ACHIEVING CLIMATE CHANGE
ABSOLUTE ACCURACY IN ORBIT
With its unprecedented accuracy,
the Climate Absolute Radiance and
Refractivity Observatory substantially
shortens the time to detect the magnitude
of climate change at the high confidence
level that decision makers need.
T
HE CLARREO VISION FROM THE NATIONAL RESEARCH
COUNCIL DECADAL SURVEY. A critical issue for climate change
observations is that their absolute accuracy is insufficient to confidently
observe decadal climate change signals (NRC 2007; Trenberth et al. 2013;
Trenberth and Fasullo 2010; Ohring et al. 2005; Ohring 2007). Observing
decadal climate change is critical to assessing the accuracy of climate model pro-
jections (Solomon et al. 2007; Masson and Knutti 2011; Stott and Kettleborough
2002) as well as to attributing climate change to various sources (Solomon et al.
2007). Sound policymaking requires high confidence in climate predictions
verified against decadal change observations with rigorously known accuracy.
The need to improve satellite data accuracy has been expressed in
Detail of CLARREO (red orbit track) obtaining matched data to serve as reference intercalibration for instruments
on a polar orbiting weather satellite (green track). For more information see Fig. 6.
BY BRUCE A. WIELICKI, D. F. YOUNG, M. G. MLYNCZ AK, K. J. THOME,
S. L
EROY, J. CORLISS, J. G. ANDERSON, C. O. AO, R. BANTGES, F. BEST,
K. B
OWMAN, H. BRINDLEY, J. J. BUTLER, W. COLLINS, J. A. DYKEMA, D. R. DOELLING, D. R. FELDMAN, N. FOX,
X. H
UANG, R. HOLZ, Y. HUANG, Z. JIN, D. JENNINGS, D. G. JOHNSON, K. JUCKS, S. KATO, D. B. KIRK-DAVIDOFF,
R. K
NUTESON, G. KOPP, D. P. KRATZ, X. LIU, C. LUKASHIN, A. J. MANNUCCI, N. PHOJANAMONGKOLKIJ, P. PILEWSKIE,
V. R
AMASWAMY, H. REVERCOMB, J. RICE, Y. ROBERTS, C. M. ROITHMAYR, F. ROSE, S. SANDFORD, E. L. SHIRLEY,
W. L. S
MITH SR., B. SODEN, P. W. SPETH, W. SUN, P. C. TAYLOR, D. TOBIN, AND X. XIONG
U.S. interagency reports (Ohring et al. 2005; Ohring
2007) and international observing system plans
(GEO 2005; GCOS 2011) and the Global Space-
Based Intercalibration System (GSICS; GSICS 2006;
Goldberg et al. 2011). Common challenges identified
in these documents include uncertain long-term cali-
bration drift, insufficient absolute accuracy, gaps in
observations, and increased uncertainty even for over-
lapped and intercalibrated instruments (GEO 2010).
The Climate Absolute Radiance and Refractiv-
ity Observatory (CLARREO;
http://clarreo.larc.nasa
.gov
) addresses these concerns by providing improved
absolute accuracy in global satellite observations that
can be traced continuously on orbit to international
physical standards such as the Système Internationale
(SI) standards for seconds, kelvins, and watts. Thus,
CLARREO should lead to different observing strate-
gies than have been employed in previous weather
and climate satellites. We will summarize this new
perspective on satellite-based observations, which we
expect will be applicable to climate change observa-
tions in general.
CLARREO aims to provide highly accurate and
SI-traceable decadal change observations sensitive
to the most critical but least understood climate
forcings, responses, and feedbacks. The required
accuracy is determined by the projected decadal
changes and the need to detect anthropogenic forced
changes against the background natural variability.
Because of the focus on longer time scales, CLARREO
measurement requirements are determined not by
instantaneous instrument noise levels, but instead by
the long-term absolute accuracy sufficient to detect
large-scale decadal changes (global, zonal, annual,
and seasonal). The result is the creation of climate
change benchmark measurements defined by three
fundamental characteristics:
• Traceable to fundamental SI standards and robust
to gaps in the measurement record;
• Time/space/angle sampling sufficient to reduce
aliasing bias error in global decadal change ob-
servations to well below predicted decadal climate
change and below natural climate variability; and
• Sufficient information content and accuracy to
determine decadal trends in essential climate
change variables.
