3. Simulations using E3SM land surface model

5. Future work

1. Introduction

2 .Field measurements

4. Regional SIF

5. Future work

Biophysical Processes and Feedback Mechanisms Controlling the Methane Budget

of an Amazonian PeatlandFenghui Yuan1, Timothy J. Griffis1*, Daniel T. Roman2, Jeffrey D. Wood3, Angela Lafuente4, Erik Lilleskov2, Hinsby Cadillo-Quiroz5,

Randall Kolka2, Jhon Ever Rengifo6, Lizardo Fachin6, Rod Chimner4, Craig Wayson2, Kristell Hergoualc’h7, Daniel M. Ricciuto8

1University of Minnesota; 2USDA Forest Service; 3University of Missouri; 4Michigan Technological University; 5Arizona State University; 6Instituto de Investigaciones de la Amazonia Peruana, Peru; 7Center for International Forest Research, Indonesia; 8Oak Ridge National Laboratory; *Correspondence: tgriffis@umn.edu (PI)

• Eddy covariance system for

measuring carbon and energy

fluxes have been established and

maintained for nearly 3 years.

• In 2021, automatic 16-

chamber-based system for

measuring soil and trunk CO2

and CH4 emissions has been

successfully installed based upon

microtopography and tree

species.

FIGURE 2. Location of PE-QFR AmeriFlux site. Also shown is

the global distribution of peatlands. (Xu et al. 2018).

FIGURE 1. Experimental and modeling approach illustrating the links among

experimental process studies and model development, validation and forecasting.

FIGURE 6. Comparing diel patterns of

modeled and observed CO2 and CH4 fluxes

for wet and dry seasons in 2019.

• Established in 2017 （AmeriFlux ID： PE-QFR）• Location: Quistococha, Peru (73º19′ 08.1" W; 3º50′ 03.9" S)

• Elevation: 104 m

• IGBP classification: Permanent Wetland

• Dominant vegetation: Mauritia flexuosa L.f. (21.3m height)

• Soil peat layer thickness: 1.92–2.45 m

• MAT=27.2 ºC

• MAP=2753.2 mm

References:[1] Griffis T.J. et al. 2020. Hydrometeorological sensitivities of net ecosystem carbon dioxide and methane

exchange of an Amazonian palm swamp peatland. Agricultural and Forest Meteorology, 295:108167.

[2] Ricciuto D.M., et al. 2021. An integrative model for soil biogeochemistry and methane processes: I.

model structure and sensitivity analysis. JGR-Biogeosciences. (In press)

[3] Yuan F., et al. 2021 An integrative model for soil biogeochemistry and methane processes : II. warming

and elevated CO2 effects on peatland CH4 emissions. JGR-Biogeosciences. (In press)

[4] Haren J., et al. 2021. A versatile gas flux chamber reveals high tree stem CH4 emissions in Amazonian

peatland. Agricultural and Forest Meteorology, 307：108504.

[5] Xu J., et al. 2018 PEATMAP: Refining estimates of global peatland distribution based on a meta-

analysis. CATENA 160: 134–140.

Acknowledgments: This work is supported by the U.S. Department of Energy, Office of Science, Office of Biological and Environmental Research, Terrestrial Ecosystem Science Program (DE-SC0020167). The data used have been submitted to the AmeriFlux program and available at: https://ameriflux.lbl.gov/sites/siteinfo/PE-QFR.

PROJECT SUMMARY

FIGURE 3. Trunk chamber gas

concentration traces for flux

measurements.

Automatic chamber-based system

• The model is derived from ELM-

SPRUCE. (Ricciuto et al. 2021)

• After parameter optimization and

several modifications of key

algorithms with observed data, the

ELM model can capture the seasonal

and diurnal CO2 and CH4 fluxes

overall, although there is still

considerable room for improvement.

• The new algorithms are mainly:

(1) soil hydraulic property function

involving an updated water retention

curve with empirical tropical soil

data; (2) dynamic vcmax25 function

according to seasonal phenological

and physiological changes of tropical

species;(3) water stress scalar

function in CH4 processes which is

improved as a dynamic function of

water table level.

• Tropical peatlands are a major CH4

source, but large uncertainties persist in

their source estimates and significant

knowledge gaps still exist in their process

understanding because of a lack of CH4

observations in the tropics.

• We aim to address these questions by

providing much needed data on CH4

fluxes in tropical peatlands and

developing and testing the Energy

Exascale Earth System Model (E3SM)

land surface component (ELM) for

tropical regions which has been

successfully applied in a boreal forest

peatland (ELM-SPRUCE version).

• Installation of air sample profile system for Lagrangian

analysis of CH4 and CO2 exchange.

• Evaluate flux partitioning and resolve vertical source/sink

distribution based on chamber and profile observations.

• Improve our understanding of the response mechanisms of

GPP to wet/dry seasons in tropical peatlands based on deeper

SIF analyses.

• Finalize the parameter optimizations and function

improvements of ELM before the end of 2021, and then

simulate the climate sensitivities of carbon fluxes to different

scenarios for this tropical peatland.

FIGURE 4. Workflow of observations‐

based off-line modeling optimization.

FIGURE 5. Comparing seasonal patterns

of modeled and observed CO2 and CH4

fluxes for wet and dry seasons in 2019.

FIGURE 7: Taylor diagrams showing correlations coefficients, standard deviations and RMSEs between daily observations and ELM simulations in wet and dry seasons.

Table 1. Parameters for sensitivity analysis.

PE-QFR

FIGURE 9: Spatial and temporal variation of sun induced

fluorescence (SIF) in the Pastaza-Marañón foreland basin.

FIGURE 10: Mean SIF of different

land cover classes within the Pastaza-

Marañón foreland basin.

• Regional patterns of photosynthesis

are being investigated using SIF and

other remote sensing products.

FIGURE 8: Sensitivity analysis for model responses of CO2 and CH4 fluxes, GPP and ecosystem respiration (ER) to 20 parameters in Table 1.

Parameter Definition Parameter DefinitionP1 bdnr Bulk denitrification rate P11 r_mort Mortality rate

P2 br_mrBase rate for maintenance

respiration (MR)P12 stem_leaf New stem carbon per new leaf carbon

P3 decomp_depth_efoldinge-folding depth for

decompositionP13 k_dom

Decomposition rate of dissolved

organic matter

P4 flnr Fraction of leaf N in RuBisco P14 m_dAceProdACmax Maximum rate of acetic acid

production from available carbon

P5 froot_leaf Fine root to leaf allocation ratio P15 m_dACMinQ10Temperature sensitivity of available

carbon mineralization

P6 frootcn Fine root C:N ratio P16 m_dGrowRAceMethanogensGrowth rate of acetoclastic

methanogens

P7 grperc Growth respiration fraction P17 m_dH2ProdAcemaxMaximum reaction rate of conversion

of H2 and CO2 to acetic acid

P8 ksomfacDecomposition fraction of soil

organic matterP18 m_dYAceMethanogens

Growth efficiency of acetoclastic

methanogens

P9 leafcn Leaf carbon/nitrogen (C:N) ratio P19 m_dKCH4OxidCH4Half-saturation coefficient of CH4

oxidation for CH4 concentration

P10 q10_mr Temperature sensitivity for MR P20 Nfix Nitrogen fixation rate