Data from: Jensen's inequality reveals how moisture variability controls soil respiration across continental scales

Pere-Rojas, Yulissa1 ; Ghezzehei, Teamrat 1

Published May 22, 2026 on Dryad. https://doi.org/10.5061/dryad.hmgqnk9zj

Data files

May 22, 2026 version files 2.70 MB

processedData.zip

2.63 MB
README.md

69.02 KB

Abstract

Understanding how temporal patterns of moisture variability influence modeled biogeochemical responses remains a central challenge in Earth system science. Jensen's inequality provides a mathematical framework for quantifying when nonlinear processes cause the response to average conditions to differ systematically from the average of responses to actual conditions. We applied this framework to continental-scale AmeriFlux soil moisture (volumetric water content, θ) and temperature data (134.5 million hourly observations from 2,004 sensors) to compare two approaches for modeling soil respiration: one using time-resolved environmental inputs and one using long-term temporal averages. Both approaches use identical nonlinear respiration functions; no flux observations are used for validation. Sensors in dry regions show large Jensen's inequality effects (median difference of -63.6% between approaches) because they experience highly skewed moisture distributions where brief wet periods drive disproportionate simulated respiratory responses. Wet regions show smaller effects (median -27.1%) because they have more uniform moisture distributions. Data density analysis reveals that sensors operating at θ ≈ 0.05 exhibit the largest differences between approaches, while sensors at θ ≈ 0.45 show minimal differences, demonstrating the mechanistic basis for where ecosystems operate on the moisture-respiration curve. Climate gradient analysis shows systematic transitions from large effects in arid systems to moderate effects in humid systems. Depth analysis reveals that surface soils show maximum effects, while deeper soils show reduced effects due to environmental buffering. Moisture-temperature coupling demonstrates systematic negative correlations in water-limited systems, indicating that environmental co-variation modulates modeled biogeochemical responses. Jensen's inequality thus serves as a diagnostic tool for identifying where temporal averaging of environmental inputs introduces the largest differences in simulated soil respiration.

Publication Information

Authors: Yulissa T. Perez Rojas¹ and Teamrat A. Ghezzehei¹

Affiliations:
¹Life and Environmental Sciences Department, University of California, Merced

Journal: JGR-Biogeosciences (under review)

Corresponding Authors:

Yulissa T. Perez Rojas (yperezrojas@ucmerced.edu)
Teamrat A. Ghezzehei (taghezzehei@ucmerced.edu)

This deposit is distributed across two linked repositories, each with the
license appropriate to its content:

Dryad — processedData.zip (computed-results portion only): released into the public domain under the [Creative Commons Zero v1.0 Universal (CC0 1.0) (https://creativecommons.org/publicdomain/zero/1.0/) dedication. This contains the authors' own analytical output (probability distributions, derived metrics, summaries, and visualization outputs) and no third-party data.

Zenodo (Software) — the analysis scripts (.R and .py files): released under the MIT License.

Zenodo (Supplemental Information) — data_zenodo.zip: released under
the [Creative Commons Attribution 4.0 International (CC BY 4.0) (https://creativecommons.org/licenses/by/4.0/) license. This archive contains three folders: rawData/ (AmeriFlux input metadata), processedData/ (AmeriFlux- and SoilGrids-derived files held out of Dryad), and forPaper/ (manuscript figures and tables).

For the CC BY 4.0 portion, you are free to:

Share — copy and redistribute the material in any medium or format
Adapt — remix, transform, and build upon the material for any purpose, even commercially

Under the following terms:

Attribution — You must give appropriate credit, provide a link to the license, and indicate if changes were made

The data_zenodo.zip portion is not released under CC0 because it incorporates data derived from third-party CC BY sources (AmeriFlux BASE and ISRIC SoilGrids; see Data Provenance below), whose attribution requirements must be preserved. The forPaper/ figures and tables are placed in Zenodo (not Dryad) to avoid licensing conflict with their journal publication.