The National Research Council (NRC) decadal
survey defined three types of CLARREO bench-
mark measurements. The first is spectrally resolved
infrared radiance (IR) emitted from Earth to space
determined with an accuracy of 0.065 K (k = 2, or 95%
confidence
1
). The infrared spectra are traced to the SI
standard for the kelvin. The second benchmark is the
phase delay rate of the signal from the low-Earth-orbit
AFFILIATIONS: WIELICKI, YOUNG, MLY NC Z AK , CORLISS, DOELLING,
J
OHNSON, KATO, KRATZ, LIU, LUKASHIN, PHOJANAMONGKOLKIJ, ROITHMAYR,
S
ANDFORD, SPETH, AND TAYLOR—NASA Langley Research Center,
Hampton, Virginia; THOME, BUTLER, JENNINGS, AND XIONG—NASA
Goddard Space Flight Center, Greenbelt, Maryland; L
EROY,
A
NDERSON, AND DYKEMA—Harvard University, Cambridge,
Massachusetts; AO, BOWMAN, AND MANNUCCI—Jet Propulsion
Laboratory, California Institute of Technology, Pasadena, California;
BANTGES AND BRINDLEY—Imperial College London, London, United
Kingdom; BEST, HOLZ, KNUTESON, REVERCOMB, SMITH, AND TOBIN—
University of Wisconsin—Madison, Madison, Wisconsin; C
OLLINS
AND FELDMAN—Lawrence Berkeley National Laboratory, Berkeley,
California; Fox—National Physical Laboratory, London, United
Kingdom; X. Huang— University of Michigan, Ann Arbor, Michigan;
Y. Huang—McGill University, Montreal, Quebec, Canada; J
IN,
R
OSE, AND SUN—Science Systems Applications, Hampton, Virginia;
J
UCKS—NASA Headquarters, Washington, D.C.; KIRK-DAVIDOFF —
University of Maryland, Greenbelt, Maryland; K
OPP, PILEWSKIE, AND
R
OBERTS—University of Colorado Boulder, Boulder, Colorado;
R
AMASWAMY—National Oceanic and Atmospheric Administration/
Geophysical Fluid Dynamics Laboratory, Princeton, New Jersey;
R
ICE AND SHIRLEY—National Institute of Standards and Technology,
Gaithersburg, Maryland; S
ODEN—University of Miami, Miami,
Florida
CORRESPONDING AUTHOR: Bruce A. Wielicki, Mail Stop 420,
NASA Langley Research Center, Hampton, VA 23681
E-mail: b.a.wielicki@nasa.gov
The abstract for this article can be found in this issue, following the table
of contents.
DOI:10.1175/BAMS-D-12-00149.1
In final form 13 February 2013
©2013 American Meteorological Society
1
In discussing absolute accuracy, the metrology community uses a coverage factor k (BIPM 2008; Datla et al. 2009) that can be
thought of simply as a more generalized version of a statistical confidence bound analogous to a Gaussian standard deviation
(σ). A value of k = 1 is analogous to a 1σ confidence bound, k = 2 to a 2σ bound. We use k instead of σ to establish a rigorous
tie between the climate and metrology communities. This interdisciplinary link is increasingly important in future climate
change studies (WMO/BIPM 2010). Use of NIST-recommended methods of evaluating and reporting uncertainty is essential
to CLARREO science objectives.
1520
OCTOBER 2013
|
Global Navigation Satellite System radio occultation
system (GNSS-RO, or simply RO) occulted by the
atmosphere, with an accuracy of 0.06% (k = 2) for a
range of altitudes from 5 to 20 km in the atmosphere.
