1
Jia Hu 1 , Chun-Ta Lai 2 , Britton Stephens 3 , James Ehleringer 2 , Russ Monson 1 and Dave Schimel 3 1 Ecology and Evolutionary Biology, University of Colorado, 2 Department of Biology, University of Utah, 3 National Center for Atmospheric Research Boundary-layer measurements of CO 2 concentration, carbon and oxygen isotopes of atmospheric CO 2 over montane forest regions in Colorado, USA Airborne Carbon in the Mountains Experiment (ACME) Introduction Using satellite data and ecosystem models, Schimel et al. (2002) showed a disproportionate carbon sink in the western montane regions to the continental United States. The mountain ecosystems, particularly areas along the Rocky Mountains, have been suggested as “islands” of high carbon uptake that are responsible for 25 to 50 % of US carbon uptake (Schimel et al., 2002). This high carbon uptake is surprising given a landscape of otherwise arid and semiarid ecosystems with relatively low carbon sequestration under current land-use practice. To understand mechanisms that control carbon exchange, potentially through carbon-water interactions and energy redistribution in forested mountain ecosystems, the Airborne Carbon in the Mountains Experiment (ACME) was proposed under multi-institutional collaborations with a focused area along the Front Range of the Colorado Rocky Mountains. Recognizing that direct measurements of carbon budget using atmospheric techniques is impossible in complex topography, ACME relies on model-data integration with a combination of measurements made at various scales, including airborne and ground-based CO 2 concentration, 13 C and 18 O of CO 2 measurements to calibrate and constrain models. Quantifying respiration in montane regions is particularly difficult owing to the complexity in the topographic effect on advection and drainage flow. ACME emphasizes the importance of characterizing respiration and its isotopic signatures. A “Keeling Plot” approach can be applied to estimate the carbon isotope ratios of nighttime respired CO 2 13 C R ). Flasks were filled in the early morning to measure concentrations, 13 C and 18 O of CO 2 in the boundary layer using a C-130 aircraft. Flasks were also filled inside forest canopies by an automated sampler on the night before a research flight. Using flask samples collected in both ground-based and airborne measurements, we were able to compare δ 13 C R values associated with surface respiration and those in the residual boundary layer. In addition, vertical profiles of CO 2 and its isotope signatures in the convective boundary layer (CBL) were measured, which were compared to tower-based measurements conducted hourly during the day. We show here preliminary results of flask measurements from two ACME campaigns conducted in May and July, 2004. Figure 1. Composite of the NSF/NCAR C-130 research aircraft flying over the Rocky Mountains in Boulder. (Photo by Carlye Calvin) Figure 2. An eddy covariance flux tower in a subalpine forest at ~ 3050 m above sea level, at the Niwot Ridge AmeriFlux site near Boulder, Colorado (Photo by Steve Oncley). Figure 3. A view from the cabin of C-130 during a research flight. Flask sample collection Flask samples were collected both on the ground and in the atmospheric boundary layer. Two sampling strategies were employed for airborne measurements: first is to fill flasks in the early morning to capture respiration signatures before the nocturnal boundary layer was broken from convective mixing. As the ground heated after sunrise and the depth of CBL increased, flasks were collected again to estimate isotope ratios of net CO 2 exchange between the atmosphere and the biosphere. Airborne flask measurements. Flasks were filled on the aircraft using the NCAR MEDUSA, originally designed for O 2 /N 2 sample collections. The MEDUSA sampling unit was customized for the ACME to fill 16 100-mL flasks per flight. Sampling involves flushing a flask for > 20 s (median time = 2.4 min), which is 11 times of the flask volume. Whole air samples were drawn from an inlet located upstream of engine exhaust, pre- dried by flowing through a magnesium perchlorate trap (120 mL) at a constant flow rate of 3.3 SLPM, then pressurized to 1 atm before being collected in 100-ml glass flasks (Kontes Glass Co., Vineland, NJ). A relief valve was set at 1.7 atm to prevent over-pressurizing glass flasks. Ground-based flask measurements. Two programmable flask samplers controlled by CR23X dataloggers (Campbell Scientific, Logan, UT) were deployed to fill flasks within and above forest canopies (Schauer et al. 2003). The first sampled at 15-min intervals the night before a research flight (9:00 pm ~ 12:30 am). These measurements were conducted within forest canopies and were used to characterize 13 C R values at the ecosystem scale. The second was programmed to sample every 45 minutes and started