The History and Future of TSI and SSI Measurements

Transcription

1 The History and Future of TSI and SSI Measurements Greg Kopp CU / LASP with thanks for contributions from: David Harber, Jerry Harder, Judith Lean, Peter Pilewskie, and Tom Woods G. Kopp, p. 1

2 TSI Is One of Four Main Inputs to Climate Models Solar Variations +0.1 K over 11-year cycle Higher S - solar input Greenhouse Gas +0.1 K per decade (trend) Smaller! - emissivity El Nino (ENSO) ±0.3 K variations Modulates! - emissivity Volcanoes -0.4 K occasionally Higher a - albedo [figure adopted from Lean, Solar Physics, 2005] G. Kopp, p. 2

3 Solar Forcing / Heating is Wavelength Dependent Ultraviolet (UV) radiation drives many atmospheric processes Near UV, visible, near infrared radiation affect surface and ocean processes Chemistry Climate Models Need SSI GISS GCM [Rind et al., 2004; Shindell et al., 2006] NCAR WACCM [Marsh et al., 2007] HAMMONIA [Schmidt and Brasseur, 2006] CMAM [Beagley et al., 1997] [Adapted from P. Pilewskie, Solar Physics, 2005] G. Kopp, p. 3

4 What Total Solar Irradiance (TSI) Data Exist? TSI Climate Data Record is nearly 30 years long and continuous. G. Kopp, p. 4

5 Total Solar Irradiance Measurement Summary Continuous and overlapping measurements are critical in maintaining a long term data record. Critical for climate Provide inputs to solar and atmospheric models NIMBUS7 ERB ( ) SMM ACRIM I ( ) ERBS ERBE ( ) UARS ACRIM II ( ) SOHO VIRGO ( ) ACRIMSat ACRIM III ( ) SORCE TIM ( ) PICARD PMO6 & SOVAP ( ) Glory TIM ( ) TSIS TIM ( )? G. Kopp, p. 5

6 What Spectral Solar Irradiance (SSI) Data Exist? SSI Climate Data Record over majority of solar spectrum began with TIMED and SORCE. G. Kopp, p. 6

7 Spectral Solar Irradiance Measurement Summary # GOES X-Ray Sensor (XRS) - since channels: 0.05 to 0.4 nm and 0.1 to 0.8 nm at s cadence; ~30% accuracy TIMED and SORCE XPS - since filter bandpasses from 0.1 to 34 nm and Ly-"; 12-24% accuracy GOES Solar X-Ray Imager (SXI) 3 filter bandpasses from 0.8 to 8 nm with 5 imaging at 1-min cadence SOHO SEM - since & 30.4 nm at 15-s cadence SNOE SXP to filter bandpasses from 2 to 35 nm; one 63-s integration every 95 min TIMED EGS - since to 200 nm with 0.4 nm resolution; one spectrum every 95 min SME Solar Ultraviolet Monitor (SUM) to 1989 (data) 115 to 302 nm G. Kopp, p. 7

8 Spectral Solar Irradiance Measurement Summary # UARS and SORCE SOLSTICE - since 1991 UARS: 115 to 425 nm with 0.2 nm resolution; data available to 2001 SORCE: 115 to 320 nm with 0.1 nm resolution; 5% accuracy; daily spectra starting 2003 UARS SUSIM to to 410 nm w/ 1 nm resolution (daily), 2% long-term precision POES SBUV/2 - since to 400 nm with 1 nm resolution, daily cadence includes Mg II core to wing SORCE SIM - since to 2400 nm with nm resolution, 4 spectra/day; 2-8% accuracy GOME - since nm, Mg II core to wing ENVISAT SCIAMACHY- since nm with 0.2 to 1.5 nm resolution SOHO VIRGO SPM - since filter bandpasses (402, 500, 862 nm) with 5 nm resolution; 1-min cadences G. Kopp, p. 8

9 Composite Time Series Created in Favorite Wavelengths Mg II composite includes data from NIMBUS7, NOAA9, NOAA11, UARS SOLSTICE, UARS SUSIM, EUMETSAT GOME, NOAA 16, and NOAA 17. [courtesy of Rodney Viereck] [courtesy of Tom Woods] G. Kopp, p. 9

10 SSI Time Series Now Available Across Solar Spectrum See poster by J. Harder et al. G. Kopp, p. 10

11 SSI to TSI Comparison Short-Term Variations Instrumental G. Kopp, p. 11

12 SSI to TSI Comparison Long-Term Variations G. Kopp, p. 12

13 What Is the Solar Cycle Spectrum and Variability? ~20% 1.2-6% 2-5% Accuracy 8% Accuracy 1% %/yr 0.006%/yr Stability [courtesy of Tom Woods] G. Kopp, p. 13

14 % over a few days What Is the Natural TSI Variability? Short duration causes negligible climate effect 0.1% over 11-year solar cycle Small but detectable effect on climate % over centuries (unknown) Direct effect on climate (Maunder Minimum and Europe s Little Ice Age) 0.1% An unequivocal link between climate change and TSI has been established over the past three decades. Magnitude of natural climate forcing needs to be known for setting present and future climate policy regulating anthropogenic forcings. Future long-term solar fluctuations, similar to historical variations, are not known from current measurements or TSI proxies. G. Kopp, p. 14

15 Science Drivers for Irradiance Requirements If accuracy is better than 1/5 of the solar cycle variability, then nonoverlapping data sets can be combined for studying long-term variations In order to precisely measure the solar cycle variations, the long-term (LT) precision / stability over 5 years should be better than 1/10 of the solar cycle variability Wavelength Range Solar Cycle Variability Req. Accuracy (1/5 SC) Req. Stability (1/10 SC / 5 yr) TSI (all #s) 0.1% 200 ppm 20 ppm/yr Visible-NIR 0.1% 200 ppm 20 ppm/yr MUV 2.5% 0.5% 0.05%/yr FUV 25% 5% 0.5%/yr EUV/XUV 100% 20% 2%/yr G. Kopp, p. 15

