EARTHSHINE OBSERVATIONS OF VEGETATION AND IMPLICATION FOR LIFE DETECTION ON OTHER PLANETS

1
Earth viewed from 6.4 109 km Voyager-1, 14th
feb. 1990
Shall we be able to detect life on an
unresolved Earth-like extrasolar planet ?
2
LOOKING FOR THE VEGETATION SIGNATUREIN THE
EARTHSHINE SPECTRUM

• Luc ARNOLD
• Observatoire de Haute-Provence, CNRS,
• 04870 Saint-Michel-lObservatoire, France
• Luc.Arnold _at_ oamp.fr

Earthshine workshop, Lund, Sweden, 14 January 2008
3
1. Context

• Preparation of High contrast imaging space
  missions
• Darwin (infrared) Despite of ESA Cosmic Vision
  conclusions
• TPF-C coronagraph (visible)
• More See-Coast, Luciola, etc.

These instruments will (hopefully) provide us
with 1/ unresolved images of extra-solar
planets in the HZ 2/ spectra to give us first
insights into planet chemistry
4

• What shall we look for ?
• Look for the signature of molecules in the
  atmosphere (biogenic products ?)
• O2, O3, CH4, H2O, CO2
• Look for biologic activity
• the planet colour vegetation, pigments
• Look for missing photons used in a photosynthesis
  
• Possible artifacts minerals, rocks

5
What are we looking for ?
The Vegetation Red Edge (VRE) VRE ( SRED
SIR) / SRED
(Clark 1999)
Light scattered (reflected / transmitted) by
leaves
6
2. Basics on Moons Earthshine

• Spectroscopy of Moons Earthshine (ES)
• -gt provides an integrated (disk-averaged) Earth
  spectrum
• -gt equivalent to the spectrum of an unresolved
  Earth

ES spectrum SRERMTE Moonlight spectrum
SRMTE Earth reflectance RE ES/Moonlight S
Sun radiance RM Moon reflectance RE Earth
reflectance TE Earth atmosphere transmittance
7
3. Review of results 3.1 Results from
observations
8
A blue Earth, O3, O2, H2O, vegetation signature
(Arnold et al 2002)
Clark 1999
morning
evening
9
Vegetation signature VRE 4 to 10 /-3 !
Test on Vega VRE -1 /- 2 ! Test on Moon VRE
0 /- 3
10
Woolf et al. 2002 VRE6
11
Turnbull et al. 2006 Woolf 2002 Vis spectrum
near IR data
12
Seager et al. 2005 VRE ?
13
Montanes-Rodrigez et al. 2005 2006 Observation
19 nov. 2003 VRE 0 to 3
14
Hamdani et al. 2006 (AA) VRE 1 to
4 Observation from Chile (NTT_at_ESO with EMMI)
Dark Earth in near-UV (lt360nm) Ozone absorption
! Rayleigh visible down to 360nm
15
3.2 Results from models

• Des Marais et al. 2000  2, or more if
  vegetated areas in view 
• Schneider 2000 VRE 5 (unpublished)
• Arnold et al. 2002 VRE 7 to 12

16

• Results from radiative transfer codes
• Tinetti et al. 2006 (Astrobiology)
• vegetation is a small feature as soon as clouds
  are present.
• VRE 4 at 60 cloudy Earth (from fig. 12,
  paper II)
• Paillet et al. 2006 (PhD)
• vegetated areas should be 10 from visible
  cloud-free surface to be detectable in S/N20
  spectra

17
4. What did we learn from ES observations ?

• Vegetation has a sharp edge at 700nm
• But is easily hidden by clouds (60 cloud cover)
• gt VRE remains a small feature, observed or
  simulated to be 0 to 10 ( typically 0 to 3 - 4
  ) above the red continuum (phase, clouds,
  seasons, etc)
• Earthshine data reduction remains difficult
  (although possible with efforts)
• Moon low above horizon, high air-mass
• Several calibration spectra sky (faint) Moon
  crescent (very bright!)
• Pollution of ES by light scattered from Moon
  crescent (Rayleigh scat.)
• Moon colour dependance with phase earthshine and
  crescent are not observed at the same phase
  angles !
• Etc.
• gt possible bias (see Hamdani et al. 2006)

18
An integral field spectrograph for Earthshine
calibrated spectroscopy

• A spectrum for each patch on the moon
• Full earthshine surface is used -gt better S/N
  wrt to a single slit spectrograph
• Would allow to characterize the color dependence
  of Moons phase function
• (in fact the extension of Qiu 2003 work done for
  broadband photometry)

19

• Is it possible to built integrated spectra from
  satellite data?
• POLDER, GOME
• Data recorded for a given solar angle by nadir
  instruments (typic.)
• Reconstructed Earth gt approximation only
• Shadowing effect between plants ? Hot spot ?
• Take into account the Bidir. Reflectance
  Distribution Function (BRDF) of biomes
• Approximated by Lambertian Earth

20

• Broad-band imaging by POLDER
• The Earth 24th June 2001 21hUT as seen from the
  Moon
• Scene reconstructed from POLDER 12th June 1997
  data
• Lambertian mask

RED 670 nm
IR 865 nm
VRE ( SRED SIR) / SRED
21
Comparison POLDER / Earthshine (Arnold et al.
2003)
Good agreement but probable positive bias from
desert -gt spectrum required
22

• Integrated Earth spectrum from GOME data

Biomes map Continental sea ice map Cloud map
(ISCCP)
23
GOME spectra ocean, ice/snow, cloud 22 biomes
(spring time)
toundra, steppe
ocean
forests
desert
cloud, ice/snow
24
Atm. Rayleigh desert corr. Mean VRE 8.7,

Mean apparent albedo(V) 0.29
25

• Vegetation is a small feature (few )
• gt high S/N 100, Rgt35,
• 100-h range exposure
• but O2 and O3 much more obvious
• On Earth O2 and O3 produced by photosynthesis
• gtif O2 O3 H2O then look for vegetation,
  i.e. a spectral feature revealing missing photons
  used in a photosynthesis (maybe not a red edge at
  700nm)
• Signature shape a sharp feature preferred
  required!...
• because a smooth feature of a few unlikely to
  be detected (S/N), and could be easily explained
  as a combination of minerals)

26

• We must take care of possible mineral artifacts a
  sharp edge can be a mineral artifact !

460nm
620nm (Mercury Sulfide)
Clark 1999

• Minerals (some at least) do show seasonal
  variations !

27

• Was the vegetation visible on the Early Earth ?
• Quaternary last climatic extrema
• Holocene optimum 6kyr BP
• 2C /today
• less continental ice
• more vegetation
• (Sahara was green)
• LGM (last glacial maximum) 21kyr BP
• -2C /today
• more continental and sea ice
• less vegetation

28
Mean VRE 21kyr -2 /today Mean VRE 6 kyr 1
/today (mainly thanks to a greener Sahara !) Mean
VRE 0 kyr 8 (To Be Confirmed) Vegetation is a
weak signal (few ) but remains visible during
last climatic extrema
29
Vegetation as a global biomarker Conclusion

• Difficulties
• Probably a small signal (clouds), high S/N
• Minerals artifacts
• Model of atmosphere
• Universality of photosynthesis signature ?
• Look for spectral features combined with O2 and
  H2O
• Look for seasonal variations
• Resolved images of exoplanets will help !
• 100-km interferometric array in space, for 10
  pixels
• across an Earth at 10pc (Labeyrie 1999)