Decadal variability in the
Earth’s reflectance as
observed by Earthshine
Enric Pallé
Big Bear Solar Observatory, NJIT
The albedo sets the input to
the climate heat engine
Pin = Cπ RE2 (1 − A); Pout = 4π RE2σ εT 4 ; T 4 =
Solar constant
Albedo
C
(1 − A);
4σε
A ~ 0.30
The Solar Constant
• We have measured the solar irradiance from space and we also
have tried to ‘reconstruct’ it for past times.
• Solar Irradiance changes can not explain the earth’s temperature
variability without the help of stretching factors (amplification)
Is the mechanism(s)
indirect ?
•EUV
•Cosmic rays
•Others?
The climate is sensitive to A
„
„
„
The average energy input from the sun is
C(1-A)/4 = 240 W/m2
Changing A by 0.01 changes this by 3.4 W/m2
This is climatologically significant
All anthropogenic greenhouse gases over last 150
years result in 2.4 W/m2
‹ Doubling CO2 results in about twice this amount
‹
„
Linearization of the power balance (absent feedbacks)
gives
dT / dA ~ -1.5K / 0.01
The earth’s albedo is highly
variable
Clear Overcast
„
Local albedo depends upon:
‹ Surface
type
‹ Meteorology (clouds)
‹ Solar zenith angle (time of day)
„
The global albedo varies with the
seasons
‹ North/South
land asymmetry
‹ Snow/ice cover
‹ Cloud patterns
Land
0.16
0.50
Ocean
0.08
0.44
Desert
0.23
Snow
0.68
The Earthshine Project: Photometry goals
„
„
„
„
„
The Moon enables us to monitor one aspect of climate change, the
earth’s reflectance
Observe earthshine to determine absolutely calibrated, large-scale,
high-precision measurements of the earth’s reflectance
Look for secular, seasonal and long-term
variations in the albedo (like over a solar
cycle)
Transient phenomena like El Niño or
volcanic eruptions
Simulate the observational results
‹
‹
Compare with observations
Calibrate treatment of cloud cover
Earthshine measurements of the Earth’s
large-scale reflectance
Waning / morning
„
„
„
„
The Earthshine is the ghostly
glow on the dark side of the
Moon
Origin of Earthshine first
explained by Leonardo da Vinci
First measured by Danjon
beginning in 1927-34 and by
Dubois 1940-60.
ES/MS = albedo (+ geometry
and moon properties)
a
b
⎛ I a Ta ⎞⎛ pb f b (θ ) ⎞⎛ Rem ⎞
⎟⎟⎜⎜
⎟⎟⎜⎜
⎟⎟
pe f e (β ) = ⎜⎜
⎝ I b Tb ⎠⎝ pa f a (θ 0 ) ⎠⎝ Re ⎠
θ=π−β
Zero-airmass
intensities
Lunar
reflectivities
θ0 ~ 1o
Rem
⎛ Res ⎞
⎜⎜
⎟⎟
⎝ Rms ⎠
2
Geometry
π
σ
A = 2 = ∫ pe f e (β ) sin β dβ
πRe −π
β
3 pe f e ( β )
; A* ≡ −2.5 log A*
A ≡
2 f L (β )
*
Re
dσ
≡ pe f e (β )Re2
dΩ
f (0) ≡ 1
2
fL
[ ]
(
π − β )cos β + sin β
(β ) =
π
A* → p* for historical reasons
The Effective and Bond Albedos
„
„
On any one night, we measure p*, the effective (or apparent) albedo
(1 direction). (different Sun-Earth-Moon reflection angle)
To obtain the Bond albedo, A, we integrate over all phases of the
moon at monthly/yearly time scales
2
A = ∫ p * f L (θ ) sin(θ )dθ
3
Coverage during one night
15/10/99
Phase = -116
Evening
In the sunlight &
Visible from the Moon
04/09/99
Phase = +110
Morning
Morning obs /Waning Moon
Evening obs /Waxing Moon
Modeling
hourly
variations
Cloudy Asia
North America
Dark Arabian Sea
Dark Atlantic
„
Temporal offset
due to cloud
cover posting
interval
(12 hours) ??
