Greg Kopp, p. 1ASIC 3 – 16 May 2006TSI Accuracy The Accuracy of Total Solar Irradiance...
-
Upload
cheyenne-vowels -
Category
Documents
-
view
216 -
download
2
Transcript of Greg Kopp, p. 1ASIC 3 – 16 May 2006TSI Accuracy The Accuracy of Total Solar Irradiance...
Greg Kopp, p. 1ASIC3 – 16 May 2006 TSI Accuracy
The Accuracy of Total Solar Irradiance Measurements
Greg Kopp, CU/LASP, Boulder, CO
Greg Kopp, p. 2ASIC3 – 16 May 2006 TSI Accuracy
Impacts of Solar Irradiance on Earth
Sun - Climate Correlations
.
.
.Across All Time
Scales
L. Hood, Univ. ArizonaSORCE Meeting 2003
Secular Trends 11-Year Solar Cycle
27-Day Solar RotationSultan Hameed, Stony Brook
SORCE Meeting 2004
(a proxy of TSI)
Greg Kopp, p. 3ASIC3 – 16 May 2006 TSI Accuracy
27-Year TSI Data RecordNone of these instruments is
calibrated end-to-end for irradiance.
Greg Kopp, p. 4ASIC3 – 16 May 2006 TSI Accuracy
Summary of Stated Instrument Accuracies
Uncertainties are 1-
Instrument
TSI Value [W/m^2]
Stated Uncertainty
[ppm]
ERB (NIMBUS 7) 1371.9 5000ACRIM I 1367.5 1000ACRIM II 1364.2 1000ACRIM III 1366.1 1000
ERBE 1365.2 833VIRGO 1365.7 1000
VIRGO-PMO 1365.7 1204VIRGO-DIARAD 1366.4 470
DIARAD-like 1366.4 600SORCE/TIM 1361.0 350
Greg Kopp, p. 5ASIC3 – 16 May 2006 TSI Accuracy
Climate Record Needs: 100 ppm Absolute Accuracy...
• Dick Willson: “...uncertainties of 100 ppm or less would be required to produce a TSI record with sufficient traceability over the multi-decadal to centennial time scales for climate change and solar physics investigations without employing and overlapping, redundant measurement...”
No TSI instrument has yet achieved this level of accuracy.
Greg Kopp, p. 6ASIC3 – 16 May 2006 TSI Accuracy
Two primary TSI composites differ by 40 ppm/yr. Caused by 2 years of marginal quality data – not even a gap!
Climate Record Needs: ... or 10 ppm/yr Stability and Overlap
?
Greg Kopp, p. 7ASIC3 – 16 May 2006 TSI Accuracy
27-Year TSI Record Relies on Continuity
Current climate record plan relies on continuity and mission overlap.
But why the offsets?
Greg Kopp, p. 8ASIC3 – 16 May 2006 TSI Accuracy
TSI Accuracy Workshop
• Organizer: Jim Butler, NASA/GSFC
• Location: NIST Gaithersburg, MD
• Dates: 18-20 July 2005
• Attendees– Representatives of several TSI instruments
• ACRIM I, II, and III
• ERBS/ERBE
• SORCE/TIM
• VIRGO/PMO
• VIRGO/DIARAD & SOLCON
– NIST, NASA
• Approach– Day 1: Accuracy (“the Day 1 Problem”)
– Day 2: Stability
– Day 3: Improved or current calibration facilities
• Dick Willson: “We haven’t had a meeting like this in 20 years!”
TSI Accuracy Workshop Questions
Multiple Radiometers Track Changes and Indicate
Consistency With Stated Uncertainties
• Review Instrument Designs– Are there systematic differences that could
cause TSI offsets?
• Review Calibrations & Uncertainties– How accurately is each instrument calibrated?
How well are uncertainties estimated?
– How well are degradation and stability understood?
– What were goals and actuals?
• Intra-instrument Consistency– Do intra-instrument cavity comparisons agree
with stated uncertainties?
– What ground calibrations or facilities would improve the future TSI record?
Agenda
Absolute Accuracy
Stability
Calibrations
Greg Kopp, p. 11ASIC3 – 16 May 2006 TSI Accuracy
Summary of Instruments
Instrument Cav #
Cavity Layout
Precision Aperture Position
V Mon
I/R Mon
Black Black Type
Active Therm Control
Dark Meas. Freq.
Comments
ERB 1 ? internal ? ? paint specular no every meas.