The measurement is traced to the SI standard for
the second. The third benchmark measurement is
spectrally resolved nadir reflectance of solar radia-
tion (RS) from Earth to space determined with an
accuracy of 0.3% (k = 2). The percentage is relative to
the mean spectral reflectance of the Earth of about
0.3. While solar spectral reflectance is a measure-
ment relative to solar spectral irradiance, use of the
solar spectral irradiance observations made by the
Total Solar Irradiance Spectrometer (TSIS) enables
traceability to the SI standards for the watt.
IR, RS, and RO measurements provide informa-
tion on the most critical but least understood climate
forcings, responses, and feedbacks associated with the
vertical distribution of atmospheric temperature and
water vapor (IR/RS/RO), broadband reflected (RS)
and emitted (IR) radiative fluxes, cloud properties
(IR/RS), and surface albedo (RS), temperature (IR),
and emissivity (IR).
CLARREO enables two new approaches to climate
analysis: benchmark spectral fingerprinting and refer-
ence intercalibration. Spectral fingerprinting signals
directly measured by CLARREO allow determina-
tion of climate response and feedbacks (Leroy and
Anderson 2010; Leroy et al. 2008a; Huang et al.
2010a,b; Feldman et al. 2011a,b; Jin et al. 2011; Kato
et al. 2011; Roberts et al. 2011). The second approach
uses CLARREO spectra to calibrate satellite instru-
ments that do not reach decadal change absolute
accuracy requirements. These include current and
future instruments such as the Cross-Track Infrared
Sounder (CrIS), Infrared Atmospheric Sounding
Interferometer (IASI), Clouds and the Earth’s
Radiant Energy System (CERES), Visible Infrared
Imager Radiometer Suite (VIIRS), Landsat, and all
geostationary satellite radiometers. In this approach,
CLARREO is an SI-traceable reference standard
in orbit, providing reference intercalibration for
other instruments to support efforts such as GSICS
(Goldberg et al. 2011). These other instruments can
then more accurately observe decadal climate changes
and can also build long-term data records by bridging
data gaps and reducing dependence on assumptions
of stability and of uninterrupted overlap. Note that
CLARREO does not include passive microwave
observations given the lack of sufficiently accurate
SI standards in this spectral region.
The National Aeronautics and Space Adminis-
tration’s (NASA’s) current budget profile includes
a CLARREO launch no earlier than 2022 for the
baseline mission, but studies of other options are
underway (see the section “Future directions”).
RELATIONSHIP TO MAJOR CHALLENGES
IN CLIMATE SCIENCE AND PREDICTION.
CLARREO decadal change observations are also
needed to reduce uncertainties in the climate feed-
backs that drive uncertainty in climate sensitivity.
These feedbacks (from largest to smallest uncertainty)
are from clouds, lapse rate/water vapor, and snow/ice
albedo (Solomon et al. 2007; Soden and Held 2006;
Bony et al. 2006; Roe and Baker 2007). In addition,
CLARREO will help quantify radiative forcing from
anthropogenic changes in land albedo, will quanti-
tatively confirm the effect of greenhouse gases on
infrared emissions to space, and will make modest
contributions to improving aerosol direct radiative
forcing.
CLARREO employs recent advances in me-
trology for more accurately calibrated solar
and infrared instruments, and uses better radio
occultation to improve capabilities to probe the
atmosphere (see “Mission and instrument design”
sidebar). CLARREO also measures with high
spectral resolution over 95% of the spectrum of
Earth’s thermal emitted radiation (200–2000 cm
–1
or 5–50-μm wavelength) and solar reflected radia-
tion (350–2300 nm) for the first time. This is the
spectrum of energy that radiatively forces climate
change and feedbacks. Because its spectral range
spans many other instruments, CLARREO is a
metrology laboratory in orbit, anchoring the global
satellite monitoring system.
While most satellite missions strive for smaller
spatial scales to improve understanding of Earth
processes (Stephens et al. 2002, Winker et al. 2010),
a climate change metrology mission like CLARREO
must focus on larger scales—for example, the spatial
patterns of critical climate feedbacks (Fig. 1). Climate
models show that these feedbacks occur on spatial
scales of 2,000 km or larger and are often very zonal
in nature.
