Modeling Environmental Controls on Net Ecosystem CO 2 Exchange
of a Tropical Bog Symon Mezbahuddin 1, Robert Grant 1 and Takashi
Hirano 2 1 Department of Renewable Resources, University of
Alberta, Canada 2 Research Faculty of Agriculture, Hokkaido
University, Japan Slide 2 Introduction Bogs are the peatland
ecosystems that get water and nutrient inputs only through
precipitation Drier weather resultant of increased climatic events
such as El-Nio could alter balance between tropical peatland
ecosystem photosynthesis and respiration and contribute to
increasing atmospheric CO 2 Eddy covariance (EC) net ecosystem
productivity (NEP) estimates of natural tropical peatlands range
from -296 to -594g C m -2 y -1 [Hirano et al. 2009] Ecosystem
processes X multiple environmental factors cause large variations
and uncertainties in predictions of peatland ecosystem response to
climate change Slide 3 Objectives To examine how wetter vs. drier
weather conditions cause diurnal, seasonal and interannual
variations in net ecosystem CO 2 exchange of a tropical bog
[Sebangau River, Palangka Raya, Central Kalimantan, Indonesia; MAT:
26.3 C and MAP: 2600 mm] by using hourly time step process based
ecosystem model ecosys (Grant et al. 2001) tested against valid
eddy covariance data. *Eddy covariance and micrometeorological data
(2002-2005) were collected from flux station over the site (2 3042
S; 114 211 E ) through AsiaFlux Network. Slide 4 Objectives contd.
2003 (Total precipitation = 2291 mm) 2005 (Total precipitation =
2620 mm) (m 3 m -3 ) Figure: Hourly measured precipitation and soil
water content at 0-20cm depth during 2003 and 2005 over Palangka
Raya peat swamp forest, Indonesia Dry year Wet year Slide 5
Conceptual Hypotheses NEP = GPP - RE Increased atmospheric dryness
Increased vapour pressure deficit Higher potential transpiration
Reduced precipitation Reduced soil water content Decreased soil,
root and canopy water potentials Increased stomatal resistance and
consequent stomatal closure Less actual transpiration at a cost of
CO 2 fixation and hence reduced GPP Hypothesis - I Slide 6
Conceptual Hypotheses contd. NEP = GPP - RE Reduced soil water
content Reduced heat capacity of soil Increased soil temperature
Water table drawdown Increased peat aeration Rapid aerobic peat
decomposition Higher heterotrophic respiration Increased ecosystem
respiration Faster biogeochemical reaction Hypothesis - II Oxic
Anoxic WT Slide 7 Hypotheses contd. Surface peat desiccation
Inhibition of heterotrophic respiration at surface layers Partial
or complete offset of any increment in heterotrophic respiration
stimulated by deeper peat aeration Reduction or elimination of
drier and warmer weather effects on NEP through its impact on RE
However, Hypothesis - III WT Desiccation Slide 8 Study Site Figure:
Map of the study site Picture courtesy: (Hirano et al. 2007) Tower
fetches: 0.4 km N; >5 km S; 0.5 km E; 1.5 km W Woody peat
>10,000 years BP Natural swamp forest ~ 4m thick Drainage,
partial thinning ~8 years before flux tower installation Slide 9
Modeling Experiment Figure: Modeling experiment layout (a)
horizontal and (b) vertical (b) A modeled tropical peat swamp
ecosystem of 100 years spin up run using repeated hourly weather
cycles (2002- 2005) The spinning period allowed the modeled
ecosystem to grow, regenerate and attain dynamic C equilibrium Soil
and vegetation characteristics are from site measured data reported
in different studies (b) (a) Slide 10 Preliminary Results Figure:
Hourly measured (dots) and simulated (lines) precipitation, soil
water content at 0-20cm depth and daily water table depths during
2003 over Palangka Raya peat swamp forest, Indonesia 2003 (dry
year) (m 3 m -3 ) Slide 11 Preliminary Results contd. (m 3 m -3 )
2005 (wet year) Figure: Hourly measured (dots) and simulated
(lines) precipitation, soil water content at 0-20cm depth and daily
water table depths during 2005 over Palangka Raya peat swamp
forest, Indonesia Slide 12 Preliminary Results contd. Figure:
Hourly measured vapour pressure deficit during 2005 and 2003 over
Palangka Raya peat swamp forest, Indonesia 2005 (wet year) 2003
