Dominant role of soil moisture in controlling Nighttime Net Ecosystem Exchange in Sub-Humid West African Ecosystems

Résumé fourni par la source

# MI_Wavelet_analysis **_Dominant role of soil moisture in controlling Nighttime Net Ecosystem Exchange in Sub-Humid West African Ecosystems_** > Last updated: 2026-07-03 This repository provides the raw data and the R / MATLAB scripts to reproduce the analyses, figures and results presented in the scientific manuscript "Dominant role of soil moisture in controlling Nighttime Net Ecosystem Exchange in Sub-Humid West African Ecosystems", submitted to JGR Biogeosciences. The workflow combines **mutual information** (to quantify statistical dependencies between NEE and its drivers) and **wavelet analysis** (to resolve those dependencies across time scales) in order to investigate the scale-dependent controls of nighttime net ecosystem exchange (NEE). ## Public Release ``` ``` --- ## Table of Contents 1. [Requirements](#requirements) 2. [Installation](#installation) 3. [Data](#data) 4. [Repository Structure](#repository-structure) 5. [How to Run](#how-to-run) 6. [Reproducing the Figures](#reproducing-the-figures) 7. [Acknowledgements](#acknowledgements) 8. [Contact](#contact) --- ## Requirements | Software | Version | Purpose | |---|---|---| | R | 4.2.0 | Data processing, gap-filling, mutual information | | MATLAB | R2022b (v9.13.0) | Wavelet power, coherence and phase analysis | **R packages** ```R install.packages(c("data.table", "lubridate", "randomForest", "varrank", "REddyProc", "openeddy", "tidyverse", "ggplot2")) ``` **MATLAB toolboxes** (included in the `Processing` subfolder) - Wavelet coherence routines from Grinsted et al. (2004) — folder `wavelet-coherence-master` (see [Acknowledgements](#acknowledgements)) - *Cumulative Arcwise Significance of Global Wavelet Power and Global Coherence Spectra* (v1.0.0.0) by Justin Schulte — folder `arcwise_sigtest`, available on [MATLAB Central File Exchange](https://fr.mathworks.com/matlabcentral/profile/authors/6932485) > Adjust the package list above to match the actual dependencies used by your scripts. --- ## Installation Download and unpack the zip archive. Set the main project path at the top of each script so that it matches your local environment before running anything. --- ## Data The data sets used in this study are provided by the **AMMA-CATCH** observatory and are available through the following DOIs. Each data set covers the two flux stations, **Nalohou** and **Bellefoungou**. | Data set | Content | DOI | |---|---|---| | AE.H2OFlux_Odc | Meteorology, fluxes and soil moisture at the flux stations | [10.17178/AMMA-CATCH.AE.SHFlux_Odc](http://dx.doi.org/10.17178/AMMA-CATCH.AE.SHFlux_Odc) | | CE.SW_Odc | Complementary soil temperature and humidity | [10.17178/AMMA-CATCH.CE.SW_Odc](http://dx.doi.org/10.17178/AMMA-CATCH.CE.SW_Odc) | | CL.Rain_Od | Rainfall | [10.17178/AMMA-CATCH.CL.Rain_Od](http://dx.doi.org/10.17178/AMMA-CATCH.CL.Rain_Od) | The scripts read these inputs and write intermediate and final outputs to the `Processing/` and `Figures/` subfolders. --- ## Repository Structure ``` MI_Wavelet_analysis/ ├── Processing/ # Analysis pipeline + dependencies │ ├── NEE_data_post_process.r # Data preparation │ ├── Mutual_information_analysis.r # Mutual information analysis │ ├── NEE_data_MDS_gap_filling.r # Gap-filling (MDS) │ ├── NEE_data_RF_gap_filling.r # Gap-filling of long gaps (Random Forest) │ ├── Global_power_analysis.m # Global wavelet power │ ├── Coherence_analysis.m # Wavelet coherence │ ├── Phase_analysis.m # Phase difference (wavelets) │ ├── wavelet-coherence-master/ # Grinsted et al. (2004) toolbox │ └── arcwise_sigtest/ # Schulte arcwise significance toolbox ├── Figures/ # Code to generate the article figures ├── LICENSE └── README.md ``` ### Pipeline overview | Stage | File | Description | |---|---|---| | 1. Preparation | `NEE_data_post_process.r` | Data preparation | | 2. Mutual information | `Mutual_information_analysis.r` | Analysis using mutual information | | 3. Gap-filling (short) | `NEE_data_MDS_gap_filling.r` | Marginal distribution sampling (MDS) | | 4. Gap-filling (long) | `NEE_data_RF_gap_filling.r` | Random Forest (RF); also prepares data for coherence analysis | | 5. Wavelet power | `Global_power_analysis.m` | Global wavelet power analysis | | 6. Wavelet coherence | `Coherence_analysis.m` | Wavelet coherence analysis | | 7. Phase | `Phase_analysis.m` | Phase difference analysis using wavelets | --- ## How to Run Run the R stages in order, then the MATLAB stages. **R (interactive)** ```R source("./Processing/NEE_data_post_process.r") source("./Processing/Mutual_information_analysis.r") source("./Processing/NEE_data_MDS_gap_filling.r") source("./Processing/NEE_data_RF_gap_filling.r") ``` **MATLAB** ```matlab run('./Processing/Global_power_analysis.m') run('./Processing/Coherence_analysis.m') run('./Processing/Phase_analysis.m') ``` Make sure the main path is set in accordance with your local settings. --- ## Reproducing the Figures Once the pipeline has produced its outputs, generate the article figures with the scripts in the `Figures/` subfolder. --- ## Acknowledgements The wavelet transform analysis makes use of the source code shared by **Grinsted, A., Moore, J. C., and Jevrejeva, S. (2004)**, *Application of the cross wavelet transform and wavelet coherence to geophysical time series*, Nonlinear Processes in Geophysics, 11, 561–566. The assessment of the statistical significance of the global wavelet power and global coherence spectra uses the toolbox *Cumulative Arcwise Significance of Global Wavelet Power and Global Coherence Spectra* (v1.0.0.0) by **Justin Schulte**, distributed through the MATLAB Central File Exchange. Eddy covariance flux post-processing and gap-filling rely on the **REddyProc** package: Wutzler, T., Lucas-Moffat, A., Migliavacca, M., Knauer, J., Sickel, K., Šigut, L., Menzer, O., and Reichstein, M. (2018): *Basic and extensible post-processing of eddy covariance flux data with REddyProc*, Biogeosciences, 15, 5015–5030, https://doi.org/10.5194/bg-15-5015-2018. Variable ranking based on mutual information uses the **varrank** package: Kratzer, G., and Furrer, R. (2018): *varrank: an R package for variable ranking based on mutual information with applications to observed systemic datasets*, http://arxiv.org/abs/1804.07134. --- ## Contact - **Renaud Koukoui** — romeo.koukoui@imsp-uac.org - **Ossénatou Mamadou** — ossenatou.mamadou@imsp-uac.org ---

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.