The Intergovernmental Panel on Climate Change
(IPCC) typically uses a 5-yr running mean filter on
decadal time series (Solomon et al. 2007) to reduce
the impact of the typical (3–5-yr period) natural vari-
ability from El Niño–Southern Oscillation (ENSO)
events. As a result, CLARREO focuses primarily on
observing annual and longer time scales, with an
initial benchmark climate record of at least 5 years.
Chung et al. (2012) confirm that a 5-yr running mean
is a lower bound on the duration needed to accurately
1521
OCTOBER 2013AMERICAN METEOROLOGICAL SOCIETY
|
MISSION AND INSTRUMENT DESIGN
C
LARREO requirements (Table 1) were used to develop
instrument designs, with the additional goal of reducing
instrument size to minimize mass, power, and cost. A wide
range of mission orbits, spacecraft, and launch vehicle
designs were considered to optimize the requirements.
Prototype designs were developed for all of the instruments,
with similar designs being used to verify calibration accuracy
tests in collaboration with NIST.
The CLARREO instruments are much smaller than
typical weather instruments such as VIIRS (252 kg), CrIS
(152 kg), and IASI (210 kg). This allows small spacecraft
and launch vehicles. The entire suite of CLARREO instru-
ments would require a satellite with mass of only one-
third to one-sixth that of the fl agship missions Terra, Aqua,
or NPP.
CLARREO instrument design represents an advance in
absolute calibration over existing instruments. Figure 2a
demonstrates how this is achieved for the thermal infrared
interferometer, including independent deep cavity
blackbodies with multiple phase change cells for tem-
perature accuracy; an infrared quantum cascade laser to
monitor blackbody emissivity as well as spectral response;
multiple deep space views to verify polarization sensitivity;
and a heated halo on the blackbody to independently verify
blackbody emissivity (Anderson et al. 2004; Dykema and
Anderson 2006; Gero et al. 2008, 2012; Best et al. 2008).
Figure 2b demonstrates the approach for the refl ected
solar spectrometer and its use of the moon as a reference
for stability in orbit, the sun with multiple attenuators
to verify instrument nonlinearity of gain across the Earth
viewing dynamic range, and the ability to directly scan
deep space to verify instrument offsets (Espejo et al. 2011;
Fox et al. 2011). Spectral response is verifi ed using solar
spectral absorption line features. One critical difference
FIG. 1. IPCC Fourth Assessment Report (AR4) climate model ensemble means of decadal feedback for tem-
perature, water vapor, surface albedo, and clouds (Soden et al. 2008). Only very large spatial scales of 2000 km
and larger are driving sensitivity of the climate system to anthropogenic forcing, and thus CLARREO’s focus
is large scale.
1522
OCTOBER 2013
|
quantify feedbacks in coupled ocean–atmosphere
models. CLARREO’s long-term focus depends on
thousands of observations and hence on accuracy,
whereas weather and climate process missions de-
pend on instantaneous observations and hence their
precision. Averaging measured spectra over large
time and space domains reduces uncertainty due to
uncorrelated random instrument noise to an insig-
nificant level over time. Thus, on annual and longer
time scales the main uncertainty in the measured
CLARREO radiances is due to systematic uncer-
tainty, not random noise.
This tolerance for moderate random instrument
noise allows CLARREO to use smaller instruments
from other instruments in orbit is that the entire in-
strument can point at the Earth, sun (every 2 weeks),
moon (monthly at 5°–10° phase angle), or deep space.
This eliminates the need for scanning mirrors with
angle-dependent calibration uncertainties, and allows
the use of depolarizers to reduce polarization sensitiv-
ity to the required levels. Scanning the instrument
view across lunar and solar disks provides images
suitable for verifying stray light performance. Finally,
any future improvements in the absolute refl ectance
of the lunar surface can be used to tie the CLARREO
solar spectrometer results to future improvements
in calibration beyond CLARREO, even should these
improvements come 10 or even 30 years from now
(Kieffer 1997; Kieffer and Stone 2005). Note that the
calibration of the refl ected solar is in terms of refl ec-
tance units, which can be converted to absolute radi-
ance using the spectral total solar irradiance provided
by instruments such as TSIS with an expected absolute
accuracy of 0.25% (Richard et al. 2011).