ACRIM I 3 cones, front to back
internal yes yes paint specular no low
ACRIM II 3 cones, front to back
internal yes yes paint specular no low
ACRIM III 3 cones, front to back
internal yes yes paint specular no low
ERBE 1 cones, front to back
internal yes yes paint specular no every meas.
one 3-min meas every 2 wks
VIRGO-PMO 2 inverted cones, front
to back
internal yes yes paint specular no
(at L1)
none low-freq. shutter
VIRGO-DIARAD
2 cylinders, side by side
internal yes yes paint diffuse no
(at L1)
none
TIM 4 cones, side by side
front of instrum.
no no NiP diffuse yes every orbit
pulse-width modulation
Greg Kopp, p. 12ASIC3 – 16 May 2006 TSI Accuracy
Reviewed Uncertainties
SORCE/TIM
Relative W/m2
Area 0.000425 0.58
Thermal efficiency 0.000130 0.18
Eectrical. Power 0.000150 0.20
Cavity absorption 0.000030 0.04
Total 0.000735 1.00
RSS 0.000470 0.64
VIRGO/DIARAD L
ACRIM
VIRGO/PMOComponent Value u c (u*c)^2Area N/APclosed 45 mW 0.0000045 5.00E+04 0.050625Popen 17 mW 0.0000017 5.00E+04 0.007225CNE 1 5.00E-04 1.40E+03 0.49CR N/A 7.00E-05 1.40E+03CSt N/A 1.00E-04 1.40E+03CLH N/A 3.00E-05 1.40E+03CApH N/A 5.00E-04 1.40E+03Cdiff N/A 1.00E-04 1.40E+03WRR-Factor 1 6.00E-04 1.40E+03 0.7056WRR/SI 1 9.00E-04 1.40E+03 1.5876
2.84105Uncertainty abs 1.6855 W/m2Uncertainty rel 1685.5 ppm95% Uncertainty 3371.1 ppm
Uncertainty of the PMO6V WRR/SI traceability @ 1400W/m2
Correction Value [ppm] SORCE Worst Case
Distance to Sun, Earth & S/C 33,537 0.1 0.1Doppler Velocity 57 0.7 0.7Shutter Waveform 100 1.0 1.0Aperture 1,000,000 30 651 Diffraction 452 47 47Cone Reflectance 250 54 108Non-Equivalence, ZH/ZR - 1 7, AC 23 46Servo Gain 16,129 0.0 0.0Standard Volt + DAC 1,000,000 7.0 100 Pulse Width Linearity 1,000,000 186 300Standard Ohm + Leads 1,000,000 17 25Dark Signal 2,693 10 25Scattered Light & IR 100 25 50Pointing 10 25Measurement Repeatability (Noise) 1.5 2.0Uncertainty due to Sampling 12 12Total RSS 206 738Cone Agreement Accuracy 301 301
Greg Kopp, p. 13ASIC3 – 16 May 2006 TSI Accuracy
Instrument
TSI Value [W/m^2]
Stated Uncertainty
[ppm]
ERB (NIMBUS 7) 1371.9 5000ACRIM I 1367.5 1000ACRIM II 1364.2 1000ACRIM III 1366.1 1000
ERBE 1365.2 833VIRGO 1365.7 1000
VIRGO-PMO 1365.7 1204VIRGO-DIARAD 1366.4 470
DIARAD-like 1366.4 600SORCE/TIM 1361.0 350
TSI Instrument Uncertainties & Intra-Instrument Variations
Uncertainties are 1-
InstrumentTSI Value [W/m^2]
Stated Uncertainty
[ppm]
Cavity Variations [ppm]
ERB (NIMBUS 7) 1371.9 5000 - ACRIM I 1367.5 1000 511ACRIM II 1364.2 1000 2046ACRIM III 1366.1 1000 1036
ERBE 1365.2 833 - VIRGO 1365.7 1000 2271
VIRGO-PMO 1365.7 1204 299VIRGO-DIARAD 1366.4 470 2858
DIARAD-like 1366.4 600 1612SORCE/TIM 1361.0 350 301
Greg Kopp, p. 14ASIC3 – 16 May 2006 TSI Accuracy
Possible Causes of Differences in Absolute Values
• Underestimated Uncertainties: Is this simply the state of the art in these radiometric measurements, with all uncertainties being underestimated?
• Apertures: Measurements from different facilities have greater variations than stated aperture measurement uncertainties.
– Does not account for 0.3% TSI differences
– Does not explain inter-cavity variations within single instrument
• Applied Power: The TIM uses DSP-controlled pulse width modulation while other radiometers apply DC power.
– Very unlikely to have 0.3% difference
• Optical – Scatter Prior to Limiting Aperture: Instruments with oversized (non-limiting) aperture near front of instrument allow much more sunlight into instrument.
– The TIM precision aperture and shutter are at the front of the instrument, so this is a difference.
