and Wright, L. A., “Differential-tilt technique for saturation-resistant profiling of atmospheric turbulence,” SPIE Proceedings on Adaptive Optical Systems Technology II 4494, (2001). z Difference of differential-tilt variances define measurement set that can be related to turbulence distribution over path z Relation can be inverted through appropriate numerical technique z Constraints can be applied to inversion using r0 estimates for profiler used as differential-image-motion monitor 5 MRW – 11/2/2006 Wave-Optics Simulation of MZA Associates Corporation Turbulence Profiling WaveTrain System Model 2 2 -15 Profile 2: C relative error = 0.059811 -15 Profile 3: C relative error = 0.10901 x 10 n x 10 n 8 8 reconstructed from data reconstructed from data input to simulation input to simulation 7 7 6 6 5 5 ) ) -2/3 4 -2/3 4 (m (m 2 n 2 n C C 3 3 2 2 1 1 0 0 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 normalized bin position normalized bin position 2 2 2 -15 Profile 4: C relative error = 0.080568 -15 Profile 5: C relative error = 0.16081 -15 Profile 6: C relative error = 0.23198 x 10 n x 10 n x 10 n 8 8 8 reconstructed from data reconstructed from data reconstructed from data input to simulation input to simulation input to simulation 7 7 7 6 6 6 5 5 5 ) ) ) -2/3 -2/3 4 -2/3 4 4 (m (m (m 2 n 2 n 2 n C C C 3 3 3 2 2 2 1 1 1 0 0 0 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 normalized bin position normalized bin position normalized bin position 6 MRW – 11/2/2006 NOP-to-Beck Site Data Collection MZA Associates Corporation z Turbulence profiler data was collected on NOP-to-Beck path 9.6 km, ~900 m altitude change April-August 2006 Other data items Scintillometer km 9.6 Tactical imagery z 33 test days with profiler 7,900 ft z 21 days of good +958 ft +1,512 ft (2,409 ,m) quality/volume for (+292 m) (+461 m) profiling z ~4500 data files z 726 profiles 5,900 ft 0000-0600: 120 5,300 ft (1,799 m) 0600-1200: 86 (1,616 m) 5,015 ft 1200-1800: 110 (1,529 m) 1800-2400: 410 7 MRW – 11/2/2006 Data Quality Checking MZA Associates Corporation APC2006144003600.mat Example of Clipping, Frame 74 z Data filtering implemented to reduce noise 30 sources in differential-tilt estimation 20 25 z Spot SNR 40 Average pixel SNR > 2 for all subapertures 20 z Saturated spot image 60 15 No saturated pixels used in centroid 80 z Spot clipping 10 clipping Subaperture spot too close to centroid 100 processing boundary 5 120 0 z Telescope jitter 20 40 60 80 100 120 Clipping Indicated by Green Circle Quantified by averaging shift over all APC2006144003600. mat 8 subapertures Gross Row Tilt Gross Column Tilt Retained data with jitter std < 7 pixels 6 z Required following attributes for profile 4 processing 2 At least 10% of frames in a file must pass all quality checks 0 Gross Tilt(pixels) At least 200 frames total passing quality checks -2 z SNR was most common reason for data -4 telescope jitter filtering -6 Especially when scintillation was high -8 0 100 200 300 400 500 600 700 800 900 1000 Frame No. 8 MRW – 11/2/2006 Profiling Example: Day 172 MZA Associates Corporation 2 2 1 3 Beck NOP 3 1 Beck NOP Beck NOP 9 MRW – 11/2/2006 Comparison with Scintillometer MZA Associates Corporation z Rytov number from turbulence profiles compared with scintillometer estimate of Rytov number (when available) z It can be shown that for any turbulence profile, the following inequality applies: z A good “rule of thumb” approximation is: z When scintillometer is in bounds implied by profiler, scintillometer and profiler give consistent Rytov number estimates 10 MRW – 11/2/2006 Comparison with Standard MZA Associates Corporation Turbulence Models “Night” 20:00-06:00 “Morning” 06:00-12:00 “Afternoon” 12:00-20:00 11 MRW – 11/2/2006 Conclusions MZA Associates Corporation z Differential-tilt turbulence profiler successfully taken from concept to development and into application ~10 km path from North Oscura Peak to valley floor at WSMR z Substantial test data has been collected and analyzed Filtering for data quality important to assuring turbulence estimate Profile estimates available around the clock Profiles and derived atmospheric propagation parameters consistent with expected trends z Profile estimates validated using independent measurements Consistency between profile Rytov and scintillometer z Diurnal trends (median conditions) Compare favorably to CLEAR-1 model given propagation height above ground Night < 0.5 x CLEAR-1/Night Morning 0.5 – 1.0 x CLEAR-1/Night Afternoon End-points (near ground) 2.0 – 4.0 x CLEAR-1/Night Otherwise, 0.5 – 1.0 x CLEAR-1/Night 12 MRW – 11/2/2006.