rsofun

R framework for site-scale simulations of ecosystem processes.
https://github.com/geco-bern/rsofun

Category: Biosphere
Sub Category: Plants and Vegetation

Keywords

dgvm growth modeling p-model simulation vegetation-dynamics

Keywords from Contributors

acclimation optimality-theory photosynthesis

Last synced: about 7 hours ago
JSON representation

Repository metadata

Implements the Simulating Optimal FUNctioning framework for site-scale simulations of ecosystem processes, including model calibration. It contains Fortran 90 modules for the P-model, SPLASH, and BiomeE models.

• Host: GitHub
• URL: https://github.com/geco-bern/rsofun
• Owner: geco-bern
• License: gpl-3.0
• Created: 2018-10-24T15:48:42.000Z (about 7 years ago)
• Default Branch: master
• Last Pushed: 2025-12-01T08:43:20.000Z (24 days ago)
• Last Synced: 2025-12-07T19:54:21.077Z (17 days ago)
• Topics: dgvm, growth, modeling, p-model, simulation, vegetation-dynamics
• Language: Fortran
• Homepage: https://geco-bern.github.io/rsofun/
• Size: 194 MB
• Stars: 30
• Watchers: 8
• Forks: 33
• Open Issues: 17
• Releases: 12
• Metadata Files:
  • Readme: README.md
  • Changelog: NEWS.md
  • License: COPYING
  • Citation: CITATION.cff

README.md

install.packages("rsofun")
library("rsofun")

if(!require(remotes)){install.packages("remotes")}
remotes::install_github("geco-bern/rsofun")
library("rsofun")

if(!require(remotes)){install.packages("remotes")}
remotes::install_github("geco-bern/rsofun", build_vignettes = TRUE)
library("rsofun")

R CMD build --no-manual --no-build-vignettes rsofun

R CMD INSTALL -c --preclean  *.tar.gz

# define model parameter values from previous
# work
params_modl <- list(
    kphio              = 0.04998,    # setup ORG in Stocker et al. 2020 GMD
    kphio_par_a        = 0.0,        # set to zero to disable temperature-dependence of kphio
    kphio_par_b        = 1.0,
    soilm_thetastar    = 0.6 * 240,  # to recover old setup with soil moisture stress
    soilm_betao        = 0.0,
    beta_unitcostratio = 146.0,
    rd_to_vcmax        = 0.014,      # value from Atkin et al. 2015 for C3 herbaceous
    tau_acclim         = 30.0,
    kc_jmax            = 0.41
  )

# run the model for these parameters
output <- rsofun::runread_pmodel_f(
  p_model_drivers,
  par = params_modl
  )

settings <- list(
  method              = "GenSA",
  metric              = cost_rmse_pmodel,
  control = list(
    maxit = 100),
  par = list(
    kphio = list(lower=0.02, upper=0.2, init = 0.05)
    )
)

# calibrate the model and optimize free parameters
pars <- calib_sofun(
    drivers = p_model_drivers,  
    obs = p_model_validation,
    settings = settings,
    # extra arguments passed to the cost function:
    targets = "gpp",             # define target variable GPP
    par_fixed = params_modl[-1]  # fix non-calibrated parameters to previous 
                                 # values, removing kphio
  )

Citation (CITATION.cff)

cff-version: 1.2.0
message: >-
  If you use this software, please cite it using the
  metadata from this file.
  The corresponding documentation paper can be found at egusphere-2025-1260
title: 'rsofun: The P-Model and BiomeE Modelling Framework'
type: software
license: GPL-3.0-only
repository: 'http://dx.doi.org/10.32614/CRAN.package.rsofun'
repository-code: 'https://github.com/fabern/zenodo_sandbox'
repository-artifact: 'https://doi.org/10.5281/zenodo.3712928'
url: 'https://doi.org/10.5281/zenodo.3712928'
contact:
- affiliation: University Bern
  family-names: Stocker
  given-names: Benjamin David
  orcid: https://orcid.org/0000-0003-2697-9096
  email: benjamin.stocker@gmail.com
authors:
- affiliation: University Bern
  family-names: Stocker
  given-names: Benjamin David
  orcid: https://orcid.org/0000-0003-2697-9096
- affiliation: '@bluegreen-labs'
  family-names: Hufkens
  given-names: Koen
  orcid: https://orcid.org/0000-0002-5070-8109
- affiliation: University Bern
  family-names: "Ar\xE1n Paredes"
  given-names: Josefa
  orcid: https://orcid.org/0009-0006-7176-2311
- affiliation: University Bern
  family-names: Bernhard
  given-names: Fabian
  orcid: https://orcid.org/0000-0003-0338-0961
- affiliation: University Bern
  family-names: Marcadella
  given-names: Mayeul
  orcid: https://orcid.org/0000-0001-8555-3808
keywords:
- dgvm
- growth
- modeling
- p-model
- simulation
- vegetation-dynamics
abstract: '<h1>rsofun</h1>