Repository Structure

This deposit lives in two linked repositories:

Dryad: `processedData.zip` (CC0)

Contains the computed-results portion of processedData/. These are the authors' own analytical output and contain no third-party data.

processedData/
├── 2025_AMF_files.csv                     # File inventory (pipeline-generated)
├── probability_analysis_summary.csv       # Probability analysis summary
├── summaries/                             # Statistical summaries
├── visuals/                               # Visualization outputs
├── Joint_prob/                            # Joint probability distributions
├── SWC_prob/                              # Soil moisture distributions
├── TS_prob/                               # Temperature distributions
├── RESP_prob/                             # Respiration distributions
├── RESP_VWC_prob/                         # VWC-driven respiration distributions
├── RESP_TEMP_prob/                        # Temperature-driven respiration distributions
├── profile_pdfs/                          # Profile visualization PDFs
└── README.txt                             # Lists files held out to Zenodo CC BY

The following files are not in the Dryad processedData.zip and live in the Zenodo data_zenodo.zip (under processedData/) because they are derived from AmeriFlux and SoilGrids data:

sensor_metadata_combo_resp.csv
sensorMetadata_final_filetered.csv
sensorMetadata_level2.csv
sensorMetadata.csv
siteMetadata.csv
soilGrids/Complete_soilgriddata.csv
soilGrids/Selected_soilgriddata.csv

Zenodo (Software, MIT): individual scripts

_porosity_density.R                        # Helper functions (sourced by scripts 4 & 5)
1_preProcessData.R                         # Data preprocessing
2_siteCharacteristics.R                    # Site metadata enrichment
3_soilGrids.R                              # Soil property extraction
3b_Rosetta.py                              # Hydraulic parameter calculation
4_soilPhysicalProperties.R                 # Porosity calculations
5_calcMoistureRespiration.R                # Respiration modeling
6_plotRespiration__FINAL.R                 # Visualization generation
7_probabilityAnalysis_V2.R                 # Probability analysis
8_generate_site_level_table.R              # Supplementary table generation

Zenodo (Supplemental, CC BY 4.0): `data_zenodo.zip`

Contains three folders. Required for re-running the pipeline; not required if using the Dryad probability outputs directly.

data_zenodo.zip
├── rawData/
│   ├── BASE_MeasurementHeight_20250430.csv    # AmeriFlux sensor metadata
│   ├── CC-BY-4-0_Sites.csv                    # AmeriFlux site characteristics
│   ├── requested_files_manifest_20250521.txt  # AmeriFlux request manifest
│   └── BASE_2025/                             # NOT INCLUDED — Download from AmeriFlux (>30 GB)
│       └── README.txt                         # Placeholder with instructions
│
├── processedData/                             # AmeriFlux/SoilGrids-derived held-outs
│   ├── sensor_metadata_combo_resp.csv
│   ├── sensorMetadata_final_filetered.csv
│   ├── sensorMetadata_level2.csv
│   ├── sensorMetadata.csv
│   ├── siteMetadata.csv
│   └── soilGrids/
│       ├── Complete_soilgriddata.csv
│       └── Selected_soilgriddata.csv
│
└── forPaper/                                  # Manuscript figures and tables
    ├── *.png                                  # Manuscript figures
    ├── main_body/                             # Main figures and result tables
    │   ├── fig_*.pdf / *.png
    │   └── jensen_table.tex
    └── schematic_diagrams/
        ├── *.pdf / *.png
        ├── Fig_Jensen_Explanation.afdesign
        └── resp_functions.R

Not redistributed

The raw AmeriFlux BASE time series (rawData/BASE_2025/, >30 GB) and the intermediate per-sensor outputs (processedData/stationData/ ~11 GB, processedData/stationData_level2/ ~21 GB) are excluded for size. The raw data is retrieved from the AmeriFlux portal using the included manifest; the intermediates are generated by Scripts 1 and 5 of the pipeline.

Assembling a working directory

To run the analysis pipeline locally, build a single repo/ directory:

repo/
├── (all scripts from Zenodo Software)
├── rawData/                # extracted from data_zenodo.zip (+ BASE_2025/ from AmeriFlux)
├── processedData/          # Dryad processedData.zip MERGED WITH data_zenodo.zip's processedData/
└── forPaper/               # extracted from data_zenodo.zip, or created empty

Scripts use relative paths and expect rawData/, processedData/, and
forPaper/ at the working-directory level.

