Results from the Atmospheric Neutral Density Experiment Risk Reduction Mission

Transcription

1 Results from the Atmospheric Neutral Risk Reduction Mission A.C. Nicholas, S.A. Budzien, J. DeYoung, L. Healy Naval Research Laboratory M. Davis Honeywell TSI

2 Outline Mission Description C D Modeling Data Flow Data Processing & Analysis Conclusions Future Work 2/20

3 ANDE Concept Objectives: 1. Provide Total Atmospheric Density for Orbit Determination and Collision Avoidance 2. Validate Fundamental Theories on the Calculation of the Drag Coefficient 3. Provide Calibration Objects for SSN 4. Establish a Method to Validate Neutral/Ion Density & Composition Derived from DMSP Sensors. Satellite Laser Ranging Site Description: Fly two 19 spheres in lead-trail orbit 400 km orbit 51 Degree inclination Passive Sphere (~25 kg) Observed with SSN and SLR; variation in observed position used to determine in-track total density Active Sphere (~50kg) Determine position wrt to passive sphere Compute total density Validate C D models Use on-board instrumentation to calculate density and composition Launch via Shuttle in CY 2009 RR deployed 21 Dec 2006 Point of Contact Andrew Nicholas andrew.nicholas@nrl.navy.mil 3/20

4 RR Mission Goals Goals: Test CAPE deployment System Determine spin rate & spin vector of sphere Prototype calibration target Science: Total neutral density, C D studies, geomagnetic storm effects, fence calibration MAA: Same size/mass/form factor as ANDE Active sphere 48 cm diameter, 52.0 kg Allows for testing of CAPE with sphere similar to ANDE active Want to measure spin rate & spin vector on orbit Retros, paint scheme Laser diodes Test backup comms system for ANDE active Space qualify new photovoltaic arrays Thermal monitoring system FCal: Design driven by NSC radar fence sensitivity needs cm diameter, mass 63.7 kg Retros Cubesat payload/comms system Thermal monitoring system Phototransitors 4/20

5 Drag Equation A Drag = -1/2!"V 2 Atmospheric density (!) is #1 source of error in LEO OD process Current atmospheric density models are empirical climatological models (MSIS, Jacchia, all with 15% to 25% error inherent). Sparse direct measurements of atmospheric density for validation. B = C D A/m, inverse ballistic coefficient (B) Coef. of drag (C D ), modeled Function of size, shape, material, roughness, temperature, atmosphere Frontal Area (A) of spacecraft Spherical symmetry is independent of orientation Mass (m) of spacecraft Velocity wrt medium V=v sc -v m Actual Winds; 5-10% of v sc during geomagnetic storms Observe V, solve for product of!" Minimize variations in B to maximize accuracy of! Expected 5/20

6 ANDERR Data Coverage

7 Atmospheric Neutral!v ~1.8 f t /s ec an t i- R ANDERR Deploy from STS-116 Dec. 21, :22 UT AM!v ~ 0.5 ft/sec!v ~ 0.6 ft/sec 7/20

8 C D Modeling The C D values were modeled for both ANDERR spacecraft (94 plates: 73 leading, 21 trailing) As flown mass and size properties Spacecraft Diameter Mass Avg. (cm) (kg) C ( D C D = 2s2 +1) e #s 2 ( + 4s4 + 4s 2 +1)erf ( s) + 2 " ( 1#$ ) T sc " s 3 2s 4 3s MAA dia T % " s = v 2R T % sc $ ' # & M w ( 1 2 FCal dia MAA FCAL 8/20

9 SLR Augmented Predictions SSN observations from Dahlgren Up to 3/day ILRS SLR observations number/day varies Special-K predictions From SSN observations Up to 3/day Predictions using SSN-only data Preprocessing of SSN observations Once/day OCEAN merge of SLR+SSN observations Once/day OCEAN-generated 2-day SLR Augmented Predictions Atmospheric Model Correction 9/20

10 C D Results Orbit determination derived C D values (MSIS residuals) are non-physical! These are results from 2-day fit spans; fitting product of!b Atmospheric models are over estimating density (!) # resulting in an artificially low C D Consistent with secular change work of Emmert et al. (GRL, 2008) modeled C D value Object Avg. Fitted C D Std. Dev. MSIS Scale MAA FCal /20

11 Comparison to Atmospheric Drivers Atmosphere has solar and geomagnetic forcing: Solar: EUV irradiance (F10.7 cm flux is a proxy) & Geomagnetic: A p,, V sw NRLMSISE-00 daily avg. density Corrected daily avg. density 11/20

12 Periodicities! residuals!"# $%&'($)*%+, Lomb-Scargle analysis performed on time series Retrieved C D exhibit periods of 9, 18 and 27 days F10.7 Strong 27-day period observed in all drivers Expected due to solar rotation period 18-day period observed in C D and F 10.7 A p 9-day periods observed in C D, A p & V sw Consistent with 2005 CHAMP density periods seen by Lei et al. (2008) and Thayer et al. (2008) V sw EUV flux Shorter periods (5-, 7-day) are observed in A P & V sw Not observed in ANDERR C D data Possibly due to time, altitude and/or inclination differences ANDERR: ~300 km, 51.2 CHAMP: ~400km, 87.3 TIMED SEE data does not show periods < 27 days Dotted line is 95% significance level 12/20

