Functional Block Diagram

1
The Orbiting Carbon Observatory (OCO)
Mission Vijay Natraj Ge152 Wednesday, 1 March
2006
2
Atmospheric CO2 the Primary Anthropogenic
Driver of Climate Change
Keeling Plot
Since 1860, global mean surface temperature has
risen 1.0 C with a very abrupt increase since
1980.
Atmospheric levels of CO2 have risen from 270
ppm in 1860 to 370 ppm today. Accumulation of
atmospheric CO2 has fluctuated from 1 6 GtC/yr
despite nearly constant anthropogenic emissions.
WHY?
3
An Uncertain FutureWhere are the Missing Carbon
Sinks?

• Only half of CO2 produced by human activities
  over the past 30 years has remained in the
  atmosphere.
• What are the relative roles of the oceans and
  land ecosystems in absorbing CO2?
• Is there a northern hemisphere land sink?
• What are the relative roles of North America/
  Eurasia?
• What controls carbon sinks?
• Why does the atmospheric buildup vary with
  uniform emission rates?
• How will the sinks respond to climate change?
• Climate prediction requires an improved
  understanding of natural CO2 sinks.
• Future atmospheric CO2 increases
• Their contributions to global change

4
The Global Carbon Cycle Many Questions

• Atmospheric CO2 has been monitored systematically
  from a network of 100 surface stations since
  1957.

The 100 GLOBALVIEW-CO2 flask network stations
and the 26 continental sized zones used for CO2
flux inversions. This network is designed to
measure back-ground CO2. It cannot retrieve
accurate source and sink locations or
magnitudes! Bousquet et al., Science 290, 1342
(2000).
5
Why Measure CO2 from Space?Improved CO2 Flux
Inversion Capabilities

• Studies using data from the 56 GV-CO2 stations
• Flux residuals exceed 1 GtC/yr in some zones
• Network is too sparse
• Inversion tests
• Global XCO2 pseudo-data with 1 ppm accuracy
• Flux errors reduced to lt0.5 GtC/yr/zone for all
  zones
• Global flux error reduced by a factor of 3.

OCO
Flux Retrieval Errors GtC/year/Zone
Fig. F.1.3 Carbon flux errors from simulations
including data from (A) the existing surface
flask network, and (B) satellite measurements of
XCO2 with accuracies of 1 ppm on regional scales
on monthly time scales
Rayner OBrien, Geophys. Res. Lett. 28, 175
(2001)
6
Why Measure CO2 from Space? Dramatically
Improved Spatiotemporal Coverage
The OCO orbit pattern (16-day repeat cycle)
7
The Orbiting Carbon Observatory (OCO) Mission

• Make the first, global, space-based observations
  of the column integrated dry air mole fraction,
  XCO2, with 1 ppm precision.
• Combine satellite data with ground-based
  measurements to characterize CO2 sources and
  sinks on regional scales on monthly to
  interannual time scales
• Fly in formation with the A-Train to facilitate
  coordinated observations and validation plans

8
XCO2 Retrieved from Bore-Sited CO2 and O2 Spectra
Taken Simultaneously

• High resolution spectroscopic measurements of
  reflected sunlight in near IR CO2 and O2 bands
  provide the data needed to retrieve XCO2
• Column-integrated CO2 abundance
• Maximum contribution from surface
• Other data needed (provided by OCO)
• Surface pressure, albedo, atmospheric
  temperature, water vapor, clouds, aerosols
• Why high spectral resolution?
• Lines must be resolved from the continuum to
  minimize systematic errors

Clouds/Aerosols, Surface Pressure
Clouds/Aerosols, H2O, Temperature
Column CO2
9
Spatial Sampling Strategy

• OCO is designed provide an accurate description
  of XCO2 on regional scales
• Atmospheric motions mix CO2 over large areas as
  it is distributed through the column
• Source/Sink model resolution limited to 1o x 1o
• High spatial resolution
• 1 km x 1.5 km footprints
• Isolates cloud-free scenes
• Provides thousands of samples on regional scales
• 16-day repeat cycle
• Provides large numbers of samples on monthly time
  scales

Spatial sampling along ground track
Ground tracks over the tip of South America
10
Operational Strategy Maximizes Information
Content and Measurement Validation Opportunities

• 115 PM near polar (98.2o) orbit
• 15 minutes ahead of EOS A-Train
• Same ground track as AQUA
• Global coverage every 16 days
• Science data taken on day side
• Nadir mode
• Highest spatial resolution
• Glint mode
• Highest SNR over ocean
• Target mode
• Validation
• Airmass dependence
• Comparison with surface FTS stations
• Calibration data taken on night side
• Solar, limb, dark, lamp

11
Sampling Biases
Q20

• 115 PM local sampling time chosen because
• Production of CO2 by respiration is offset by
  photosynthetic uptake
• Instantaneous XCO2 measurement is within ?0.3 ppm
  of the diurnal average (see figure)
• Atmospheric transport desensitizes OCO
  measurements to the clear-sky bias
• Air passes through clouds on a time-scale short
  compared to the time needed to affect significant
  changes in XCO2
• Mixing greatly reduces the influence of local
  events point sources on XCO2

MAY
Fig. F.2.4 a) Calculated monthly mean, 24 hour
average XCO2 (ppm) during May using the NCAR
Match model driven by biosphere and fossil fuel
sources of CO2. b) XCO2 differences (ppm)
between the monthly mean, 24 hour average and
the 115 PM value
12
Will it Work?