approximately two hours before the airborne flights (4:30 am ~ 3:00 pm). These measurements overlapped aircraft observations, providing information about daytime CO 2 exchange in the atmosphere-forest interface. Figure 5. One of the two automated flask samplers deployed in Niwot Ridge AmeriFlux site. Figure 4. “MESUSA” sampling unit customized for filling 100-ml flasks on C-130. Figure 6. A GC-IRMS system for high precision analyses of concentration, 13 C and 18 O of atmospheric CO 2 at SIRFER laboratory, University of Utah. Isotope analyses Flasks were analyzed at the Stable Isotope Ratio Facility for Environmental Research (SIRFER), University of Utah using an integrated GC-IRMS system. Four replicate of 1- mL aliquots were sub-sampled from a 100-mL flask to analyze CO 2 concentrations, 13 C and 18 O of atmospheric CO 2 . An outlier test (± 1 S.D.) was subsequently performed offline to reduce analytical noise. Measurement precision was determined as 0.4 ppm for CO 2 concentration, 0.06 ‰ for 13 C and 0.09 ‰ for 18 O (Schauer et al. RCMS, in press). Figure 7. Vertical profiles of CO 2 concentration, 13 C and 18 O of CO 2 measured in the residual boundary layer (RBL) on the morning of May 20 and July 22, 2004. Sampling height is referenced to the mean sea level (MSL). Profiles of 13 C closely mirrored those of CO 2 concentration, while profiles of 18 O resembled those of 13 C. For the early morning profile on both sampling days, the impact of respiration was evident in the NBL, where higher CO 2 , lower 13 C and lower 18 O values were found below 3500 m. More negative 13 C values inside NBL were associated with CO 2 released from decomposition of 13 C-depleted plant material and microbial respiration. The smaller values of 18 O in CO 2 within NBL reflected the fact that soil CO 2 efflux carried the 18 O signature of soil water, which was relatively depleted in 18 O compared to that of leaf water. The strong 18 O gradient close to the ground suggested that soil respiration has a profound impact on ecosystem respiration in mountain ecosystems. A transition layer where air entrained from RBL into NBL can be clearly identified in the early morning owing to the capping of thermal inversion. The vertical gradient observed on May 20 was much smaller compared to that measured at 6:00 am on July 22, which can be explained by the lower respiration rate in spring associated with lower temperature. Another explanation is that a deeper NBL on May 20 might dilute respired CO 2 with a greater air volume. On July 22, a second profile measured at 8:00 am indicated a more well mixed boundary layer, suggesting a rapid mixing from thermal convection. Results Vertical profiles Figure 8. Comparison between vertical profiles of CO 2 concentration, 13 C and 18 O of atmospheric CO 2 measured in the morning and in the afternoon of July 20, 2004. Lower altitude was achieved during morning flights in order to collect air samples in the NBL. While respiration clearly showed its impact on NBL during morning hours, afternoon profiles showed a reversed trend with lower CO 2 concentration and higher 13 C values closer to the ground. This reversed vertical profile was due to CO 2 uptake via photosynthesis where plants were known to discriminate against 13 C, leaving the atmosphere enriched in 13 CO 2 . Figure 9. Comparison between 3 vertical profiles of CO 2 concentration, 13 C and 18 O of atmospheric CO 2 measured on the morning of July 29, 2004. All 3 profiles showed great similarities for the vertical structure of atmospheric boundary layer except a more negative 18 O value at 2000 m at 8:00. Figures 7-9 indicated that a transition layer where air aloft entrained into surface boundary layer commonly occurred between 2000 - 3000 m MSL. The top of the nocturnal boundary layer extended as high as 1000 m above the ground during these flights. Figure 10. Comparison between diurnal variations of CO 2 concentration, 13 C and 18 O of atmospheric CO 2 measured above forest canopies and in the CBL. Two days were selected representing spring and summer conditions. First notice the rapid decrease in CO 2 concentration above forests in the early morning, owing to the break up of nocturnal boundary layer. Above-canopy CO 2 concentration remains roughly constant in the afternoon, a well-mixed condition achieved by strong convective turbulence. Midday values of CO 2 concentration were about 7 ppm lower on July 29 than May 20. This difference in the CBL CO 2 concentration reflects the higher photosynthetic uptake in the northern hemisphere in the summer. Values of 13 C largely mirror those of CO 2 concentration. There were times that measurements in CBL were nearly identical to those above the canopy (e.g., early morning on May 20 and in the afternoon of July 29). However, there were also times when CBL measurements were clearly different from those measured above forests. For instance, CO 2 concentration was higher while 13 CO 2 was lower in the CBL from 7 to 10 am on July 29. This may be due to effects of remnant CO 2 from the previous night forming a residual boundary layer on the aircraft measurements. Such influences were dismissed after the convective boundary layer grew in the afternoon from surface heating, giving closer comparisons to canopy values. 