  <p>An R package for Simulating Optimal FUNctioning (rsofun). A model for site-scale
  simulations of ecosystem processes. The package contains the following modules:</p>

  <p>- P-model for leaf-level acclimation of photosynthesis from <a href="https://gmd.copernicus.org/preprints/gmd-2019-200/">Stocker
  et al. (2019)</a>.<br>- SPLASH for bioclimatic variables, including the surface
  radiation budget and the soil water balance from <a href="https://doi.org/10.5194/gmd-10-689-2017">Davis
  et al. (2017)</a>.<br>- BiomeE for comprehensive simulations of ecosystem carbon
  and water cycling, tree growth, and tree cohort-explicit forest dynamics following
  the Perfect Plasticity Approximation, from <a href="https://doi.org/10.5194/bg-12-2655-2015">Weng
  et al., (2015)</a>.</p>

  <h2>Installation</h2>

  <h3>Stable release</h3>

  <p>To install the current stable release use a CRAN repository:</p>

  <p><code>install.packages("rsofun")</code><br><code>library("rsofun")</code></p>

  <h3>Development release</h3>

  <p>To install the latest development release of the package run the following commands
  to install rsofun directly from GitHub:</p>

  <p><code>if(!require(remotes)){install.packages("remotes")}</code><br><code>remotes::install_github("geco-bern/rsofun")</code><br><code>library("rsofun")</code><br><br></p>

  <p>**NOTE:** Installing from GitHub requires compilation of Fortran and C source
  code contained in {rsofun}. To enable compiling source code, install <a href="https://cran.r-project.org/bin/windows/Rtools/">Rtools</a>
  on Windows, or <a href="https://developer.apple.com/xcode/">Xcode</a> and the <a
  href="https://github.com/fxcoudert/gfortran-for-macOS">GNU Fortran compiler on Mac</a>
  (see also ''Mandatory tools'' <a href="https://mac.r-project.org/tools/">here</a>).
  On Linux, the gfortran compiler is usually installed already.</p>

  <p>Vignettes are not rendered by default, if you want to include additional documentation
  please use:</p>

  <p><code>if(!require(remotes)){install.packages("remotes")}</code><br><code>remotes::install_github("geco-bern/rsofun",
  build_vignettes = TRUE)</code><br><code>library("rsofun")</code></p>

  <h3>From source&nbsp;</h3>

  <p>Assuming <code>rsofun</code> is the location of the source directory, on can
  build the R package (with extension .tar.gz) from the command line using in a bash
  shell:<br><br><code>R CMD build --no-manual --no-build-vignettes rsofun</code><br><br>The
  package can then be installed with:</p>

  <p><code>R CMD INSTALL -c --preclean &nbsp;*.tar.gz</code></p>

  <p>, where the star <code>*</code> can be replaced by the name of the package produced
  at the previous step.</p>

  <h2>Use</h2>

  <p>Below sections show the ease of use of the package in terms of model parameter
  specification and running both a single run or optimizing the parameters for a given
  site (or multiple sites). For an in depth discussion we refer to the <a href="https://geco-bern.github.io/rsofun/articles/">vignettes</a>.</p>

  <h3>Running model</h3>

  <p>With all data prepared we can run the P-model using <code>runread_pmodel_f()</code>.
  This function takes the nested data structure and runs the model site by site, returning
  nested model output results matching the input drivers. In R:</p>

  <p><br><code># define model parameter values from previous</code><br><code># work</code><br><code>params_modl
  &lt;- list(</code><br><code>&nbsp; &nbsp; kphio &nbsp; &nbsp; &nbsp; &nbsp; &nbsp;
  &nbsp; &nbsp;= 0.04998, &nbsp; &nbsp;# setup ORG in Stocker et al. 2020 GMD</code><br><code>&nbsp;
  &nbsp; kphio_par_a &nbsp; &nbsp; &nbsp; &nbsp;= 0.0, &nbsp; &nbsp; &nbsp; &nbsp;#
  set to zero to disable temperature-dependence of kphio</code><br><code>&nbsp; &nbsp;
  kphio_par_b &nbsp; &nbsp; &nbsp; &nbsp;= 1.0,</code><br><code>&nbsp; &nbsp; soilm_thetastar
  &nbsp; &nbsp;= 0.6 * 240, &nbsp;# to recover old setup with soil moisture stress</code><br><code>&nbsp;
  &nbsp; soilm_betao &nbsp; &nbsp; &nbsp; &nbsp;= 0.0,</code><br><code>&nbsp; &nbsp;
  beta_unitcostratio = 146.0,</code><br><code>&nbsp; &nbsp; rd_to_vcmax &nbsp; &nbsp;
  &nbsp; &nbsp;= 0.014, &nbsp; &nbsp; &nbsp;# value from Atkin et al. 2015 for C3
  herbaceous</code><br><code>&nbsp; &nbsp; tau_acclim &nbsp; &nbsp; &nbsp; &nbsp;
  = 30.0,</code><br><code>&nbsp; &nbsp; kc_jmax &nbsp; &nbsp; &nbsp; &nbsp; &nbsp;
  &nbsp;= 0.41</code><br><code>&nbsp; )</code></p>

