JPL/Caltech NASA 23 Jul 2009, C.S. Jacobs Page 1 Extending ICRF to 24, 32, and 43 GHz C.S. Jacobs JPL/Caltech/NASA 23 July 2009 Extending the ICRF to Higher

JPL/Caltech NASA

23 Jul 2009, C.S. Jacobs Page 1

Extending ICRF to 24, 32, and 43 GHz

C.S. Jacobs JPL/Caltech/NASA23 July 2009

Extending the ICRF to Higher Radio Frequencies:24, 32, and 43 GHz Global Astrometry

URS208977

JPL/Caltech NASA



• Motivation for a radio frame above 8 GHz

• Radio source structure/position/frequency effects

• Overview of Radio frames at 8, 24, 32, and 43 GHz

• K-band (24 GHz) frame

• X/Ka-band (32 GHz) frame

• Conclusions

Outline

JPL/Caltech NASA



Motivation

• Astrometry, Geodesy and Deep Space navigation have been at 8.4 GHz (X-band) with 2.3 GHz (S-band) plasma calibrations

Going to Higher radio frequencies allows

• Potentially more compact sources Potentially more stable positions • Higher Telemetry Rates to Spacecraft• Avoid 2.3 GHz RFI issues• Ionosphere & solar plasma down 15X !! at 32 GHz (Ka-band) compared to 8 GHz.

Drawbacks of Higher radio frequencies:• More weather sensitive• Weaker sources, shorter coherence times• Many sources resolved• Antenna Pointing more difficult

Picture credit: SOHO/ESA/NASA

JPL/Caltech NASA



The Atmosphere Effects Sensitivity

Murphy, R. et al., 1987, Earth Observing System Volume IIe: HMRR High-Resolution Multifrequency Microwave Radiometer. Published by NASA, Goddard Space Flight Centre, Greenbelt, Maryland 20771, USA, 59pp.

The three curves show absorption in a dry atmosphere, in the same atmosphere with 20 kg/m2 of added water vapour, and with both water vapour and 0.2 kg/m2 of stratus cloud added. .

Valleys are microwave windows

Murphy, R. et al., 1987, Earth Observing System Volume IIe: HMRR High-Resolution Multifrequency Microwave Radiometer. Published by NASA, Goddard Space Flight Centre, Greenbelt, Maryland 20771

Atmosphere Background:

3K/atmo @ 8 GHz

10-15K/atmo @32 GHz

JPL/Caltech NASA



Source Structure:

Natural radio sourcesat 8 to 43 GHz

are not point-like

Credit for this section:

E. B. Fomalont, NRAO. Charlottesville, VA

& C.S. Jacobs, JPL/Caltech/ NASA, Pasadena, CApresented at DDA meeting, Virginia Beach, May 2009

JPL/Caltech NASA



Structure vs. Frequency: scaled to beam

S-band X-band K-band Q-band2.3 GHz 8.6 GHz 24 GHz 43 GHz13.6cm 3.6cm 1.2cm 0.7cm

Ka-band32 GHz0.9cm

The sources become better ----->

Image credit: P. Charlot, Bordeaux Observatory & KQ VLBI Collaboration

0450-020 = J0501-0159

JPL/Caltech NASA



Structure vs. Frequency: Absolute scale

S-band X-band K-band K-band2.3 GHz 8.6 GHz 24 GHz 24 GHz6x3 mas 2x1 mas 2x1 mas 0.7x0.3masPk = 1.58 Jy Pk = 1.31 Jy Pk = 0.95 Jy Pk = 0.92 Jy

The sources becomes more compact ----->

Contours: -1, 1, 2, 4, 8, 25, 50, 75, 99% of maxAngular scale in milli-arcsec. Same scale for each diagram

0450-020 = J0501-0159

Credit: Petrov et al; Charlot et al., AJ, submitted Nov 2008.

Same resolutionas 8.6 GHz

Nominal resolution

JPL/Caltech NASA



Features of AGN (from Marscher)

Features of AGN: Note the Logarithmic length scale.Credit: Alan Marscher, `Relativistic Jets in Active Galactic Nuclei and their relationship to the Central Engine,’Proc. of Science,VI Microquasar Workshop: Microquasars & Beyond, Societa del Casino, Como, Italy, 18-22 Sep 2006.

View angleDoppler enhanced

One-sidedMost sources

>0.1 Jy

~1 masRs~

0.1 as

| |>100 2.3 GHz Radio Distant Radio Core Jets Lobes

JetBase

BlackHole

JPL/Caltech NASA



Overview of

Radio-based Celestial Frames

at 8.4, 24, 32, and 43 GHz

JPL/Caltech NASA


Extending ICRF to 24, 32, and 43 GHzS/X (2/8 GHz) ICRF1: 608

Sources

608 sources from S/X data up to 1995Errors Inflated: scale plus 250 µas RSS’edCredit: Ma et al, AJ, v. 116, 516, 1998.

JPL/Caltech NASA



S/X Defining: 212 Sources

212 “Defining” sources determine conventional rotationErrors scaled plus 250 µas RSS’edCredit: Ma et al, AJ, v. 116, 516, 1998.

