Instruments and Platforms for Measuring Atmospheric CO2 Concentration

Britton StephensNCAR Atmospheric Technology Division

Outline• Measurement technologies

• Instruments

• Precision, potential bias, costs, CDAS implementation

• Platforms

• Costs, CDAS implementation

Accuracy, or more formally reproducibility, is determined by calibration procedures

Measurement expenses are largely a function of the cost of the people who carry them out.

Expected from fossil fuel emissions

Observations

TransCom1 FF Gradients

What do existing flask measurements tell us?

Dry Mole Fraction of CO2

Brazil Boulder 370 molecules of CO2 370 molecules of CO2

779630 molecules of N2 779630 molecules of N2

210000 molecules of O2 210000 molecules of O2

10000 molecules of Ar 10000 molecules of Ar

40000 molecules of H2O 5000 molecules of H2O

Mole Fraction = 355.8 ppm Mole Fraction = 368.2 ppm

Dry Mole Fraction = 370.0 ppm Dry Mole Fraction = 370.0 ppm

For ± 0.1 ppm consistency, need to stabilize to ± 300 ppm H2O (or dry to dewpoint of -25 ºC)

Absolute Measurement Techniques: Manometric and Gravimetric

NOAA/CMDL Manometer:

Reproducibility of 0.06 ppm

(C. Zhao et al., 1997)

Relative Measurement Techniques: Non-dispersive Infrared (NDIR) Spectroscopy

(from www.tsi.com)

• Broadband IR radiation filtered for 4.26 um

• Cooled emitter and detector

• Pulsed emitter or chopper wheel

• 1 or 2 detection cells

Advantages: Robust, precise

Disadvantages: Non-linear; sensitivity to pressure, temperature, and optical conditions; pressure broadening; power consumption

NDIR measurements require calibration gases tied to a common scale

NOAA/CMDL scheme for

propagation of WMO CO2 scale:

At least 4 points needed for 0.1 ppm consistency

Recalibration needed ~ every 3 years due to drift

$2000

LiCor, Inc. CO2 AnalyzerCMDL Flask Analysis System

Typical LiCor Output Signals

(from www.cmdl.noaa.gov/ccgg)

Intra- and Inter-laboratory agreement still not better than 0.2 ppm

CDAS v 1.0 Implementation

Sensor Cost

Short-term Precision

Zero Drift

rate

Accuracy

$7-10,000 0.2 ppm in one second

<1 ppm per hour

Depends on calibration scheme

Setup Cost Operational Cost Precision Bias

$50-$150,000 depending on surface platform (+$10,000 per level on tall towers)

$50-100,000 per year depending on surface platform, $1000 per flight on light aircraft

0.05 ppm (± 1σ)

0.2 ppm (± 1σ)

LiCor Analyzer Specifications

Manual Flask Sampling

CDAS v 1.0 Implementation

Setup Cost Operational Cost Precision Bias Sampling

$5000 $200 per flask pair (returned as one sample value) + $20*annual flask total

0.05 ppm (± 1σ)

0.2 ppm (± 1σ)

Weekly interval1

1Selective sampling based on radon in v 2.0

Automated Flask Sampling

CDAS v 1.0 ImplementationSetup Cost Operational Cost Precision Bias Sampling

$50,000 ($75,000 on buoy)

$50 per flask sample ($150 on buoy) + $20*annual flask total

0.1 ppm (± 1σ)

0.2 ppm (± 1σ)

Specified 2-24 day interval1

(from www.cmdl.noaa.gov/ccgg)

Atmospheric Observing Systems

Boulder, Colorado

NOAA SBIR

Robust, Precise, CO2 Analyzer for Unattended

Field Use

Under Development

• 1 part in 3000 in 1 minute or less

• 30 cc/min or less gas usage

• Unattended field deployment for 6 months or more

• $5,000 or less when manufactured in quantity

CDAS v 1.0 Implementation

Sensor Cost

Short-term Precision

Zero Drift

rate

Accuracy

$5,000 0.1 ppm in one minute

minimal Depends on calibration scheme

Setup Cost Operational Cost Precision Bias

$500,000 one-time development + $15,000 ($20,000 on light aircraft) (+$5,000 per level on tall towers)

$15,000 per year, $500 per flight on light aircraft

0.1 ppm (± 1σ)

0.2 ppm (± 1σ)

NOAA/AOS Analyzer Specifications

Niwot Ridge Pilot Study August 2002

Jeilun Sun, Steve Oncley, Alex Guenther, Dave Schimel, Don Lenschow, Britt Stephens, Russ Monson, and others.

Prototype Inexpensive/Autonomous CO2 System

Research items:

• Stability of CO2 in aluminum LPG cylinders

• Correction for zero drift between calibrations

Goals:

• 1-2 year service schedule

• Total installation ~ $2000

• 0.5 ppm accuracy

RMT, Ltd., Russia

CDAS v 1.0 Implementation

Sensor Cost

Short-term Precision

Zero Drift

rate

Accuracy

$2,000 0.5 ppm in one second

< 25 ppm per hour

Depends on calibration scheme

Setup Cost Operational Cost Precision Bias

$200,000 one-time development + $5,000 (+$2,000 per level on tall towers)

$5,000 per year 0.2 ppm (± 1σ)

0.5 ppm (± 1σ)

NCAR/RMT Analyzer Specifications

Sensors in Development

Remote Sensing Instruments(not implemented in CDAS v 1.0)

• Upward-looking Fourier-transform infrared (FTIR) spectrometers (~$300K+)

• Active and passive satellite CO2 measurements ($200- $300M)

Both ~ 1-3 ppm in column average with potential biases due to land cover type, aerosols, clouds, viewing and sun angle

• Virtual tall towers (~$50K), 0.2-1.0 ppm for weekly-monthly mid-boundary layer values

Ground Station

CDAS ImplementationSetup Cost Location

v 1.0 $5000 Off of coast allowed to simulate clean air sampling

v 2.0

or

v 3.0

Scaled based upon location

Only allowed on land, co-located with existing meteorological station

Point Barrow Alaska

Very Tall Tower

CDAS ImplementationSetup Cost Location

v 1.0 $100 per meter

Anywhere

v 2.0

or

v 3.0

Co-located with existing tower or budgeted at $1000 per meter

WLEF, Wisconsin, USA

Towers over 650 feet AGL in U.S. and proximity

Light Aircraft Profile

CDAS ImplementationSetup Cost Location

v 1.0 $750 per flight

Anywhere

v 2.0

or

v 3.0

Only allowed over or near land, clear air selection allowed

Carr, Colorado, USA

(from www.cmdl.noaa.gov/ccgg)

Volunteer Observing Ship

CDAS ImplementationSetup Cost Location

v 1.0 $20,000 Great circle route between any two

points

v 2.0

or

v 3.0

Only allowed co-located with existing shipping lanes

Freighter Travel, Ltd.

Blue Star Line

Buoy

CDAS ImplementationSetup Cost Location

v 1.0 $25,000 Anywhere

v 2.0 Co-located with existing buoy or budgeted at $100,000

ATLAS Buoy, TAO Array

(from NOAA/PMEL)

• Research Aircraft ($1,500-$20,000 per hour)

• Research Ships (~$20,000 per day)

Additional Platforms(not implemented in CDAS v 1.0)

Conclusions

• There are wide ranges of available instruments and platforms for CO2 concentration measurements

• All have different setup and operation costs, precisions, and potential biases

• Complexity requires a tool for quantitatively assessing creatively designed networks