  <p><code># run the model for these parameters</code><br><code>output &lt;- rsofun::runread_pmodel_f(</code><br><code>&nbsp;
  p_model_drivers,</code><br><code>&nbsp; par = params_modl</code><br><code>&nbsp;
  )</code><br><br></p>

  <h3>Parameter optimization</h3>

  <p>To optimize new parameters based upon driver data and a validation dataset we
  must first specify an optimization strategy and settings, as well as a cost function
  and parameter ranges. In R:</p>

  <p><br><code>settings &lt;- list(</code><br><code>&nbsp; method &nbsp; &nbsp; &nbsp;
  &nbsp; &nbsp; &nbsp; &nbsp;= "GenSA",</code><br><code>&nbsp; metric &nbsp; &nbsp;
  &nbsp; &nbsp; &nbsp; &nbsp; &nbsp;= cost_rmse_pmodel,</code><br><code>&nbsp; control
  = list(</code><br><code>&nbsp; &nbsp; maxit = 100),</code><br><code>&nbsp; par =
  list(</code><br><code>&nbsp; &nbsp; kphio = list(lower=0.02, upper=0.2, init = 0.05)</code><br><code>&nbsp;
  &nbsp; )</code><br><code>)</code><br><br></p>

  <p>`rsofun` supports both optimization using the `GenSA` and `BayesianTools` packages.
  The above statement provides settings for a `GenSA` optimization approach. For this
  example the maximum number of iterations is kept artificially low. In a real scenario
  you will have to increase this value orders of magnitude. Keep in mind that optimization
  routines rely on a cost function, which, depending on its structure influences parameter
  selection. A limited set of cost functions is provided but the model structure is
  transparent and custom cost functions can be easily written. More details can be
  found in the "Parameter calibration and cost functions" vignette.</p>

  <p>In addition starting values and ranges are provided for the free parameters in
  the model. Free parameters include: parameters for the quantum yield efficiency
  `kphio`, `kphio_par_a` and `kphio_par_b`, soil moisture stress parameters `soilm_thetastar`
  and `soilm_betao`, and also `beta_unitcostratio`, `rd_to_vcmax`, `tau_acclim` and
  `kc_jmax` (see `?runread_pmodel_f`). Be mindful that with newer versions of `rsofun`
  additional parameters might be introduced, so re-check vignettes and function documentation
  when updating existing code.</p>

  <p>With all settings defined the optimization function `calib_sofun()` can be called
  with driver data and observations specified. Extra arguments for the cost function
  (like what variable should be used as target to compute the root mean squared error
  (RMSE) and previous values for the parameters that aren''t calibrated, which are
  needed to run the P-model).</p>

  <p><br><code># calibrate the model and optimize free parameters</code><br><code>pars
  &lt;- calib_sofun(</code><br><code>&nbsp; &nbsp; drivers = p_model_drivers, &nbsp;</code><br><code>&nbsp;
  &nbsp; obs = p_model_validation,</code><br><code>&nbsp; &nbsp; settings = settings,</code><br><code>&nbsp;
  &nbsp; # extra arguments passed to the cost function:</code><br><code>&nbsp; &nbsp;
  targets = "gpp", &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; # define target variable
  GPP</code><br><code>&nbsp; &nbsp; par_fixed = params_modl[-1] &nbsp;# fix non-calibrated
  parameters to previous&nbsp;</code><br><code>&nbsp; &nbsp; &nbsp; &nbsp; &nbsp;
  &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp;#
  values, removing kphio</code><br><code>&nbsp; )</code><br><br></p>

  <h2>Data and code for model documentation paper (Paredes et al., in rev.)</h2>

  <p>Versioned releases of this repository are deposited on Zenodo (see badge at the
  top of the README file). Code to reproduce the analysis and plots presented here
  is contained in this repository (subdirectory `<code>analysis/</code>`) and is demonstrated
  on the model documentation website (<a href="https://geco-bern.github.io/rsofun/">https://geco-bern.github.io/rsofun/</a>,
  article &lsquo;Sensitivity analysis and calibration interpretation&rsquo;).</p>