JPL/Caltech NASA


Extending ICRF to 24, 32, and 43 GHzS/X-band (2/8 GHz): 3500*

Sources

*Plotted ~2400 sources with Declination uncertainty < 1 masGSFC-2008b-astro data to June 2008 (some K-band [22 GHz] data)Credit: Petrov, Kovalev, Fomalont, & Gordon, VCS-6, AJ, 2008, 136, 580.

JPL/Caltech NASA



K-band (24 GHz): 275 Sources

VLBA Data 10 sessions 2002-07Formal errors with no inflationCredit: G.E. Lanyi et al, IVS meeting 2008, St. Petersburg

JPL/Caltech NASA



Q-band (43 GHz): 132 Sources

VLBA Data 4 sessions 2002-03Formal errors with no inflationCredit: G.E. Lanyi et al, AJ, submitted Nov 2008

JPL/Caltech NASA



X/Ka (8.4/32 GHz): 339 Sources

DSN 2005-09, Dec to - 45 deg. Credit: Jacobs & Sovers, 2008

JPL/Caltech NASA



K-band (24 GHz)Celestial Frame

based on VLBA Observations

JPL/Caltech NASA



Game Plan for Observations

• Phase I - Until 32 GHz available, bracket with VLBA’s closest bands 24 GHz (K-band) 43 GHz (Q-band) - Interpolate behavior to 32 GHz (Ka-band)

- identifies likely detectable sources at Ka, time variations - maps sources to provide structure information - accuracy cross-check of X/Ka during Phase-II

• Phase II - 32 GHz and 8.4 GHz (Ka/X-band) DSN observations 34m Beam-waveguide antennas, 2 baselines:

Goldstone California to Madrid, Spain 8,400 km Tidbinbilla, Australia 10,500 km

• Phase III - Add more antennas to the network - VLBA: 32 GHz proposal in the works

- Other sites have Ka, but none with X/Ka to cal plasma

JPL/Caltech NASA



• D. Boboltz USNO

• P. Charlot Observatory of Bordeaux

• A. Fey USNO

• E. B. Fomalont NRAO-Charlottesville

• B. Geldzahler NASA HQ

• D. Gordon Goddard/NASA

• C. S. Jacobs JPL/Caltech NASA

• G. E. Lanyi JPL/Caltech NASA

• C. Ma Goddard/NASA

• C. J. Naudet JPL/Caltech NASA

• J. Romney NRAO-Socorro

• O. J. Sovers RSA Systems/JPL

• L. D. Zhang JPL/Caltech NASA

KQ VLBI Collaboration institutions:

JPL, GSFC, USNO, NRAO, and Bordeaux Observatory •

Phase I: KQ Collaborators

QuickTime™ and aTIFF (Uncompressed) decompressorare needed to see this picture.

JPL/Caltech NASA



24 & 43 GHz Observations

Mauna Kea OVRO Brewster N. Liberty Hancock

Kitt Peak Pie Town Ft. Davis Los Alamos St. Croix

(photos credit NRAO/NSF/AUI http://www.aoc.nrao.edu/vlba/html/vlbahome/thesites.html)

• VLBA ten 25m antennas

10 sessions each 24 hours

83K delay/rate obs. pairs

~65-180 sources /session

3-5 snapshots, 2 min each

400 MHz spanned band

128 Mbps record rate

• Simultaneous astrometry and imaging

• No Southern Stations

JPL/Caltech NASA



Results: 24 GHz Celestial Frame

Credit: G.E. Lanyi et al, IVS meeting 2008, St. Petersburg

JPL/Caltech NASA



∆RA, ∆ Dec: K vs. S/X

Accuracy tested vs. S/X ICRF-2 proto-type

RA: 108 µas = 0.5 nrad Dec: 207 µas = 1.0 nrad

ICRF-2 Credit: Ma et al, IERS Tech Note 35, in prep. 2009.

JPL/Caltech NASA



K-band zonal Errors: ∆Dec vs. Dec

K-band vs. S/Xtilts vs. Declinationdue to uncalibratedIonosphere and other effects

No Dual-bandPlasma cals,No stations inthe south

Without Ion cals-10 +- 0.3 µas/deg

With GPS Ion cals-5 +- 0.3 µas/deg(shown at right)

Credit: G.E. Lanyi et al, AJ, 2008, in prep., Ma et al, in prep, 2009

JPL/Caltech NASA



X/Ka Frame

from Deep Space Network 8.4/ 32 GHz observations

JPL/Caltech NASA



Phase II: X/Ka

• Phase I - Until 32 GHz available, bracket with VLBA’s closest bands24 GHz (K-band)43 GHz (Q-band)- Interpolate behavior to 32 GHz (Ka-band)

- identifies likely detectable sources at Ka- maps sources to provide structure information

• Phase II - 32 GHz and 8.4 GHz (Ka/X-band) DSN observations 34m Beam-waveguide antennas, 2 baselines:

Goldstone California to Madrid, Spain 8,400 km

Tidbinbilla, Australia 10,500 km

• Phase III - Add more antennas to the network - VLBA: 32 GHz proposal in the works