16 Long-Term Solar Variability The Maunder Minimum in the late 1600 s is a significant long-term change Solar output decreased % for 70 years Earth temperatures were ~ C colder than the early 1900s (Little Ice Age) Want to resolve <0.1% change over ~100 years This solar variability rate roughly matches current 0.001%/year instrument stability Improved absolute accuracy helps this detection over long time scales G. Kopp, p. 16

17 Solar Variability Desired Detection 0.1% / 100 yrs Solar Evolution? G. Kopp, p. 17

18 What Are Instrument Requirements for Trend Detection? To detect desired long-term trends the Solar Irradiance Climate Data Record requires: Data continuity with overlap (6+ months on orbit) and Instrument stability (!0.001%/year relative accuracy) Absolute accuracy (!0.01%) - or - Applies to TSI and SSI since requirements are solar variability driven Improved absolute accuracy lessens the time required for trend detection and makes record less reliant on continuity. Glory/TIM Requirements Accuracy Stability 0.01% (1 $) 0.001%/yr (1 $) G. Kopp, p. 18

19 TIM Stability Meets TSI Requirements... Intermittent, simultaneous solar measurements with cavity pairs track primary cavity s degradation. Degradation is low and follows a classic exponential. Degradation is corrected in Data Processing. Parameter LT Precision (Stability) Requirement LT Precision (Stability) Actual Value (1 $) 10 ppm / yr 8.7 ppm / 3 yrs G. Kopp, p. 19

20 ...but TSI Accuracies and Stabilities vs. Requirements There remain unexplained differences between modern TSI instruments in both absolute accuracy and long-term precision (stability). G. Kopp, p. 20

21 TSI Radiometer Facility (TRF) Validates Accuracy The TRF addresses absolute accuracy 1. Improve the calibration accuracy of future TSI instruments, 2. Establish a new ground-based radiometric irradiance standard, and 3. Provide a means of comparing existing ground-based TSI instruments against this standard under flight-like operating conditions. No flight TSI instrument has been calibrated end-to-end First facility to measure irradiance at solar power levels in vacuum at desired accuracies See poster by D. Harber et al. G. Kopp, p. 21

22 TSI Record Currently Relies on Continuity Until absolute accuracy is established The Climate Change and Variability Panel believes that the current strategy for continuity, of ensuring overlap between measurements, should be continued as recommended by GCOS (2003), CCSP (2003) and others (Ohring et al., 2005). NRC Report 2007 Earth Science and Applications from Space: National Imperatives for the Next Decade and Beyond? G. Kopp, p. 22

23 SSI Accuracies and Stabilities vs. Requirements ~20% 1.2-6% 2-5% Accuracy 8% Accuracy 1% %/yr 0.006%/yr Stability G. Kopp, p. 23

24 Long-Term Change" "Detection Via Instrument Stability Solid grey lines show possible solar variability with time Want to detect Maunder Minimum levels of 0.1% / 100 years (0.001%/yr) Desired Detection 0.1% / 100 yrs G. Kopp, p. 24

25 Long-Term Change" "Detection Via Instrument Stability Solid grey lines show possible solar variability with time Want to detect Maunder Minimum levels of 0.1% / 100 years (0.001%/yr) Current TSI instrument stabilities are comparable to desired longterm solar variability detection desired, making detection marginal. Current SSI stabilities are insufficient for long-term trend detection. G. Kopp, p. 25

26 Long-Term Change" Detection Via Instrument Accuracy With good absolute accuracy (horizontal portions of colored lines), can detect this rate of change given time The better the absolute accuracy, the less time required for trend detection Good absolute accuracy frees data record from reliance on continuity Times required for detection Uncertainties are 1-$. A 3-$ detection would require 3 times as long. Improved instrument accuracies speed trend detection G. Kopp, p. 26

27 How Long Must Record Rely on Continuity and Stability? Answer: Until absolute accuracy alone can maintain record For a 3-! detection certainty with a 100 ppm (1-!) instrument, this is ~30!years. Time required for 1-$ detection Uncertainties are 1-$. A 3-$ detection would require 3 times as long. Improved instrument accuracies speed trend detection G. Kopp, p. 27

28 What Is the Maximum Allowable Probability of a Data Gap from a Scientific Perspective?? Need 5-yr launch spacing This spacing is similar to the 4-year average intervals of the 7 TSI instruments in the 29-year record that contribute to current composites. G. Kopp, p. 32

29 Future Needs for Solar Irradiances for Climate Solar irradiance climate data record requirements Absolute accuracy of 0.01% or Stability of 0.001%/yr and continuity Can currently easily detect changes over solar rotation and solar cycle TSI provides long-term record for current solar forcing sensitivities New SSI measurements adding knowledge of solar activity causing irradiance changes and effects on Earth s atmosphere Detection of Maunder Minimum or long-term trend of 0.1% / 100 yrs TSI: Detection marginal with either absolute accuracy (requires 35 yrs 1-$ currently) or stability (0.001%/yr over >1 solar cycle) TSI record currently relies on instrument stability and continuity Imminent improved absolute accuracies will shorten detection time SSI: Detection requires improved stability to benefit from continuity Absolute accuracy improvements to 0.1% likely, facilitating long-term trend detection over 100 yrs This is not too different from how the TSI data record began... G. Kopp, p. 33