18 O values in CO 2 measured in CBL were very comparable to those measured above forests, which clearly showed an enrichment effects of photosynthesis on the morning of July 29. Diurnal variations of CO 2 , 13 C and 18 O Keeling plot R C CO Slope C 13 2 13 Keeling (1958) GCA 13:322 Figure 11. Using a linear regression between the inverse of CO 2 concentration and carbon isotope ratio, one can estimate the carbon isotope ratio of ecosystem respiration, denoted here as 13 C R . This is known as the Keeling plot approach. Data shown here are collected in a C 3 -C 4 tallgrass prairie, which is not a part of the ACME study. It is shown to demonstrate that flux-weighted 13 C R values should fall between boundaries set by C 3 and C 4 photosynthesis had both sources were mixed into the sampling footprint. This is more likely to happen for samples collected in the CBL which integrates sources over a large spatial area. Figure 12. Sunrise at the Jeffco airport, Broomfield Colorado, shortly before an early morning flight. Air samples were collected in the early morning to capture respiration signals before the nocturnal boundary layer dissipates. Carbon isotope ratio of respiration Figure 13. Comparisons between carbon isotope ratios of respiration measured in CBL ( 13 C CBL ) and within forest canopies ( 13 C R ). For the 3 selected days, values of 13 C CBL were slightly more positive, considerably more positive and indistinguishable from those of 13 C R on May 20, May 28 and July 22 respectively. More positive values in CBL could be interpreted as that samples collected in CBL integrate fluxes over a larger spatial area, therefore they are likely to be affected by non-local sources of C 4 plants afar from the mountain regions. Here sampling height is referenced to the mean sea level (MSL). Table 1. Comparisons between 13 C CBL and 13 C R values for aircraft data when ground-based measurements were also available. All the 13 C CBL values were consistently greater than or indistinguishable from 13 C R values in both May and July periods. Our interpretation is that aircraft data were subject to influences of C 4 photosynthesis several kilometers away from the sampling area, whereas 13 C R measurements only represent respiration from coniferous forests below. Values of 13 C R were relatively constant and more robust compared to 13 C CBL values, which is partially due to greater CO 2 ranges obtained near the ground. Discussion Measurements of CO 2 concentrations, 13 C and 18 O of atmospheric CO 2 were conducted by filling flasks in the atmospheric boundary layer and on the ground in a subalpine forest in the Colorado mountain regions. These measurements are an integral part of an extensive program (ACME) to investigate carbon sequestration in mountain landscapes, potentially responsible for a significant fraction of carbon uptake in the U.S. The data shown here will provide an independent check to ensure the quality of online CO 2 concentration measurements. Simultaneous measurements in the boundary layer and above forest canopies also provide a unique opportunity to characterize boundary layer dynamics, including venting of surface respired CO 2 from the nocturnal boundary layer and the mixing of convective turbulence. While the data are preliminary, they promise to provide a great utility for model calibration and validation. Quantifying carbon sequestration in the mountains will have to rely on models to interpret observations, whereas model results are only useful if they can be verified by actual measurements. The isotope data will provide additional information to constrain modeled fluxes from inverse calculations based on CO 2 concentration measurements. Isotopic discrimination by terrestrial biomes is an important parameter in global inverse models. Lai et al. (2004, GBC:18, GB1041; 2004 GCB, in press) proposed a useful experimental approach to estimate carbon isotope discrimination during photosynthesis at the canopy scale. However, current estimates of the discrimination factor are sensitive to uncertainties in the background atmosphere. Flask measurements conducted here will allow for further examinations on this issue. Acknowledgements: This research is supported, in part, by the Terrestrial Carbon Processes (TCP) program by the office of Science (BER), U.S. Department of Energy under Grant No. DE-FG03-00ER63012, the NSF Biocomplexity program, and NCAR Biogeosciences initiative.