The original CLARREO decadal survey mission
called for three spacecraft at 90° inclination (NRC
2007; Kirk-Davidoff et al. 2005) to assure full 24-h
diurnal sampling accuracy on regional, zonal, and
global averages. The more recent development of
the CLARREO accuracy requirements referenced to
natural variability, combined with additional orbital
sampling studies for IR and RS, demonstrated that the
mission could be reduced to a single 90° orbit, signifi -
cantly reducing mission cost. The 90° orbit is unique to
CLARREO and assures full diurnal cycle sampling for
spectral fi ngerprints as well as full reference intercali-
bration sampling over all climate regimes and all satellite
orbit thermal conditions.
TABLE 1. Instrument and mission requirements. NEDT = noise equivalent differential temperature. FTS =
Fourier transform spectrometer. S/N = signal-to-noise ratio. TRIG = Tri GPS GNSS RO Sensor. RAAN =
right ascension of ascending node.
IR spectrometer RS spectrometer GNSS radio occultation Spacecraft orbit
Systematic error <0.06 K
(k = 2)
Systematic error <0.3% (k = 2) of
Earth mean reflectance
Systematic error <0.06%
refractivity (k = 2) for
5–20 km
90° ± 0.1° orbit for full diurnal
sampling twice per year
200–2000 cm
–1
spectral
coverage
320–2300-nm spectral coverage GPS and Galileo GNSS
frequencies
Global coverage 90°
inclination
0.5 cm
–1
unapodized
spectral resolution
4-nm spectral samples; 8-nm
resolution
5–20-km altitude range
refractivity
609 ± 0.2-km altitude, 61-day
repeat
NEDT < 10 K for
200–600 cm
–1
, and
>1600 cm
–1
, all others <2 K
S/N > 33 for 0.3 scene
reflectance, at a solar zenith angle
of 75°. S/N > 25 for λ > 900 nm
>1000 occultations per day
to control sampling noise
RAAN of 0° or 180°
to optimize reference
intercalibration
25–100-km nadir FOV 0.5-km nadir FOVs for a 100-km-
wide swath
5-yr initial mission record
length
<200 km between
successive spectra along
the ground track
Polarization sensitivity <0.5%
(k = 2) for λ < 1000 nm, <0.75%
(k = 2) for λ > 1000 nm
Orbits repeat exactly each
year to avoid diurnal/seasonal
cycle aliasing
Nadir pointing, with
systematic error <0.2°
Pointable in azimuth and elevation
for solar, lunar, reference
intercalibration views
RS and IR fly on same
spacecraft or in close
formation
Prototype design: 4-port
FTS, 76-kg mass, 124-W
avg power, 2.5 GB day
–1
Prototype design: Dual Grating
Spectrometer, 69-kg total mass,
96-W avg power, 30 GB day
–1
Prototype design: TRIG
receiver, 18-kg mass, 35-W
avg. power, 1.2 GB day
–1
IR/RO- or RS-fueled
spacecraft mass 370 kg, can fit
on small launch vehicles
1523
OCTOBER 2013AMERICAN METEOROLOGICAL SOCIETY
|
![FIG. 2. CLARREO concepts for improved SI-traceable absolute accuracy in orbit. (a) The infrared relies on phase change cells at –39°, 0°, and 30°C to verify thermistor accuracy, quantum cascade laser and heated halos to verify blackbody emissivity, optics design to verify polarization sensitivity, and the quantum cascade laser with integrating sphere to verify instrument spectral response. (b) The verification of nadir spectral reflectance accuracy relies on rotating the entire instrument to view the moon at constant phase angle as a single-level stable reflectance source [similar to the Sea-viewing Wide Field-of-View Sensor (SeaWiFS)], the sun in combination with filters and precision apertures for nonlinearity determination, and the use of depolarizers to control polarization sensitivity.](/figures/fig-2-clarreo-concepts-for-improved-si-traceable-absolute-1zr6jtkc.webp)
