Werner SchmutzPMOD/WRC, Switzerland

TOSCA WorkshopBerlin, May 14, 2012

Measurements of TSI and SSI

Werner Schmutz 2

Overview

14. May 2012

• Total Solar Irradiance

• Absolute calibration (first light PREMOS)

• Composites (relative calibration)

• Spectral Solar Irradiance (SSI)

• SIM/SORCE

• VIRGO/SOHO

PICARD

PREMOS – SOVAP – SODISM

Werner Schmutz 314. May 2012Total Solar Irradiance

Filter Radiometers

Werner Schmutz 4

TSI calibration

14. May 2012

PREMOS A is the first and only radiometer in space with a SI-traceable irradiance calibration in vacuum

Traceable to the irradiance calibration facility at LASP in Boulder (TRF)

Werner Schmutz 5

Traceability of PREMOS-TSI

14. May 2012

PREMOS B

Comparison to cryogenic rad.

(power in vacuum)

NPL

Comparison to cryogenic rad.

(power and irradiance in

vacuum)TRF @ LASPPREMOS A3

Werner Schmutz 6

Uncertainty of the calibration

14. May 2012

+ absolute uncertainty of TRF facility (70 ppm)

= uncertainty of TRF comparison (220 ppm)

PREMOS APREMOS A

TRF radiometerTRF radiometer

Werner Schmutz 7

Calibration uncertainty budget

14. May 2012

Traceable via TRF, LASP, Boulder to NIST

• Irradiance in vacuum PREMOS A uncertainty: ± 280 ppm

(± 0.4 W/m2)

Table compiled by Greg Kopp for an ISSI workshop March 2012

Flight-spare recalibrations

Werner Schmutz 814. May 2012

Comparison PREMOS – TIM

Werner Schmutz 9

Status of PREMOS-TSI

14. May 2012

„PREMOS is in excellent health“

o PREMOS-TSI is the most accurate absolute measurement;

±0.4 W/m2 or ±290 ppm

o After 2 years, PREMOS-TSI has at most 50 ppm relative deviation to TIM/SORCE.

The Future of TSI observations:Are relative observations sufficient?

Werner Schmutz 1014. May 2012

Werner Schmutz 11

There are three TSI composites

20.12.2011

TSI-composites normalized

Werner Schmutz 1220.12.2011

!normalized 2004-2005

DIARAD

PMOD, ACRIM

Composite 1996-2010

Werner Schmutz 1314. May 2012

!0.2 W/m2

± 0.2 W/m2 / 10-years

Werner Schmutz 14

Is there a long-term trend?

14. May 2012

Fröhlich 2009, A&AL 501, L27-L30

Werner Schmutz 15

Could we detect a long-term trend with a composite?

14. May 2012

Werner Schmutz 16

Requirements for a TSI monitoring

14. May 2012

„Any plan to rely on an unbroken chain of measurements is broken“

o Not only because of a potential gap;o But mainly because of the uncertainty

is continuously increasing with time !

Accurate absolute measurements are required !

Werner Schmutz 17

Requirements for a TSI monitoring

14. May 2012

Accurate absolute measurements are required: Nowadays possible !

But we certainly also want to assess the variations of TSI and therefore, we still need to aim for continues and overlapping data !

Werner Schmutz

TSI monitoring today …

1814. May 2012

Presently, 4 operational space experiments observing TSI:

- VIRGO (launched 1995)- ACRIM III (launched 2000)- TIM (launched 2003)- PREMOS (launched 2010)

Part II:Spectral Solar Irradiance

Werner Schmutz 1914. May 2012

Werner Schmutz 2014. May 2012

The open question !

Is the SIM observation really correct?

Or is it rather a degradation problem?

The bands 410-470 and 480-730 nm are anti-correlated to TSI variations

Compensated by larger (than TSI) UV variations

Werner Schmutz 2114. May 2012

Anti-correlation in models

Contrast between active and quite Sun (SSN 150 vs SSN 0)

Black: Bright+DarkRed: bright componentsBlue: dark components

Werner Schmutz 2214. May 2012

VIRGO and PREMOS bands

Werner Schmutz

215 nm PREMOS vs SOLSTICE

2314.05.2012

Independant correction of PREMOS

Strong correlation of13.5 and 27 days modulation

PREMOS sampling is faster Rotational modulation more accurate

PREMOS

SOLSTICE

PREMOS

NIST SSI workshop, February 2012

VIRGO SSI time series 1996 - now

Christoph Wehrli & VIRGO TeamPMOD/WRC Davos

An attempt to assess instrument degradation in a self consistent way by:

• referring operational measurements to occasional backup operations

• correcting the backup channel by initial ageing of operational channel

NIST SSI workshop, February 2012

VIRGO Sun Photometers

• Interference filter radiometer with 3 channels centered at 862nm, 500nm and 402nm (R,G,B); FWHM bandwidths 5nm; silicon PD detectors; rad-hard windows.