Changes in the Earth’s albedo over the last
20 years
„
„
„
„
„
„
Earthshine Observations: December 1998 – present
ISCCP data June 1983 – September 2001 (to be
updated)
International Satellite Cloud Climatology Project
(ISCCP) provides ~100 daily cloud variables on a
(280 km)2 grid
For each observation, calculate double-projected (E-S
and E-M) area average of these variables
Regress observed A* anomaly against the most
significant of these
This allows us to reconstruct the earth’s albedo as
seen from BBSO since 1983
Multiple regression on p*
Common period December 1998 – September 2001
Regress On:
9
9
9
cloud cover
optical thickness
surface reflectance
Earth’s albedo 1983-2003
Albedo
Effective
Ground-level
solar
Irradiance
And we are not alone: tropical satellite data OLR and SW
The proxy implications
„
Confidence in our results based on:
94-95 earthshine data agreement
‹ Positive/negative phases are similar
‹ Scrambling the data in mock reconstructions time/space
support the trend
‹
„
Variation is large
‹
Albedo change is 6 W/m2 ; GHG up to now is 2.4 W/m2
‹
Equivalent to 2% increase in solar irradiance, a
factor
20!! the typical maxima to minima variations (0.3 W/m2)
Reversibility suggests natural variations.
‹ GCM do not show such variations
‹ What is the climatic impact? Recent warming
acceleration?
‹
Not so surprising…
Although A does not only
depend on mean cloud
amount….
….ISCCP data show
reduction in cloud amount
1983-2001
Source: ISCCP web site
The ES results are not inconsistent
with satellite observations
Radiation anomalies
within ± 20o of the
Equator
B. Wielicki et al.,
Science p. 841, v. 295
(2002)
Coverage of multiple stations
•2nd Station now operating in Crimea; 3rd ready to go to Yunnan
•Canary Islands optimal complement to BBSO
•Prototype robotic telescope under construction for 8-station network
Coverage with simultaneous
observations
Four station simulation
Earthshine Spectroscopy
Earth’s spectral albedo for 2004/11/19
Montañés Rodriguez et al. (in prep.)
Rayleigh Scattering
Chappuis Ozone band
B-O2 A-O2
Atmospheric
Water vapor
Summary
„
„
„
„
„
„
„
ES is a viable way to monitor the climate system on large scales and
over long times
By combining ES and ISCCP data, we have a 20-year record of the
earth’s SW reflectance that
‹ Shows surprising interannual coherence and a large decadal
variability that is likely natural (why??)
‹ Is not reproduced by current models
Continued observations with a global network are warranted
Visible and near-IR spectroscopic observations of ES are being
analysed
Are these changes due to GHGs? …. No
Are these changes solar? …. humm!!
Are they natural variability? ….. probably
The End
Some references
„
„
„
„
„
Pallé E, Goode P.R., Montañes-Rodriguez P., Koonin S.E., Changes in
Earth’s reflectance over the past two decades, Science, May 28th issue
(2004).
Qiu J., Goode PR, Pallé E, Yurchyshyn V, Hickey J, MontañesRodriguez P., Chug MC, Kolbe E, Brown CT, Koonin SE, Earthshine
and the Earth’s albedo 1.: Precise and large-scale nightly
measurements, Journal Geophysical Research, 108(D22), 4709,
doi: 10.1029/2003JD003610 (2003).
Pallé E, Goode PR, Qiu J, Yurchyshyn V, Hickey J, MontañesRodriguez P., Chu MC, Kolbe E, Brown CT, Koonin SE, Earthshine and
the Earth’s albedo 2.: Observations and simulations over 3 years,
Journal Geophysical Research, 108(D22), 4710,
doi: 10.1029/2003JD003611 (2003).
Goode PR, Qiu J, Yurchyshyn V, Hickey J, Chu MC, Kolbe E, Brown
CT, Koonin SE, Geophysical Research Letters 28, 1671 (2001).
MacDonald, GJ and Koonin, SE, Earthshine and Climate, The
Observatory, 112, 59 (1992).
http://www.bbso.njit.edu/~epb/
The global Earth is not too
green
60% of Earth is covered
by clouds at any time
For 2003/11/19:
10:30
UT
13:00 UT
Oceans
79%
71%
Snow/ice
8%
11%
Land
13%
18%
Vegetation
9%
14%
Unobstructed
by clouds
5%
8%
Regression against ISCCP
14 10-degree lunar phase bins
±(65-135)
310 nights 1999-mid 2001
r = 0.6; P << 0.001
Coefficients vary smoothly
with lunar phase
Cloud/Radiation changes this century
ISCCP
Solar Surface radiation
(global dimming)
Sunshine records
Synoptic cloud observations
Earthshine
1900
1950
2000
Comparison with photometric
albedos and models for one night
Montañés Rodriguez et al. (in preparation)