Bayesian integration of flux tower data into a process-based simulator for quantifying uncertainty in simulated output
Parameters of a process-based forest growth simulator are difficult or impossible to obtain from field observations. Reliable estimates can be obtained using calibration against observations of output and state variables. In this study, we present a Bayesian framework to calibrate the widely use...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Copernicus Publications
2018-01-01
|
| Series: | Geoscientific Model Development |
| Online Access: | https://www.geosci-model-dev.net/11/83/2018/gmd-11-83-2018.pdf |
| _version_ | 1811343043479994368 |
|---|---|
| author | R. Raj C. van der Tol N. A. S. Hamm A. Stein |
| author_facet | R. Raj C. van der Tol N. A. S. Hamm A. Stein |
| author_sort | R. Raj |
| collection | DOAJ |
| description | Parameters of a process-based forest growth simulator are difficult or
impossible to obtain from field observations. Reliable estimates can be
obtained using calibration against observations of output and state
variables. In this study, we present a Bayesian framework to calibrate the
widely used process-based simulator Biome-BGC against estimates of gross
primary production (GPP) data. We used GPP partitioned from flux tower
measurements of a net ecosystem exchange over a 55-year-old Douglas fir stand
as an example. The uncertainties of both the Biome-BGC parameters and the
simulated GPP values were estimated. The calibrated parameters leaf and fine
root turnover (LFRT), ratio of fine root carbon to leaf carbon (FRC : LC),
ratio of carbon to nitrogen in leaf (C : N<sub>leaf</sub>), canopy water
interception coefficient (<i>W</i><sub>int</sub>), fraction of leaf nitrogen in
RuBisCO (FLNR), and effective soil rooting depth (SD) characterize the photosynthesis and carbon and nitrogen
allocation in the forest. The calibration improved the root mean square error
and enhanced Nash–Sutcliffe efficiency between simulated and flux tower
daily GPP compared to the uncalibrated Biome-BGC. Nevertheless, the seasonal
cycle for flux tower GPP was not reproduced exactly and some overestimation
in spring and underestimation in summer remained after calibration. We
hypothesized that the phenology exhibited a seasonal cycle that was not
accurately reproduced by the simulator. We investigated this by calibrating
the Biome-BGC to each month's flux tower GPP separately. As expected, the
simulated GPP improved, but the calibrated parameter values suggested that
the seasonal cycle of state variables in the simulator could be improved. It
was concluded that the Bayesian framework for calibration can reveal features
of the modelled physical processes and identify aspects of the process
simulator that are too rigid. |
| first_indexed | 2024-04-13T19:22:46Z |
| format | Article |
| id | doaj.art-c978e2e56dfb4bdf8d911256f609cdf4 |
| institution | Directory Open Access Journal |
| issn | 1991-959X 1991-9603 |
| language | English |
| last_indexed | 2024-04-13T19:22:46Z |
| publishDate | 2018-01-01 |
| publisher | Copernicus Publications |
| record_format | Article |
| series | Geoscientific Model Development |
| spelling | doaj.art-c978e2e56dfb4bdf8d911256f609cdf42022-12-22T02:33:29ZengCopernicus PublicationsGeoscientific Model Development1991-959X1991-96032018-01-01118310110.5194/gmd-11-83-2018Bayesian integration of flux tower data into a process-based simulator for quantifying uncertainty in simulated outputR. Raj0C. van der Tol1N. A. S. Hamm2A. Stein3Faculty of Geo-Information Science and Earth Observation (ITC), University of Twente, P.O. Box 217, 7514 AE Enschede, the NetherlandsFaculty of Geo-Information Science and Earth Observation (ITC), University of Twente, P.O. Box 217, 7514 AE Enschede, the NetherlandsFaculty of Geo-Information Science and Earth Observation (ITC), University of Twente, P.O. Box 217, 7514 AE Enschede, the NetherlandsFaculty of Geo-Information Science and Earth Observation (ITC), University of Twente, P.O. Box 217, 7514 AE Enschede, the NetherlandsParameters of a process-based forest growth simulator are difficult or impossible to obtain from field observations. Reliable estimates can be obtained using calibration against observations of output and state variables. In this study, we present a Bayesian framework to calibrate the widely used process-based simulator Biome-BGC against estimates of gross primary production (GPP) data. We used GPP partitioned from flux tower measurements of a net ecosystem exchange over a 55-year-old Douglas fir stand as an example. The uncertainties of both the Biome-BGC parameters and the simulated GPP values were estimated. The calibrated parameters leaf and fine root turnover (LFRT), ratio of fine root carbon to leaf carbon (FRC : LC), ratio of carbon to nitrogen in leaf (C : N<sub>leaf</sub>), canopy water interception coefficient (<i>W</i><sub>int</sub>), fraction of leaf nitrogen in RuBisCO (FLNR), and effective soil rooting depth (SD) characterize the photosynthesis and carbon and nitrogen allocation in the forest. The calibration improved the root mean square error and enhanced Nash–Sutcliffe efficiency between simulated and flux tower daily GPP compared to the uncalibrated Biome-BGC. Nevertheless, the seasonal cycle for flux tower GPP was not reproduced exactly and some overestimation in spring and underestimation in summer remained after calibration. We hypothesized that the phenology exhibited a seasonal cycle that was not accurately reproduced by the simulator. We investigated this by calibrating the Biome-BGC to each month's flux tower GPP separately. As expected, the simulated GPP improved, but the calibrated parameter values suggested that the seasonal cycle of state variables in the simulator could be improved. It was concluded that the Bayesian framework for calibration can reveal features of the modelled physical processes and identify aspects of the process simulator that are too rigid.https://www.geosci-model-dev.net/11/83/2018/gmd-11-83-2018.pdf |
| spellingShingle | R. Raj C. van der Tol N. A. S. Hamm A. Stein Bayesian integration of flux tower data into a process-based simulator for quantifying uncertainty in simulated output Geoscientific Model Development |
| title | Bayesian integration of flux tower data into a process-based simulator for quantifying uncertainty in simulated output |
| title_full | Bayesian integration of flux tower data into a process-based simulator for quantifying uncertainty in simulated output |
| title_fullStr | Bayesian integration of flux tower data into a process-based simulator for quantifying uncertainty in simulated output |
| title_full_unstemmed | Bayesian integration of flux tower data into a process-based simulator for quantifying uncertainty in simulated output |
| title_short | Bayesian integration of flux tower data into a process-based simulator for quantifying uncertainty in simulated output |
| title_sort | bayesian integration of flux tower data into a process based simulator for quantifying uncertainty in simulated output |
| url | https://www.geosci-model-dev.net/11/83/2018/gmd-11-83-2018.pdf |
| work_keys_str_mv | AT rraj bayesianintegrationoffluxtowerdataintoaprocessbasedsimulatorforquantifyinguncertaintyinsimulatedoutput AT cvandertol bayesianintegrationoffluxtowerdataintoaprocessbasedsimulatorforquantifyinguncertaintyinsimulatedoutput AT nashamm bayesianintegrationoffluxtowerdataintoaprocessbasedsimulatorforquantifyinguncertaintyinsimulatedoutput AT astein bayesianintegrationoffluxtowerdataintoaprocessbasedsimulatorforquantifyinguncertaintyinsimulatedoutput |