• Accuracies of 1 ppm needed to identify CO2
  sources and sinks
• Realistic, end-to-end, Observational System
  Simulation Experiments
• Reflected radiances for a range of
  atmospheric/surface conditions
• line-by-line multiple scattering models
• Comprehensive description of
• mission scenario
• instrument characteristics
• Results
• Retrieve XCO2 from single clear sky nadir
  sounding to 0.3-2.5 ppm precision
• Rigorous constraints on the distribution and
  optical properties of clouds and aerosols

End-to-end retrievals of XCO2 from individual
simulated nadir soundings at SZAs of 35o and 75o.
The model atmospheres include sub-visual cirrus
clouds (0.02??c? 0.05), light to moderate aerosol
loadings (0.05??a? 0.15), over ocean and land
surfaces. INSET Distribution of XCO2 errors
(ppm) for each case
13
Cloud, Aerosol and Cirrus Interference
Clouds, aerosols and sub-visible cirrus (high
altitude ice clouds) prevent measurement of the
entire atmospheric column. Sub-visible cirrus
clouds are effective at scattering near infrared
light because the light wavelengths and particle
sizes are both 1 2 µm. An analysis of
available global data suggests that a space-based
instrument will see cloud-free scenes only
10 of the time. Geographically persistent
cloud cover will be especially problematic and
will induce biases in the data.
Number of cloud-free scenes per month anticipated
for space-based sampling averaged into 3??6?
(Lat?Lon) bins based on AVHRR cloud data
(OBrien, 2001).
14
OCO Performance Improves with Spatial Averaging
Accuracy of OCO XCO2 retrievals as a function of
the number of soundings for optimal (red) and
degraded performance (blue) for a typical case
(37.5? solar zenith angle, albedo0.05, and
moderate aerosol optical depth, ?a0.76 ?m
0.15). Results from end-to-end sensitivity
tests (solid lines) are shown with shaded
envelopes indicating the range expected for
statistics driven by SNR (N1/2) and small-scale
biases (N1/4).
15
Validation Program Ensures Accuracy and Minimizes
Spatially Coherent Biases

• Ground-based in-situ measurements
• NOAA CMDL Flask Network Tower Data
• TAO/Taurus Buoy Array
• Uplooking FTS measurements of XCO2
• 3 funded by OCO
• 4 upgraded NDSC
• Aircraft measurements of CO2 profile
• Complemented by Laboratory and on-orbit
  calibration

Buoy Network
CMDL
16
The Pushbroom Spectrometer Concept
It is possible to obtain many ground-track
spectra simultaneously if the instantaneous field
of view (IFOV) is imaged onto a 2D detector
array. In this case, wavelength information is
dispersed across one dimension and cross-track
scenes are dispersed along the other
dimension. The instrument acquires spectra
continuously along the ground track at a rate of
4.5 Hz. This results in 70 spectra/sec and 9000
spectra per 4??5? region every 16 days.
17
OCO Data Product Pipeline
AIRS T, P, H2O

• The OCO data flow from space through an automated
  pipeline which yields Level 1 and 2 data
  products.
• Level 3 and Level 4 products are produced by
  individual Science Team members.
• Preliminary tests of the retrieval algorithm
  demonstrate the OCO mission concept
• (Kuang et al., Geophys. Res. Lett., 29 (15)
  2001GL014298, 2002).

Space-borne Data Acquisition
Level 2
Calibration Validation Data
Spectral Radiances
Level 3
Ancillary Data FTIR XCO2 GVCO2 CO2 MODIS
Aerosol NCEP Fields
Global 1 ppm XCO2 Maps
Data Assimilation Models
Inversion Models
Level 4
Temporally Varying CO2 Source/Sink Maps
18
Retrieval Algorithm
19
Summary

• Climate Forcing/Response
• T/H2O/O3 AIRS/TES/MLS
• Clouds CloudSat
• Aerosols CALIPSO
• CO2 OCO

• OCO will provide critical data for
• Understanding the carbon cycle
• Essential for developing carbon management
  strategies
• Predicting weather and climate
• Understanding sources/sinks essential for
  predicting CO2 buildup
• O2 A Band will provide global surface pressure
  measurements
• OCO validates technologies critically needed for
  future operational CO2 monitoring missions
• Satisfies a measurement need that has been
  identified by NPOESS, for example

XCO2 (ppm)