Jia Hu 1, Chun-Ta Lai 2, Britton Stephens 3, James Ehleringer 2, Russ Monson 1 and Dave Schimel 3 1 Ecology and Evolutionary Biology, University of Colorado,

Embed Size (px)

Citation preview

Page 1: Jia Hu 1, Chun-Ta Lai 2, Britton Stephens 3, James Ehleringer 2, Russ Monson 1 and Dave Schimel 3 1 Ecology and Evolutionary Biology, University of Colorado,

Jia Hu1, Chun-Ta Lai2, Britton Stephens3, James Ehleringer2, Russ Monson1 and Dave Schimel3

1Ecology and Evolutionary Biology, University of Colorado, 2Department of Biology, University of Utah, 3National Center for Atmospheric Research

Boundary-layer measurements of CO2 concentration, carbon and oxygen isotopes of atmospheric CO2 over

montane forest regions in Colorado, USA

Airborne Carbon in the Mountains Experiment (ACME)

Introduction

Using satellite data and ecosystem models, Schimel et al. (2002) showed a

disproportionate carbon sink in the western montane regions to the continental United

States. The mountain ecosystems, particularly areas along the Rocky Mountains, have

been suggested as “islands” of high carbon uptake that are responsible for 25 to 50 %

of US carbon uptake (Schimel et al., 2002). This high carbon uptake is surprising

given a landscape of otherwise arid and semiarid ecosystems with relatively low

carbon sequestration under current land-use practice. To understand mechanisms that

control carbon exchange, potentially through carbon-water interactions and energy

redistribution in forested mountain ecosystems, the Airborne Carbon in the Mountains

Experiment (ACME) was proposed under multi-institutional collaborations with a

focused area along the Front Range of the Colorado Rocky Mountains. Recognizing

that direct measurements of carbon budget using atmospheric techniques is impossible

in complex topography, ACME relies on model-data integration with a combination of

measurements made at various scales, including airborne and ground-based CO2

concentration, 13C and 18O of CO2 measurements to calibrate and constrain models.

Quantifying respiration in montane regions is particularly difficult owing to the

complexity in the topographic effect on advection and drainage flow. ACME

emphasizes the importance of characterizing respiration and its isotopic signatures. A

“Keeling Plot” approach can be applied to estimate the carbon isotope ratios of

nighttime respired CO2 (δ13CR). Flasks were filled in the early morning to measure

concentrations, 13C and 18O of CO2 in the boundary layer using a C-130 aircraft.

Flasks were also filled inside forest canopies by an automated sampler on the night

before a research flight. Using flask samples collected in both ground-based and

airborne measurements, we were able to compare δ13CR values associated with surface

respiration and those in the residual boundary layer. In addition, vertical profiles of

CO2 and its isotope signatures in the convective boundary layer (CBL) were measured,

which were compared to tower-based measurements conducted hourly during the day.

We show here preliminary results of flask measurements from two ACME campaigns

conducted in May and July, 2004.

Figure 1. Composite of the NSF/NCAR C-130 research aircraft flying over the Rocky Mountains in Boulder. (Photo by Carlye Calvin)

Figure 2. An eddy covariance flux tower in a subalpine forest at ~ 3050 m above sea level, at the Niwot Ridge AmeriFlux site near Boulder, Colorado (Photo by Steve Oncley).

Figure 3. A view from the cabin of C-130 during a research flight.

Flask sample collection  

Flask samples were collected both on the ground and in the atmospheric boundary layer.

Two sampling strategies were employed for airborne measurements: first is to fill flasks in

the early morning to capture respiration signatures before the nocturnal boundary layer was

broken from convective mixing. As the ground heated after sunrise and the depth of CBL

increased, flasks were collected again to estimate isotope ratios of net CO2 exchange

between the atmosphere and the biosphere.

Airborne flask measurements. Flasks were filled on the aircraft using the NCAR MEDUSA,

originally designed for O2/N2 sample collections. The MEDUSA sampling unit was

customized for the ACME to fill 16 100-mL flasks per flight. Sampling involves flushing a

flask for > 20 s (median time = 2.4 min), which is 11 times of the flask volume. Whole air

samples were drawn from an inlet located upstream of engine exhaust, pre-dried by flowing

through a magnesium perchlorate trap (120 mL) at a constant flow rate of 3.3 SLPM, then

pressurized to 1 atm before being collected in 100-ml glass flasks (Kontes Glass Co.,

Vineland, NJ). A relief valve was set at 1.7 atm to prevent over-pressurizing glass flasks.