Data Availability

Raw AmeriFlux Data

Included in data_zenodo.zip under rawData/:

✓ BASE_MeasurementHeight_20250430.csv - Sensor height and metadata information
✓ CC-BY-4-0_Sites.csv - Site characteristics (coordinates, climate, vegetation)
✓ requested_files_manifest_20250521.txt - List of AmeriFlux files used

NOT included (must be downloaded separately):
The raw sensor time series data (>30 GB) must be downloaded from the AmeriFlux Data Portal.

Download Instructions:

Visit the AmeriFlux Management Project database: https://ameriflux.lbl.gov/
For each site used in this study, download the BASE data product (CC-BY-4.0 license)

The complete list of sites is available in:
- rawData/requested_files_manifest_20250521.txt (inside data_zenodo.zip)
- Or after running Script 1: processedData/2025_AMF_files.csv
Place all downloaded AmeriFlux .csv files in the repo/rawData/BASE_2025/ directory

Data Use Policy:
All AmeriFlux data are subject to AmeriFlux data use policies. Please cite individual sites according to their data use requirements.

Data Provenance and Third-Party Licensing

Some files in the Zenodo portion of this deposit are derived from third-party sources released under CC BY 4.0. They are redistributed under CC BY 4.0 with attribution to the original providers; this is the reason those files are in Zenodo (CC BY 4.0) rather than Dryad (CC0).

AmeriFlux BASE. rawData/CC-BY-4-0_Sites.csv,
rawData/BASE_MeasurementHeight_20250430.csv, and the derived metadata
files in data_zenodo.zip's processedData/ (siteMetadata.csv and the
sensorMetadata* files) are derived from AmeriFlux BASE data products.
The specific sites used are listed in
rawData/requested_files_manifest_20250521.txt.

Sites cited: 313 sites were obtained and screened; 242 sites met the soil-sensor co-location and record-length requirements and contributed data to the reported results. Per-site BASE DOIs for the 242 cited sites are provided in the manuscript references (see ameriflux_filtered.bib). Six downloaded sites were not under the AmeriFlux CC-BY-4.0 policy and contributed no data to the analyzed set; they are not cited.

Funding for the AmeriFlux data service was provided by the U.S. Department of Energy Office of Science. The AmeriFlux BASE data were accessed from the AmeriFlux data portal on 21 May 2025. The BASE data pipeline is described by:

Chu, H., Christianson, D. S., Cheah, Y.-W., et al. (2023). AmeriFlux
BASE data pipeline to support network growth and data sharing.
Scientific Data, 10, 614. https://doi.org/10.1038/s41597-023-02531-2

ISRIC SoilGrids. Soil property extracts in data_zenodo.zip's
processedData/soilGrids/ are derived from SoilGrids 2.0, released under
CC BY 4.0:

Poggio, L., de Sousa, L. M., Batjes, N. H., Heuvelink, G. B. M.,
Kempen, B., Ribeiro, E., and Rossiter, D. (2021). SoilGrids 2.0:
producing soil information for the globe with quantified spatial
uncertainty. SOIL, 7, 217–240. https://doi.org/10.5194/soil-7-217-2021

SoilGrids Data

Soil property data are automatically downloaded from the ISRIC SoilGrids database (https://soilgrids.org/) by Script 3. No manual download is required.

Processed Data

Included in the Dryad processedData.zip (CC0):

Probability distributions for all variables
Jensen's inequality calculations (final results)
Statistical summaries and correlations
Visualization outputs
The pipeline-generated file inventory (2025_AMF_files.csv)

Included in data_zenodo.zip under processedData/ (CC BY 4.0):

All sensor and site metadata CSV files (sensor_metadata_combo_resp.csv,
sensorMetadata_final_filetered.csv, sensorMetadata_level2.csv,
sensorMetadata.csv, siteMetadata.csv)
Soil properties from SoilGrids (soilGrids/Complete_soilgriddata.csv,
soilGrids/Selected_soilgriddata.csv)

These are held out of Dryad because they are derived from third-party CC BY 4.0 data (AmeriFlux and SoilGrids); see Data Provenance above.

NOT included (generated by running the pipeline):

processedData/stationData/ - Individual sensor time series (~11 GB, generated by Script 1)
processedData/stationData_level2/ - Enhanced time series with respiration (~21 GB, generated by Script 5)

Total size when pipeline completes: ~38 GB (includes large time series folders)

Note: Users can reproduce all main analyses and figures using only the included metadata files without re-running the full time series processing. Scripts 6–8 use the metadata files and will run directly after merging the Dryad processedData.zip with the held-out files from data_zenodo.zip into a single processedData/ directory alongside the scripts.

