Comprehensive ecosystem model-data synthesis using multiple data sets at two temperate forest free-air CO2enrichment experiments: Model performance at ambient CO2concentration
Free‐air CO2 enrichment (FACE) experiments provide a remarkable wealth of data which can be used to evaluate and improve terrestrial ecosystem models (TEMs). In the FACE model‐data synthesis project, 11 TEMs were applied to two decadelong FACE experiments in temperate forests of the southeastern U.S.—the evergreen Duke Forest and the deciduous Oak Ridge Forest. In this baseline paper, we demonstrate our approach to model‐data synthesis by evaluating the models' ability to reproduce observed net primary productivity (NPP), transpiration, and leaf area index (LAI) in ambient CO2 treatments. Model outputs were compared against observations using a range of goodness‐of‐fit statistics. Many models simulated annual NPP and transpiration within observed uncertainty. We demonstrate, however, that high goodness‐of‐fit values do not necessarily indicate a successful model, because simulation accuracy may be achieved through compensating biases in component variables. For example, transpiration accuracy was sometimes achieved with compensating biases in leaf area index and transpiration per unit leaf area. Our approach to model‐data synthesis therefore goes beyond goodness‐of‐fit to investigate the success of alternative representations of component processes. Here we demonstrate this approach by comparing competing model hypotheses determining peak LAI. Of three alternative hypotheses—(1) optimization to maximize carbon export, (2) increasing specific leaf area with canopy depth, and (3) the pipe model—the pipe model produced peak LAI closest to the observations. This example illustrates how data sets from intensive field experiments such as FACE can be used to reduce model uncertainty despite compensating biases by evaluating individual model assumptions.
%0 Journal Article
%1 Walker_2014
%A Walker, Anthony P.
%A Hanson, Paul J.
%A Kauwe, Martin G. De
%A Medlyn, Belinda E.
%A Zaehle, Sönke
%A Asao, Shinichi
%A Dietze, Michael
%A Hickler, Thomas
%A Huntingford, Chris
%A Iversen, Colleen M.
%A Jain, Atul
%A Lomas, Mark
%A Luo, Yiqi
%A McCarthy, Heather
%A Parton, William J.
%A Prentice, I. Colin
%A Thornton, Peter E.
%A Wang, Shusen
%A Wang, Ying-Ping
%A Warlind, David
%A Weng, Ensheng
%A Warren, Jeffrey M.
%A Woodward, F. Ian
%A Oren, Ram
%A Norby, Richard J.
%D 2014
%I American Geophysical Union (AGU)
%J Journal of Geophysical Research: Biogeosciences
%K dukeface face facemds facemip ornlface
%N 5
%P 937--964
%R 10.1002/2013jg002553
%T Comprehensive ecosystem model-data synthesis using multiple data sets at two temperate forest free-air CO2enrichment experiments: Model performance at ambient CO2concentration
%U https://doi.org/10.1002%2F2013jg002553
%V 119
%X Free‐air CO2 enrichment (FACE) experiments provide a remarkable wealth of data which can be used to evaluate and improve terrestrial ecosystem models (TEMs). In the FACE model‐data synthesis project, 11 TEMs were applied to two decadelong FACE experiments in temperate forests of the southeastern U.S.—the evergreen Duke Forest and the deciduous Oak Ridge Forest. In this baseline paper, we demonstrate our approach to model‐data synthesis by evaluating the models' ability to reproduce observed net primary productivity (NPP), transpiration, and leaf area index (LAI) in ambient CO2 treatments. Model outputs were compared against observations using a range of goodness‐of‐fit statistics. Many models simulated annual NPP and transpiration within observed uncertainty. We demonstrate, however, that high goodness‐of‐fit values do not necessarily indicate a successful model, because simulation accuracy may be achieved through compensating biases in component variables. For example, transpiration accuracy was sometimes achieved with compensating biases in leaf area index and transpiration per unit leaf area. Our approach to model‐data synthesis therefore goes beyond goodness‐of‐fit to investigate the success of alternative representations of component processes. Here we demonstrate this approach by comparing competing model hypotheses determining peak LAI. Of three alternative hypotheses—(1) optimization to maximize carbon export, (2) increasing specific leaf area with canopy depth, and (3) the pipe model—the pipe model produced peak LAI closest to the observations. This example illustrates how data sets from intensive field experiments such as FACE can be used to reduce model uncertainty despite compensating biases by evaluating individual model assumptions.
@article{Walker_2014,
abstract = {Free‐air CO2 enrichment (FACE) experiments provide a remarkable wealth of data which can be used to evaluate and improve terrestrial ecosystem models (TEMs). In the FACE model‐data synthesis project, 11 TEMs were applied to two decadelong FACE experiments in temperate forests of the southeastern U.S.—the evergreen Duke Forest and the deciduous Oak Ridge Forest. In this baseline paper, we demonstrate our approach to model‐data synthesis by evaluating the models' ability to reproduce observed net primary productivity (NPP), transpiration, and leaf area index (LAI) in ambient CO2 treatments. Model outputs were compared against observations using a range of goodness‐of‐fit statistics. Many models simulated annual NPP and transpiration within observed uncertainty. We demonstrate, however, that high goodness‐of‐fit values do not necessarily indicate a successful model, because simulation accuracy may be achieved through compensating biases in component variables. For example, transpiration accuracy was sometimes achieved with compensating biases in leaf area index and transpiration per unit leaf area. Our approach to model‐data synthesis therefore goes beyond goodness‐of‐fit to investigate the success of alternative representations of component processes. Here we demonstrate this approach by comparing competing model hypotheses determining peak LAI. Of three alternative hypotheses—(1) optimization to maximize carbon export, (2) increasing specific leaf area with canopy depth, and (3) the pipe model—the pipe model produced peak LAI closest to the observations. This example illustrates how data sets from intensive field experiments such as FACE can be used to reduce model uncertainty despite compensating biases by evaluating individual model assumptions.},
added-at = {2019-08-16T17:37:49.000+0200},
author = {Walker, Anthony P. and Hanson, Paul J. and Kauwe, Martin G. De and Medlyn, Belinda E. and Zaehle, Sönke and Asao, Shinichi and Dietze, Michael and Hickler, Thomas and Huntingford, Chris and Iversen, Colleen M. and Jain, Atul and Lomas, Mark and Luo, Yiqi and McCarthy, Heather and Parton, William J. and Prentice, I. Colin and Thornton, Peter E. and Wang, Shusen and Wang, Ying-Ping and Warlind, David and Weng, Ensheng and Warren, Jeffrey M. and Woodward, F. Ian and Oren, Ram and Norby, Richard J.},
biburl = {https://www.bibsonomy.org/bibtex/2bbfca1b4c324d1230bd8a52c88bfd00e/karinawilliams},
doi = {10.1002/2013jg002553},
interhash = {a806221c47276c1a901157f21ce2ee91},
intrahash = {bbfca1b4c324d1230bd8a52c88bfd00e},
journal = {Journal of Geophysical Research: Biogeosciences},
keywords = {dukeface face facemds facemip ornlface},
month = may,
number = 5,
pages = {937--964},
publisher = {American Geophysical Union ({AGU})},
timestamp = {2019-08-16T17:37:49.000+0200},
title = {Comprehensive ecosystem model-data synthesis using multiple data sets at two temperate forest free-air {CO}2enrichment experiments: Model performance at ambient {CO}2concentration},
url = {https://doi.org/10.1002%2F2013jg002553},
volume = 119,
year = 2014
}