The SDS, and Variations of the Solar Diameter: Flight 11 (October 16, 2009) Sabatino Sofia Yale University New Haven, CT, USA.

The SDS, and Variations of the Solar Diameter: Flight 11 (October 16, 2009)

Sabatino SofiaYale UniversityNew Haven, CT, USA

The SDS is a balloon-borne metrologic instrument that measures the SOLAR DIAMETER AND ASPHERICITY

It has flown in Fall 1992, 1994, 1995, 1996, and 2009 from Fort Sumner, NM

Future flights, especially connected with PICARD, are being considered.

RESEARCH TEAM

American University (U.J. Sofia)

CNRS/S.d’A (G. Thuillier, D. Djafer)

CSA (Stella Melo)

NASA/GSFC (W. Heaps, L. Twigg, E. Georgieva)

Yale University (S. Sofia)

SDS Principle

Optical System

The stability of the optics needed by thisinstrument is of the level commonly used inoptical interferometry. For this reason, theballoon flight version currently utilizessimilar techniques (molecular bonding) andmaterials (quartz and Zerodur). Thus theoptics consist of:

1. beam splitting wedge,2. Cassegrainian telescope,3. relay lens to achieve the required

focal length, and4. detector support.

The Beam Splitting Wedge

The wedge consists of two fused silica flatsseparated by an annular silica ring polishedto an angle of about 1000 arc-s. Molecularcontact bonding is used to hold the assemblyin alignment. The surfaces are flat to 1/50wave at 630 nm and have dielectric coatingsto define the bandpass and reduce the solartransmission to an acceptable level. Themirrored surfaces have a high reflectivity (R>.9) so that the intensities of the two imagesare approximately in the ratio of 5:4.

The Telescope

A 17.8 cm aperture ruggedized Questartelescope is used with a reduced apertureof 12.7 cm to accommodate the effectivewedge aperture, and has a focal length of2.5 m. Its optical elements are made offused silica and Zerodur, and the mainbody and mounting parts are made ofInvar to minimize thermal effects.

The Relay Lens

To provide the plate scale to match theresolution of available Charge CoupledDevice (CCD) detector elements, amagnification of the image is required. Amagnification by a factor of 8 isprovided by a multi-element Barlowlens, to an effective focal length of 20 m.

A NEW OPTICALLY CONTACTEDBARLOW LENS IS IN FABRI CATION,SUPPORTED BY CNES

The Detectors

The CCDs are mounted on a ceramiccircuit board with a thermal coefficient ofexpansion near that of the CCD cases anda high thermal conductivity.

The CCDs are Texas Instrument virtualphase 1728 element devices having pixeldimensions of 12.7 x 12.7 microns. Abandpass filter provides the properwavelength of light to match the solarinput. This system provides a resolutionof 0.11 arc s per pixel.

S = (D – d)/F = W (1 – d/D)

Differences between existing diameter measurements

• Ground-based vs. space based• Wavelength of observation• Analysis Method• Calibration

These issues are extensively described in a paper byDjafer, Thuillier and Sofia, ApJ, 676, 651, 2008.

Ground-based measurements are affected by terrestrial atmosphere

Seeing is 1”-4”, and we need sensitivity of mas.This cannot be simply solved by statistics, since atmospheric turbulence is not random.

Once you go to space, there are 2 measurements:

SoHO/MDI

SDS

SoHO/MDI is not a metrologic instrument. It has not been calibrated before launch, and cannot be calibrated in space.

The only metrologic instrument to measure the solar diameter is the SDS.

The MDI results claim an accuracy of a few mas. over more than a decade

This is equivalent to knowing the effective focal length to an accuracy of a few microns over this time period.

The instrument is frequently refocused.Large corrections are made through “characterization”i.e. on the basis of a “thermal model” not calibrated beforeflight.There are corrections for agingEtc.

The previous corrections are made in addition to distance corrections that are well understood.

By contrast, the SDS can separate instrumental changes (regardless of its origin) from changes of the solar diameter.

PICARD can calibrate scale in two separate ways:

Stellar pairsWedges (similar to SDS)

Philosophical difference:

SDS-PICARD: We determine the scale (arc sec/mm) and its changes regardless of the cause.

MDI, RHESSI *, etc: They correct for each known instrumental process, and assume the rest is solar change.

* RHESSI only claims accuracy for the asphericity determination.

Determining the solar diameter is a complex process

When we look at the image of the Sun obtained in any detector, we do not see the Sun, but an image obtainedthrough a typically complex optical system.

Besides the peculiar solar issues described earlier, we also have general optical distortion effects that we have long ago learned from stellar astrometry

ORIGIN OF PRINCIPAL OPTICAL DISTORTION

TILT OF THE DETECTOR PLANECOLORCHARGE TRANSFER EFFECTCLASSICAL DISTORTION COMAPOSSIBLE CROSS TERMS

subtract background level

read data (HK+CCD)

apply photometric coefs

preliminary edge detection

subtract ghost images

get next cycle filename

final edge detection

transform to focal-plane x,y

correct for distortion

fit direct & reflected circles

find minimum gap

output gap, sep, Rd, Rr, etc.

odd/even offset & spike filter

flight-cycle filename list

output relevant HK

“New” Yale SDS Flight Data Reduction Pipeline

distortion coefs

photom. coefs

PDS xy's of CCDs

normalize to 1 AU

cull/average by rotation angle

correct for refraction

cycle#, gap, sep, Rd, Rr, etc.

output diameter & oblateness

“New” Yale SDS Flight Data Reduction Pipeline(cont.)

cycle#, HK (time, rotat., etc)

Output from program #1 becomes input for program #2:

When you need sensitivity of 1 part in 106, and very complex analysis procedures, you run the risk that the results reflect the method rather than the object measured.

The incompatible results claimed by different authors regarding the solar diameter confirm this.

WE NEED TO PROCESS THE DATA TO THE POINTOF PRODUCING THE DIAMETER, AND THEN PUBLISH IT AS SOON AS POSSIBLE FOR THE BROAD SCIENTIFIC COMMUNITY TO EXPLORE ITS IMPLICATIONS

SDS/PICARD SYNERGY

PICARD measurements are complex, and their analysis is very demanding.

Analyzing the SDS data obtained before the PICARD launch has assisted development and testing of the PICARD algorithms. This has accelerated PICARD science productivity.

The 2009 SDS flight, added to the earlier and the proposedfollow up flights during and after PICARD will enhance the science value of the mission by:

Providing validation of measurementsAllowing normalization of past resultsAllowing the extension of the PICARD mission beyond itslifetime.