- Other sites have Ka, but none with X/Ka to calibrate plasma

JPL/Caltech NASA



X/Ka Observations

Deep Space Network 34m Beam-waveguide antennas

• 44 sessions, 9390 delays/rates, ~2K parameters very small data set by S/X standards of 5 x 10^6 delays

• nominally 24 hours each to cover full range of RA

• Record rate: now 112 Mbps (MkIV mode A) future 896 Mbps (Mk5-A) = 3X sensitivity

• 360 MHz spanned bandwidth (receiver 500 MHz, VC limited)

• 1-2 snapshots, ~2 min each

JPL/Caltech NASA



Reference Frame Modelling

X/Ka cannot achieve the results which follow without S/Xmodels! Thus, we cannot get rid of S/X for years to come!

• Rotational alignment set to S/X ICRF

• Nutations: MHB model tuned to S/X VLBI

• Tidal models tuned to fit S/X VLBI

• UTPM: GPS + S/X VLBI data

JPL/Caltech NASA



X/Ka results: 339 Sources detected

DSN data 2005-09 Dec to - 45 deg

Ecliptic

JPL/Caltech NASA



∆RA, ∆ Dec: X/Ka vs. S/X

Accuracy tested vs. S/X ICRF-2 proto-type

RA: 185 µas = 0.90 nrad Dec: 260 µas = 1.27 nrad

ICRF-2 Credit: Ma et al, in prep. 2009.

JPL/Caltech NASA



Zonal Errors: ∆Dec vs. Dec

Slope=1.76 sig

JPL/Caltech NASA



Results limited by Ka-band SNR

log10( SNR(Ka) )

Solution:

1) More bits

Mk5-A: Gbps3X SNR increase

2) Ka pointing

Potential for50% increasein SNR fromImproved pointingcalibrations

JPL/Caltech NASA



Results limited by No Ka-band Phase cal

Problem:180 psec~diurnal effect

Solution:

Ka-bandPhasecalPrototypeDemo’d --- >

Operations~2010

JPL/Caltech NASA



Troposphere Solution 2: Better Calibration

• JPL Advanced Water Vapor Radiometer ~ 1 deg beam better matches VLBI improved gain stability improved conversion of brightness temperature to path delay

• Initial demos show 1mm accuracy Goldstone-Madrid 8000 km baseline using X/Ka phase delays (Jacobs et al, AAS Winter 2005).

• A-WVRs deployed at Goldstone & Madrid Seeking funding for Tidbinbilla, Australia

• A-WVR not used yet for X/Ka catalog

-1.5

-1

-0.5

0

0.5

1

10 10.2 10.4 10.6 10.8 11

VLBI Delay Residuals DOY 200 Ka-Band DSS26-DSS55

VLBI Residuals

AWVR Calibrated VLBI

Time(hrs)

JPL/Caltech NASA



X/Ka Error Summary

• 339 sources north of -45 Declination

∆RA ~ 185 µas wRMS (0.9 nrad) ∆Dec ~ 260 µas wRMS (1.27 nrad)

• Potential for better than 100 µas accuracy: SNR: Mk-5A Gbps allows 3X sensitivity increase better pointing is realizing 50% better SNR

Instrumentation: Ka-band phase cal scheduled for ~2010, Digital Back End (RDBE)

Troposphere: Estimation with spatial/temporal correlations Advanced Water Vapor Radiometer cals

JPL/Caltech NASA



Conclusions: High Frequency ICRF

• ICRF now extended to 24 and 32 GHz ~200 µas accuracy!~300 sources north of -45 Declination

Sources unevenly observed both in numbers and geometry Caveat: models & cals still rely on S/X; still need S/X!

Observations: 24 GHz: 10 VLBA sessions, 43 GHz: 4 VLBA sessions32 GHz: 44 DSN sessions

Current Accuracy: 100-300 µas wRMS (0.5 - 1.5 nrad) Future Accuracy: 50 µas? Better?• Current limiting errors: Lack of southern stations: all bands Troposphere fluctuations: all bands Ionosphere: 24 GHz Instrumental calibrations: 32 GHz SNR: 32 and 43 GHz

Copyright 2009 California Institute of Technology. Government sponsorship acknowledged.

JPL/Caltech NASA



Troposphere Solution 1: Better Estimation

Day

few km

Night

12 3 CorrelatedFluctuations:Temporal &Spatial!

• Modify Least Squares to account for observation correlations -- both temporal and spatial

• Use Kolmogorov frozen flow model of Treuhaft & Lanyi (Radio Sci. 1987)

• This model increases the information available to the estimation process thereby enabling both 1) Reduced parameter biases 2) Reduced parameter sigmas

• Validation: Improves agreement of X/Ka vs. S/X catalogs by about 10% in Declinations (less in RA)

JPL/Caltech NASA



Dec-Dec inter-source correlations

Documents

JPL/Caltech NASA 23 Jul 2009, C.S. Jacobs Page 1 Extending ICRF to 24, 32, and 43 GHz C.S. Jacobs JPL/Caltech/NASA 23 July 2009 Extending the ICRF to Higher