(dry year) Slide 13 Preliminary Results contd. 2005 (wet year)
Figure: Hourly measured (dots) and simulated (lines) ecosystem net
radiation (R n ), latent heat (LE) and sensible heat (H) fluxes
during 2005 and 2003 over Palangka Raya peat swamp forest,
Indonesia 2003 (dry year) + influxes - effluxes H/LE=0.36 (R n
>700 Wm -2 ) H/LE=0.62 (R n >700 Wm -2 ) Slide 14 Preliminary
Results contd. 2005 (wet year) Figure: Hourly measured (closed
symbol), gap filled (open symbol) and simulated (lines) ecosystem
CO 2 fluxes during 2005 and 2003 over Palangka Raya peat swamp
forest, Indonesia 2003 (dry year) + influxes - effluxes 0.25 mm
Slide 15 Preliminary Results contd. ParametersUnitsn 1a1a 2b2bR2R23
RMSD 4 RMSE CO 2 flux molm -2 s -1 121870.790.900.854.5 RnRn Wm -2
26304 160.930.9910 - LE flux Wm -2 13924-311.070.767057 H flux Wm
-2 142820.940.564432 (0-20cm) m 3 m -3 263040.110.700.520.02- T s
(5cm) CC 263042.900.800.610.47- Table: Regression results of
simulated vs measured hourly parameters (2003-2005) n 1a1a
2b2bR2R23 RMSD 444-0.120.870.770.10 1 intercept and 2 slope from
regression of simulated over measured data; 3 root mean square for
differences from regression of measured over simulated data; 4 root
mean square of random measurement errors calculated for the
measured data based on algorithms described in Richardson et al.
(2006) Table: Regression results of simulated vs measured daily
water table depths (m) (2004-2005) R n = Net radiation; LE = Latent
heat flux; H= Sensible heat; = Soil water content; T s = Soil
temperature Slide 16 Preliminary Results contd. Figure: Hourly
measured (dots) and simulated (lines) soil water content; three day
moving averages of measured (dots) and simulated (lines)
evapotranspiration and net ecosystem productivity during 2003 over
Palangka Raya peat swamp forest, Indonesia (m 3 m -3 ) Open symbols
have >50% gap filled data +NEP = C sink - NEP = C source 2003
(dry year) Slide 17 Preliminary Results contd. Figure: Hourly
measured (dots) and simulated (lines) soil water content; three day
moving averages of measured (dots) and simulated (lines)
evapotranspiration and net ecosystem productivity during 2005 over
Palangka Raya peat swamp forest, Indonesia (m 3 m -3 ) Open symbols
have >50% gap filled data +NEP = C sink - NEP = C source 2005
(wet year) Slide 18 Preliminary Results contd. Components of
ecosystem carbon budget 2003 (drier year)2005 (wetter year)Values
from other studies EcosysECEcosysEC
-------------------------------------------------g C m -2
-----------------------------------------------------------------
GPP33983461 1 3578-3370 5 RaRa 1570-1431-- R root 795-954-- RhRh
1450-1324-- R soil 2245-2278-2100 4 2130 5 RE38153838 1 3709--
NEP-217-378 1 -131-441 1 -542 2 522 3 Table: Annual ecosystem
carbon balance for Palangka Raya Drained forest, Indonesia during
2003 and 2005 1 Hirano et al. 2007; 2 Henson 2005; 3 Suzuki et al.
1999; 4 Melling et al. 2005; 5 Kosugi et al. 2008; 5 Jauhianen et
al. 2005 Slide 19 Summary There was not a complete shift in
ecosystem energy balance due to water stress resultant of prolonged
dry season. Still a reduction in Bowen ratio (H/LE) due to drier
weather condition resulted lower GPP contributing to lower NEP as
evident in both EC-gap filled and simulated CO 2 fluxes. Slide 20
Ecosys couldnt up to this point fully capture the large negative
NEPs at the onset of the dry season. So, we remain inconclusive
about our hypothesis of a possible increase in ecosystem
respiration stimulated by water table drawdown in drier years. We
are currently working on that. Peatlands widely vary in their
physicochemical characteristics. Rigorous testing of process based
models based on our current fundamental understanding across
contrasting peatlands would be interesting to improve our insights
on peat carbon cycling. Summary contd. Slide 21 Acknowledgements
1.Mary Louise Imrie Travel Grant 2.AsiaFlux Network 3.NSERC
Discovery Grant 4.Faculty of Graduate Studies and Research,
University of Alberta 5.University of Alberta Computing Facilities
6.All of my lab group members. Thank You