• Active (SPM-A) and Backup (SPM-B) instruments– SPM-A: exposed continuously for helioseismology

application– SPM-B: exposed rarely for solar spectral irradiance

measurements• Calibrated by EG&G FEL lamps, NBS 1973

traceable

NIST SSI workshop, February 2012

VIRGO SPM: Level1 data

1996 1998 2000 2002 2004 2006 2008 2010 20120

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

2

spec

tral

irra

dia

nce

[W

m-2

nm

-1]

VIRGO SPM-A

70%

20%

5%

Sen

sitivity after 5825 days

1996 1998 2000 2002 2004 2006 2008 2010 20120

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

2

spec

tral

irra

dia

nce

[W

m-2

nm

-1]

VIRGO SPM-B

Number of Backups 178Total exposure time 2.6 days

SPM-B:Number of backups 178Total exposure time 2.6 days

NIST SSI workshop, February 2012

1996 1998 2000 2002 2004 2006 2008 2010 20120.990

0.992

0.994

0.996

0.998

1.000

1.002

1.004

1.006

1.008

1.010

rel.

Var

iati

on

(T

SI*

5)

VIRGO SPM-B and TSI

Raw variations of SPM-B

VIRGO-SPM-B changes are about 5 times larger than the (real) solar variations in TSI.

Solar cycle 24 SSI variation is not obvious (smaller), hidden in instrumental ageing.

5*(TSI/<TSI>) - 4

NIST SSI workshop, February 2012

Initial ageing of SPM-A

Steep degradation during first hours!

17.Jan 18.Jan 19.Jan 20.Jan0.990

0.995

1.000

1.005

1.010

1.015

1.020

1.025

1.030

1.035

Date 1996

rel.

Var

iatio

n

VIRGO SPM-A First Light

Linear degradation during first month

{Commissioning activities until 29.03.1996}

28.Jan 04.Feb 11.Feb 18.Feb 25.Feb 03.Mar0.990

0.995

1.000

1.005

1.010

1.015

Date 1996

rel.

Var

iatio

n [T

SI*

100]

VIRGO SPM-A Operational

∂TSI (*100)

NIST SSI workshop, February 2012

Ageing of SPM-A and SPM-B versus exposure time

0 0.5 1 1.5 2 2.50.985

0.990

0.995

1.000

1.005

1.010

1.015

exposure time [d]

rel.

Var

iatio

n

VIRGO SPM A & B

polynomial fit SPM-A

NIST SSI workshop, February 2012

SPM-B corrected by operational degradation of SPM-

A

Instrumental effects dominating over solar cycle

1995 1997 2000 2002 2005 2007 2010

0.992

0.994

0.996

0.998

1.000

1.002

1.004

1.006

1.008

rel.

Var

iatio

n

VIRGO SPM-B corrected for SPM-A operational degradation

1%

NIST SSI workshop, February 2012

Empirical Approach

• SSI timeseries represent a mixture of Solar Cycle and instrumental effects

• Active & Backup SPM degrade differently in time or exposure time

• Linear correction accounts for probable decline of SSI, i.e. first order estimation of instrumental effect.

• Exponential correction eliminates most of solar cycle variation as well

2000 2002 2004 2006 2008 2010 20121.815

1.820

1.825

1.830

SP

M-B

500

[Wm

-2n

m-1

]

VIRGO SSI

0.5%

NIST SSI workshop, February 2012

Linear vs. Exponential Detrending: what does it to

TSI ?

1000 2000 3000 4000 5000 6000

1363.0

1364.0

1365.0

1366.0

1367.0

1368.0

MissionDay

TS

I

1000 2000 3000 4000 5000 6000

-3.0

-2.0

-1.0

+0.0

+1.0

+2.0

MissionDay

TS

I (re

sid

ual

s)

1363 1364 1365 1366 1367 1368-2.5

-2.0

-1.5

-1.0

-0.5

+0.0

+0.5

+1.0

+1.5

TSI

TS

I (re

sid

ual

s)

Lslope 0.54 [ 0.47 0.62]Eslope 0.32 [ 0.27 0.36]

NIST SSI workshop, February 2012

Linear Detrending 2000-2012

1000 2000 3000 4000 5000 60001362

1364

1366

1368

MissionDay

TS

I

1000 2000 3000 4000 5000 6000-5

0

5

MissionDay

SS

I500

(res

)

1363 1364 1365 1366 1367 1368-5

-4

-3

-2

-1

0

1

2

3

TSI

SS

I500

(res

idu

als)

slope 1.20 [ 1.02 1.38] [1/µm]

NIST SSI workshop, February 2012

Summary

Normalization of SPM-A by SPM-B:– Larger than expected variations of Backup channel

“instrumental effects”• Rapid initial degradation in Active channelversus• ‘Early increase’ of Backup (not observed in operational

channel)

Empirical correction of SPM-B: fitting degradation in time with:

– Linear or exponential detrending yields positive correlation with solar cycle (TSI) in

all 3 visible channels !!!

None of VIRGOS’s wavelength bands 862 nm, 500 nm, 402 nm

isanti-correlated to TSI variations

Thank you for your attention

Werner Schmutz 3514. May 2012

PREMOS PICARD

NIST SSI workshop, February 2012

Alternative analysis including proxies

(C. Fröhlich, EGU 2011)

Empiric correction versus Time (double exponential), Temperature (linear + Boltzmann), TSI and Mg-II Index.

SORCE