Ground-based flask measurements. Two programmable flask samplers controlled by CR23X

dataloggers (Campbell Scientific, Logan, UT) were deployed to fill flasks within and above

forest canopies (Schauer et al. 2003). The first sampled at 15-min intervals the night before a

research flight (9:00 pm ~ 12:30 am). These measurements were conducted within forest

canopies and were used to characterize 13CR values at the ecosystem scale. The second was

programmed to sample every 45 minutes and started approximately two hours before the

airborne flights (4:30 am ~ 3:00 pm). These measurements overlapped aircraft observations,

providing information about daytime CO2 exchange in the atmosphere-forest interface. Figure 5. One of the two automated flask samplers deployed in Niwot Ridge AmeriFlux site.

Figure 4. “MESUSA” sampling unit customized for filling 100-ml flasks on C-130.

Figure 6. A GC-IRMS system for high precision analyses of concentration, 13C and 18O of atmospheric CO2 at SIRFER laboratory, University of Utah.

Isotope analyses

Flasks were analyzed at the Stable Isotope Ratio Facility for Environmental Research

(SIRFER), University of Utah using an integrated GC-IRMS system. Four replicate

of 1-mL aliquots were sub-sampled from a 100-mL flask to analyze CO2

concentrations, 13C and 18O of atmospheric CO2. An outlier test (± 1 S.D.) was

subsequently performed offline to reduce analytical noise. Measurement precision

was determined as 0.4 ppm for CO2 concentration, 0.06 ‰ for 13C and 0.09 ‰ for

18O (Schauer et al. RCMS, in press).

Figure 7. Vertical profiles of CO2 concentration, 13C and 18O of CO2

measured in the residual boundary layer (RBL) on the morning of May 20

and July 22, 2004. Sampling height is referenced to the mean sea level

(MSL). Profiles of 13C closely mirrored those of CO2 concentration, while

profiles of 18O resembled those of 13C. For the early morning profile on

both sampling days, the impact of respiration was evident in the NBL, where

higher CO2, lower 13C and lower 18O values were found below 3500 m.

More negative 13C values inside NBL were associated with CO2 released

from decomposition of 13C-depleted plant material and microbial respiration.

The smaller values of 18O in CO2 within NBL reflected the fact that soil CO2

efflux carried the 18O signature of soil water, which was relatively depleted

in 18O compared to that of leaf water. The strong 18O gradient close to the

ground suggested that soil respiration has a profound impact on ecosystem

respiration in mountain ecosystems. A transition layer where air entrained

from RBL into NBL can be clearly identified in the early morning owing to

the capping of thermal inversion. The vertical gradient observed on May 20

was much smaller compared to that measured at 6:00 am on July 22, which

can be explained by the lower respiration rate in spring associated with lower

temperature. Another explanation is that a deeper NBL on May 20 might

dilute respired CO2 with a greater air volume. On July 22, a second profile

measured at 8:00 am indicated a more well mixed boundary layer, suggesting

a rapid mixing from thermal convection.

Results

Vertical profiles

Figure 8. Comparison between vertical profiles of CO2

concentration, 13C and 18O of atmospheric CO2 measured in the

morning and in the afternoon of July 20, 2004. Lower altitude was

achieved during morning flights in order to collect air samples in

the NBL. While respiration clearly showed its impact on NBL

during morning hours, afternoon profiles showed a reversed trend

with lower CO2 concentration and higher 13C values closer to the

ground. This reversed vertical profile was due to CO2 uptake via

photosynthesis where plants were known to discriminate against 13C, leaving the atmosphere enriched in 13CO2.

Figure 9. Comparison between 3 vertical profiles of CO2

concentration, 13C and 18O of atmospheric CO2 measured

on the morning of July 29, 2004. All 3 profiles showed great

similarities for the vertical structure of atmospheric

boundary layer except a more negative 18O value at 2000 m

at 8:00. Figures 7-9 indicated that a transition layer where

air aloft entrained into surface boundary layer commonly

occurred between 2000 - 3000 m MSL. The top of the

nocturnal boundary layer extended as high as 1000 m above

the ground during these flights.

Figure 10. Comparison between diurnal variations of CO2 concentration, 13C and 18O

of atmospheric CO2 measured above forest canopies and in the CBL. Two days were

selected representing spring and summer conditions. First notice the rapid decrease in

CO2 concentration above forests in the early morning, owing to the break up of

nocturnal boundary layer. Above-canopy CO2 concentration remains roughly constant in

the afternoon, a well-mixed condition achieved by strong convective turbulence. Midday

values of CO2 concentration were about 7 ppm lower on July 29 than May 20. This

difference in the CBL CO2 concentration reflects the higher photosynthetic uptake in the

northern hemisphere in the summer. Values of 13C largely mirror those of CO2

concentration. There were times that measurements in CBL were nearly identical to

those above the canopy (e.g., early morning on May 20 and in the afternoon of July 29).

