Addressing Challenges in Simulating Inter–annual Variability of Gross Primary Production

Created in BioRender. De, R. (2024) https://BioRender.com/i01x768

Description

This repository contains codes to perform analysis and reproduce figures of our research paper:

De, R., Bao, S., Koirala, S., Brenning, A., Reichstein, M., Tagesson, T., Liddell, M., Ibrom, A., Wolf, S., Šigut, L., Hörtnagl, L., Woodgate, W., Korkiakoski, M., Merbold, L., Black, T. A., Roland, M., Klosterhalfen, A., Blanken, P. D., Knox, S., Sabbatini, S., Gielen, B., Montagnani, L., Fensholt, R., Wohlfahrt, G., Desai, A. R., Paul-Limoges, E., Galvagno, M., Hammerle, A., Jocher, G., Ruiz Reverter, B., Holl, D., Chen, J., Vitale, L., Arain, M. A., and Carvalhais, N. (2025). Addressing Challenges in Simulating Inter–annual Variability of Gross Primary Production. Journal of Advances in Modeling Earth Systems, 17(5), e2024MS004697. https://doi.org/10.1029/2024MS004697

We used majorly the following two models in our study. It is highly recommended to get acquainted with the following two research papers before using our codes.

1. Optimality-based model: P-model of Mengoli Mengoli, G., Agustí-Panareda, A., Boussetta, S., Harrison, S. P., Trotta, C., and Prentice, I. C. (2022). Ecosystem photosynthesis in land-surface models: A first-principles approach incorporating acclimation, Journal of Advances in Modeling Earth Systems, 14, https://doi.org/10.1029/2021MS002767

2. Semi-empirical model: Bao model Bao, S., Wutzler, T., Koirala, S., Cuntz, M., Ibrom, A., Besnard, S., Walther, S., Šigut, L., Moreno, A., Weber, U., Wohlfahrt,695 G., Cleverly, J., Migliavacca, M., Woodgate, W., Merbold, L., Veenendaal, E., and Carvalhais, N. (2022). Environment-sensitivity functions for gross primary productivity in light use efficiency models, Agricultural and Forest Meteorology, 312, 108 708, https://doi.org/10.1016/j.agrformet.2021.108708

Disclaimer

The codes are written to be compatible with computing platforms and filestructure of MPI-BGC, Jena. It maybe necessary to adapt the certain parts of codes to make them compatible with other computing platforms. All the data should be prepared in NetCDF format and variables should be named as per the code. While the actual data used for analysis is not shared in this repository due to large sizes, all the data source are cited in the relevant paper and openly accessible. Corresponding author (Ranit De, rde@bgc-jena.mpg.de or de.ranit19@gmail.com) can be contacted in regards to code usage and data preparation. Any usage of codes are sole responsibility of the users.

Structure

• site_info: This folder contains two .csv files: (1) SiteInfo_BRKsite_list.csv, this one is necessary so that the code knows data for which all sites are available and can access site specific metadata for preparing results, such as data analysis and grouping of sites according to site characteristics, (2) site_year_list.csv lists all the site–years available for site–year specific optimization. This list also contains site–years which are not of good quality, and later gets excluded during data processing steps.
• src: This folder basically contains all source codes. It has four folders: (1) common folder contains all the scripts which are common for both the Optimality-based (P-model and its variations) and the semi-empirical model (Bao model and its variations), (2) lue_model contains model codes and cost function specific to the semi-empirical model (Bao model and its variations), (3) p_model contains model codes and cost function specific to the Optimality-based (P-model and its variations), and (4) postprocess contains all the scripts to prepare exploratory plots after parameterization and forward runs.
• prep_figs: This folder contains all the scripts to reproduce the figures which are presented in our research paper and its supplementary document. All modelling experiments and their relevant data must be available to reproduce the figures and their relative paths should be correctly mentioned at result_path_coll.py.

How to run codes?