13 Timed SEE Periodicities 0-10 nm nm nm nm nm nm nm nm nm nm nm nm nm nm nm nm nm nm nm Dotted line is 95% significance level Investigated periodicities in the TIMED SEE Irradiance data Flux summed in 10 nm bands ~27-day period in three bands: a nm b nm c nm ~20-day period observed at nm Also observed in F /20

14 Periodicity Timeline Atmospheric Neutral Wavelet analysis performed on time series to investigate when periodicities occur CD response is time lagged with F10.7 for the 18-day period Short term periods more prevalent in first half of 2007!/!msis White contour is 95% significance level 14/20

15 Conclusions ANDERR fitted B values show MSIS and Jacchia overestimate atmospheric density by ~25%. Both models show correlation B with indices for the atmospheric drivers (solar and geomagnetic) ANDERR provides an absolute atmospheric correction term Periodicities were observed in drag-derived! residuals and the atmospheric drivers 27-day periods were observed in! residuals and with all drivers (F 10.7, A p, V sw, SEE irr ) 18-day period was observed in! residuals and F 10.7! residuals response is time lagged from F day periods were observed in! residuals and geomagnetic drivers (A p, V sw ) Technique allows one to separate solar irradiance forcing from geomagnetic forcing ANDERR results are similar to CHAMP results (Lei et al & Thayer et al. 2008) 1 st half of 2007 shows similar periodicities in A p, V sw as in 2005 ANDERR exhibits strong 9-day but weak 5- and 7-day periods (CHAMP showed all 3) ANDERR orbit (~300 km, 51.2º), CHAMP (~400 km, 87.3º) Future Work Investigate altitude and inclination differences between ANDERR/CHAMP data sets Investigate 18-day period Apply corrections to other objects (Champ/GRACE/ISS) Compare with other indices/models (E 10.7, SOHO EUV, DST, MgII, Emmert, JB2008, etc ) Full ANDE Mission STS-127, NET May 2009 Different part of solar cycle 15/20

16 ANDE Timeline ANDERR Re-entry: ICU Cyl 2 (2/9/07) ICU Cyl 1 (4/18/07) ICU Avionics Deck (4/24/07) ANDERR MAA (12/25/07) ANDE RR FCal (5/25/08) ANDE: Launch in NET May 2009 via Space Shuttle STS-127 Goal: 1. Measure V s/c (ground based sensing: radar, SLR, GPS) 2. Measure V m (in-situ: WATS) 3. Measure! (in-situ: WATS) Instrument Measurement Description Developer A Drag = -1/2!"(V s/c -V m ) 2 ANDE Instrument Payload GPS Position single frequency GPS receiver WATS Ion & neutral in-situ wind and winds and temperature temperatures spectrometer imesa Electron density & temperature miniature electrostatic analyzer Univ. Texas, Austin Naval Research Lab US Air Force Academy Gyroscopes and accelerometers Spin rate and orientation 3 gyros and 3 MEMS accelerometers in each spacecraft Local High Schools: Marchall Academy, Chantilly Academy, Westfields High School Thermal Monitoring Temperature of spacecraft Array of thermistors for internal and skin temperature Naval Research Lab 16/20

17 Atmospheric Neutral Questions? 17/20

18 Back-up Slides 18/20

19 Prediction FCal with MAA Scaling 19/20

20 Science Summary Science Obj. Instrument SSN & SLR GPS TMS WATS MESA Model Instrument or Mission Total Neutral Density A,P,R A A A* MSIS, J70, TMC SSLUI, CHAMP STARSHINE, GRACE Proof of Concept R(x4) A A A C D Studies A,P,R A A,R A A Raytheon STARSHINE, CHAMP, GRACE Attitude Determination R A * A * Composition A A MSIS,GAIM SSULI Temperature A A MSIS SSULI Cross-Track Winds A HWM CHAMP Geomagnetic Storm Effects A,P,R A A,R A A MSIS STEREO, SSLUI, CHAMP, GRACE SSN Calibration/SOI A,P,R Special K OCEAN A = Active Sphere P = Passive Sphere R = Risk Reduction * = Inferred 20/20

21 Neutral Atmospheric Specification Evolution Neutral Atmospheric Specification Evolution Improved Operation Capability Hydrostatic Climatology Data Driven Models Real-Time (Requires Prediction) Now Cast STD ATM J70 HASDM (Drag Data, ANDE) Sapphire Dragon (AF SSN) Exponential MSIS NRL MSIS Data-Driven (ANDE, UV Limb Data) NRL MSIS DTM GAIM/NADIR MURI (UV Limb/NADIR, ANDE, Need neutral density to initiate model) GAIM (AFWA) NADIR MURI (AFOSR) NRLMSISE-00 HWM Data Enhanced Wind Models (ANDE) ANDE Missions Assimilate All Data (New Data ATM Model) 21/20

22 Secular Change Solar min Solar Min Solar Max Depletions of the thermosphere occurring at solar minimum Emmert et al. (GRL, 2008) 22/20