However, there were also times when CBL measurements were clearly different from

those measured above forests. For instance, CO2 concentration was higher while 13CO2

was lower in the CBL from 7 to 10 am on July 29. This may be due to effects of remnant

CO2 from the previous night forming a residual boundary layer on the aircraft

measurements. Such influences were dismissed after the convective boundary layer grew

in the afternoon from surface heating, giving closer comparisons to canopy values. 18O

values in CO2 measured in CBL were very comparable to those measured above forests,

which clearly showed an enrichment effects of photosynthesis on the morning of July

29.

Diurnal variations of CO2, 13C and 18O

Keeling plot

RCCO

SlopeC 13

2

13

Keeling (1958) GCA 13:322

Figure 11. Using a linear regression between the inverse of CO2 concentration

and carbon isotope ratio, one can estimate the carbon isotope ratio of ecosystem

respiration, denoted here as 13CR. This is known as the Keeling plot approach.

Data shown here are collected in a C3-C4 tallgrass prairie, which is not a part of

the ACME study. It is shown to demonstrate that flux-weighted 13CR values

should fall between boundaries set by C3 and C4 photosynthesis had both

sources were mixed into the sampling footprint. This is more likely to happen

for samples collected in the CBL which integrates sources over a large spatial

area.Figure 12. Sunrise at the Jeffco airport,

Broomfield Colorado, shortly before an

early morning flight. Air samples were

collected in the early morning to capture

respiration signals before the nocturnal

boundary layer dissipates.

Carbon isotope ratio of respiration

Figure 13. Comparisons between carbon isotope ratios of

respiration measured in CBL (13CCBL) and within forest canopies

(13CR). For the 3 selected days, values of 13CCBL were slightly

more positive, considerably more positive and indistinguishable

from those of 13CR on May 20, May 28 and July 22 respectively.

More positive values in CBL could be interpreted as that samples

collected in CBL integrate fluxes over a larger spatial area,

therefore they are likely to be affected by non-local sources of C4

plants afar from the mountain regions. Here sampling height is

referenced to the mean sea level (MSL).Table 1. Comparisons between 13CCBL and

13CR values for aircraft data when ground-

based measurements were also available. All

the 13CCBL values were consistently greater

than or indistinguishable from 13CR values in

both May and July periods. Our interpretation

is that aircraft data were subject to influences

of C4 photosynthesis several kilometers away

from the sampling area, whereas 13CR

measurements only represent respiration from

coniferous forests below. Values of 13CR were

relatively constant and more robust compared

to 13CCBL values, which is partially due to

greater CO2 ranges obtained near the ground.

Discussion

• Measurements of CO2 concentrations, 13C and 18O of atmospheric CO2 were conducted by filling flasks in the

atmospheric boundary layer and on the ground in a subalpine forest in the Colorado mountain regions. These

measurements are an integral part of an extensive program (ACME) to investigate carbon sequestration in

mountain landscapes, potentially responsible for a significant fraction of carbon uptake in the U.S. The data

shown here will provide an independent check to ensure the quality of online CO2 concentration measurements.

Simultaneous measurements in the boundary layer and above forest canopies also provide a unique opportunity

to characterize boundary layer dynamics, including venting of surface respired CO2 from the nocturnal boundary

layer and the mixing of convective turbulence.

• While the data are preliminary, they promise to provide a great utility for model calibration and validation.

Quantifying carbon sequestration in the mountains will have to rely on models to interpret observations, whereas

model results are only useful if they can be verified by actual measurements. The isotope data will provide

additional information to constrain modeled fluxes from inverse calculations based on CO2 concentration

measurements.

• Isotopic discrimination by terrestrial biomes is an important parameter in global inverse models. Lai et al.

(2004, GBC:18, GB1041; 2004 GCB, in press) proposed a useful experimental approach to estimate carbon

isotope discrimination during photosynthesis at the canopy scale. However, current estimates of the

discrimination factor are sensitive to uncertainties in the background atmosphere. Flask measurements

conducted here will allow for further examinations on this issue.

Acknowledgements: This research is supported, in part, by the Terrestrial Carbon Processes (TCP)

program by the office of Science (BER), U.S. Department of Energy under Grant No. DE-FG03-

00ER63012, the NSF Biocomplexity program, and NCAR Biogeosciences initiative.