– Scatter will erroneously increase the signal through the limiting aperture.
• Optical – Diffraction: This is a 0.12% effect in ACRIM and is not corrected
• Darks: Uncertainties in dark corrections are large. – These are large corrections, depend on FOV, and vary with temperature.
– Darks are not measured regularly on several instruments.
• Aperture Heating: Uncertainties in heating due to different aperture materials, conduction, mounting, emissivities
Greg Kopp, p. 15ASIC3 – 16 May 2006 TSI Accuracy
Roadmap for Future Improvements
• Complete aperture comparison measurements– Include ACRIM apertures in NIST aperture comparison
• Power comparison– NIST power comparison to trapped diode transfer standard
– NPL power trap comparison
– 0.05% accuracy
• Scatter/diffraction measurement– NIST to monitor changes in signal as beam expands to overfill entrance
aperture
• Would like an irradiance comparison against an absolute reference radiometer with 0.01% accuracy– JPL Table Mountain Observatory inter-comparisons are merely relative
– PMOD World Radiation Reference is linked to an absolute scale but without desired accuracy
– NASA’s Glory program is creating the TSI Radiometer Facility to compare TSI instruments on an absolute scale
Greg Kopp, p. 16ASIC3 – 16 May 2006 TSI Accuracy
Scatter Can Erroneously Increase Signal
Baffles
Cavity
Precision Aperture
View-Limiting Aperture
Precision Aperture
View-Limiting Aperture
TIM Cutaway
ACRIM III Cutaway
View-Limiting Aperture
Precision Aperture
View-Limiting Aperture
Precision Aperture
SunlightSunlight
Additional light allowed into
instrument can scatter into cavity
• All instruments except TIM put primary aperture close to the cavity– Could cause systematically higher TSI values reported
Majority of light is blocked before
entering instrument
Greg Kopp, p. 17ASIC3 – 16 May 2006 TSI Accuracy
Diffraction Can Erroneously Change Signal
View-Limiting Aperture
Precision Aperture
View-Limiting Aperture
Precision Aperture
SunlightSunlight
Failure to correct for light diffracted
into cavity erroneously
increases signal
• All instruments except TIM put primary aperture close to the cavity
Failure to correct for light diffracted
out of cavity erroneously
decreases signal
Greg Kopp, p. 18ASIC3 – 16 May 2006 TSI Accuracy
TSI Instrument Uncertainties & Intra-Instrument Variations
NIST calculates diffraction should lower these results.
Greg Kopp, p. 19ASIC3 – 16 May 2006 TSI Accuracy
TSI Instrument Uncertainties – With Diffraction Correction
Correction not yet approved or applied by ACRIM or ERBE Teams
Greg Kopp, p. 20ASIC3 – 16 May 2006 TSI Accuracy
Address Applied Power: Trap Diode Power Comparison
• NIST and LASP are preparing to do optical power comparisons between a transfer standard and ground TSI instruments– NPL has done similar power comparisons before, but in air
stabilized laser
vacuum window
beamsplitter (1%)
TSI instrument
trap diode
vacuum chamber
Greg Kopp, p. 21ASIC3 – 16 May 2006 TSI Accuracy
“Benchmark observations of total solar irradiance and spectrally resolved solar irradiance to an accuracy of 0.03 percent referenced to NIST standards are required to elucidate the origin of climate change.”
[NRC Committee on Earth Science and Applications from Space]
Intend to Create Facility for TSI Calibrations
• Have strong community support for a NIST-traceable cryogenic radiometer facility to calibrate TSI instruments to ~100 ppm absolute accuracy (TSI Accuracy Workshop 2005, NEWRAD 2005)
• Glory Science will create this facility to enable TSI instrument comparisons against an absolute standard
Greg Kopp, p. 22ASIC3 – 16 May 2006 TSI Accuracy
TSI Instrument in Vacuum Tank
Cryogenic Radiometer
1-D Stage
Light Source
TSI Radiometer Facility Layout
Radiometer RequirementsParameter Requirement Nominal Measured Irradiance Level 1360 W/m2 Nominal Measured Power Level 68 mW (i.e. 8 mm diameter aperture) Accuracy (Relative Standard Uncertainty, k=1) for measurements of nominal inputs
100 ppm
Modes Irradiance, Power Noise, k=1 <2 ppm (0.14 µW) Measurement Time 10 min Electrical Power Linearity 10 ppm (0.68 µW) k=1 from 1 mW to 100
mW Spectral Range 532 nm nominal, but operational over 400-
1550 nm Optical Setup capable of sharing single optical window
with TSI instrument Vacuum Setup capable of sharing common vacuum with
TSI instrument Minimum Continuous Operational Time 13 hrs