  <p>The model forcing and evaluation data is based on the publicly available FLUXNET2015
  data for the site FR-Pue, prepared by FluxDataKit v3.4.2 (10.5281/zenodo.14808331),
  taken here as a subset of the originally published data for years 2007-2012. It
  is accessible through the {rsofun} R package and contained as part of this repository
  (subdirectory&nbsp;<code>data/</code>) as CSV and as files. Outputs of the analysis
  presented here are archived in the <code>analysis/paper_results_files/</code>&nbsp;subfolder.</p>

  <p>The model documentation paper is currently under review.<br>A preprint is available
  at: https://www.biorxiv.org/content/10.1101/2023.11.24.568574v3</p>

  <h2>References</h2>

  <p>Stocker, B. D., Wang, H., Smith, N. G., Harrison, S. P., Keenan, T. F., Sandoval,
  D., Davis, T., and Prentice, I. C.: P-model v1.0: an optimality-based light use
  efficiency model for simulating ecosystem gross primary production, Geosci. Model
  Dev., 13, 1545&ndash;1581, https://doi.org/10.5194/gmd-13-1545-2020, 2020.</p>

  <p>Davis, T. W., Prentice, I. C., Stocker, B. D., Thomas, R. T., Whitley, R. J.,
  Wang, H., Evans, B. J., Gallego-Sala, A. V., Sykes, M. T., and Cramer, W.: Simple
  process-led algorithms for simulating habitats (SPLASH v.1.0): robust indices of
  radiation, evapotranspiration and plant-available moisture, Geoscientific Model
  Development, 10, 689&ndash;708, doi:10.5194/gmd-10-689-2017, URL http: //www.geosci-model-dev.net/10/689/2017/,
  2017.</p>

  <p>Weng, E. S., Malyshev, S., Lichstein, J. W., Farrior, C. E., Dybzinski, R., Zhang,
  T., Shevliakova, E., and Pacala, S. W.: Scaling from individual trees to forests
  in an Earth system modeling framework using a mathematically tractable model of
  height-structured competition, Biogeosciences, 12, 2655&ndash;2694, https://doi.org/10.5194/bg-12-2655-2015,
  2015.</p>

  <h2>Acknowledgements</h2>

  <p>The {rsofun} is part of the LEMONTREE project and funded by Schmidt Futures and
  under the umbrella of the Virtual Earth System Research Institute (VESRI).&nbsp;</p>'

Owner metadata

• Name: GECO
• Login: geco-bern
• Email:
• Kind: organization
• Description: Geocomputation and Earth Observation at the University of Bern
• Website: geco-group.org
• Location: Switzerland
• Twitter: SteniBocker
• Company:
• Icon url: https://avatars.githubusercontent.com/u/93120344?v=4
• Repositories: 22
• Last ynced at: 2023-03-15T20:55:13.905Z
• Profile URL: https://github.com/geco-bern

GitHub Events

Total

• Create event: 37
• Release event: 2
• Issues event: 90
• Watch event: 6
• Delete event: 26
• Issue comment event: 193
• Push event: 312
• Pull request review event: 96
• Pull request review comment event: 69
• Pull request event: 61
• Fork event: 4

Last Year

• Create event: 21
• Release event: 1
• Issues event: 21
• Watch event: 5
• Delete event: 16
• Issue comment event: 48
• Push event: 192
• Pull request review comment event: 22
• Pull request review event: 37
• Pull request event: 41
• Fork event: 2

Committers metadata

Last synced: 3 days ago

Committer domains:

• creaf.uab.cat: 2
• unibe.ch: 2
• pop-os.localdomain: 2
• usys.ethz.ch: 1
• gmail.comm: 1
• gl141-210.ethz.ch: 1
• eu-login-47.euler.ethz.ch: 1
• eu-login-24-ng.euler.ethz.ch: 1
• eu-login-18.euler.ethz.ch: 1
• eu-login-11-ng.euler.ethz.ch: 1
• eu-login-07-ng.euler.ethz.ch: 1
• gl142-110.ethz.ch: 1
• imperial.ac.uk: 1

Issue and Pull Request metadata

Last synced: 8 days ago

Package metadata

• Total packages: 2
• Total downloads:
  • cran: 164 last-month
• Total dependent packages: 0 (may contain duplicates)
• Total dependent repositories: 0 (may contain duplicates)
• Total versions: 9
• Total maintainers: 1

Dependencies

Score: 12.17496890047884