• Create a conda environment and install dependencies. Dependencies are listed in requirements.yml.
• Open model_settings.xlsx and specify all the experiment parameters from dropdown or by typing as described in the worksheet.
• Run main_opti_and_run_model.py (except PFT specific optimization). For PFT specific optimization, run submit_pft_opti_jobs.py. If you want parallel processing on a high performance computing (HPC) platform, other settings are necessary based on the platform you are using. PFT specific optimization and global optimization can only be performed using parallel processing on a HPC as multi-site data must be used. See send_slurm_job.sh for a sample job submission recipie to a HPC platform using slurm as a job scheduler.

How to cite?

Research paper: - BibTeX @article{De_IAV_GPP_2025, author = {De, Ranit and Bao, Shanning and Koirala, Sujan and Brenning, Alexander and Reichstein, Markus and Tagesson, Torbern and Liddell, Michael and Ibrom, Andreas and Wolf, Sebastian and Šigut, Ladislav and Hörtnagl, Lukas and Woodgate, William and Korkiakoski, Mika and Merbold, Lutz and Black, T. Andrew and Roland, Marilyn and Klosterhalfen, Anne and Blanken, Peter D. and Knox, Sara and Sabbatini, Simone and Gielen, Bert and Montagnani, Leonardo and Fensholt, Rasmus and Wohlfahrt, Georg and Desai, Ankur R. and Paul-Limoges, Eugénie and Galvagno, Marta and Hammerle, Albin and Jocher, Georg and Reverter, Borja Ruiz and Holl, David and Chen, Jiquan and Vitale, Luca and Arain, M. Altaf and Carvalhais, Nuno}, title = {Addressing Challenges in Simulating Inter–Annual Variability of Gross Primary Production}, journal = {Journal of Advances in Modeling Earth Systems}, volume = {17}, number = {5}, pages = {e2024MS004697}, doi = {https://doi.org/10.1029/2024MS004697}, url = {https://agupubs.onlinelibrary.wiley.com/doi/abs/10.1029/2024MS004697}, year = {2025} } - APA De, R., Bao, S., Koirala, S., Brenning, A., Reichstein, M., Tagesson, T., Liddell, M., Ibrom, A., Wolf, S., Šigut, L., Hörtnagl, L., Woodgate, W., Korkiakoski, M., Merbold, L., Black, T. A., Roland, M., Klosterhalfen, A., Blanken, P. D., Knox, S., Sabbatini, S., Gielen, B., Montagnani, L., Fensholt, R., Wohlfahrt, G., Desai, A. R., Paul-Limoges, E., Galvagno, M., Hammerle, A., Jocher, G., Ruiz Reverter, B., Holl, D., Chen, J., Vitale, L., Arain, M. A., and Carvalhais, N. (2025). Addressing Challenges in Simulating Inter–annual Variability of Gross Primary Production. Journal of Advances in Modeling Earth Systems, 17(5), e2024MS004697. https://doi.org/10.1029/2024MS004697

This repository: - BibTeX @software{De2025Codes, author = {De, Ranit}, title = {{Scripts for analyses presented in ``Addressing challenges in simulating inter–annual variability of gross primary production''}}, month = apr, year = 2025, publisher = {Zenodo}, version = {v1.3-published}, doi = {10.5281/zenodo.13729514}, url = {https://github.com/de-ranit/iav_gpp_p_bao} } - APA De, R. (2025). Scripts for analyses presented in "Addressing challenges in simulating inter–annual variability of gross primary production" (v1.3-published). Zenodo. https://doi.org/10.5281/zenodo.13729514

Change Log:

v1.3-published - updated readme with reference to the published article.

v1.2-preprint - updated upper limit of LUEmax model parameter for Bao model and its variations. - added an analysis on GPP uncertianty at annual scale. - added statistical significance testing for model performance across PFTs and Bioclimatic regions. - updated codes for figures as per reviewers' suggestions. - updated license.

v1.1-preprint - Updated readme with citations for the preprint and the Zenodo repository.

v1.0-preprint - Initial code for submission to a Zenodo repository and publication of preprint.

License

This work is licensed under a MIT License.

License logo is created by ExcaliburZero, used under CC BY 3.0 license