Software Requirements

R Environment

R Version: 4.0 or higher recommended

Required R Packages:

dplyr - Data manipulation
data.table - Fast data reading/writing
ggplot2 - Data visualization
maps - Geographic maps
mapproj - Map projections
scales - Scaling functions
patchwork - Plot composition
gridExtra - Grid graphics
grid - Low-level graphics
Ternary - Ternary plots
xtable - LaTeX table generation
parallel - Parallel processing

Installation:

install.packages(c("dplyr", "data.table", "ggplot2", "maps", "mapproj",
                   "scales", "patchwork", "gridExtra", "grid",
                   "Ternary", "xtable", "parallel"))

Python Environment

Python Version: 3.7 or higher

Required Python Packages:

rosetta-soil - Rosetta pedotransfer functions
pandas - Data manipulation
numpy - Numerical computations

Installation:

pip install rosetta-soil pandas numpy

Quick Start Guide

There are two ways to use this repository:

Option A: Reproduce Visualizations and Analysis (No Raw Data Needed)

Best for: Quickly reproducing figures and tables from the manuscript

Since the deposit includes all processed metadata and final results, you can immediately generate visualizations without downloading raw AmeriFlux data or running the full pipeline.

Download and assemble the working directory:

# From Dryad: processedData.zip
# From Zenodo: the individual scripts and data_zenodo.zip
mkdir repo && cd repo
mv /path/to/zenodo/*.R /path/to/zenodo/*.py .
unzip /path/to/zenodo/data_zenodo.zip          # creates rawData/, processedData/, forPaper/
unzip /path/to/dryad/processedData.zip -d . # merges with the held-out processedData/

Install R dependencies:

install.packages(c("dplyr", "data.table", "ggplot2", "maps", "mapproj",
                   "scales", "patchwork", "gridExtra", "grid",
                   "Ternary", "xtable", "parallel"))

Run visualization scripts (from R, with repo/ as working directory):

setwd("/path/to/repo")

# Generate all manuscript figures and analysis
source("6_plotRespiration__FINAL.R")
source("8_generate_site_level_table.R")

Output: Figures and tables will be generated in forPaper/ directory

Option B: Run Complete Pipeline from Raw Data

Best for: Full reproducibility from raw AmeriFlux data

Download and assemble the working directory:

# From Dryad: processedData.zip
# From Zenodo: the individual scripts and data_zenodo.zip
mkdir repo && cd repo
mv /path/to/zenodo/*.R /path/to/zenodo/*.py .
unzip /path/to/zenodo/data_zenodo.zip          # creates rawData/, processedData/, forPaper/
unzip /path/to/dryad/processedData.zip -d . # merges with the held-out processedData/

Download raw AmeriFlux data:
- Follow instructions in Data Availability section
- Place files in repo/rawData/BASE_2025/
Install all dependencies:
- R packages (see Software Requirements)
- Python packages: pip install rosetta-soil pandas numpy

Run complete pipeline (from R, with repo/ as working directory):

setwd("/path/to/repo")

# Step 1-3: Data preprocessing and soil properties
source("1_preProcessData.R")
source("2_siteCharacteristics.R")
source("3_soilGrids.R")                # Requires internet (SoilGrids API queries)

Then in terminal:

cd /path/to/repo
python 3b_Rosetta.py

Back in R:

# Step 4-5: Soil properties and respiration calculations
source("4_soilPhysicalProperties.R")
source("5_calcMoistureRespiration.R")  # Most computationally intensive step

# Step 6-8: Visualization and analysis (can run in parallel)
source("6_plotRespiration__FINAL.R")
source("7_probabilityAnalysis_V2.R")
source("8_generate_site_level_table.R")

Note: Runtime depends on system specifications and number of CPU cores available. Scripts 5-7 use parallel processing where applicable.

Analysis Pipeline

Pipeline Overview

This pipeline processes AmeriFlux soil moisture and temperature sensor data through multiple stages: data preprocessing, site characterization, soil property enhancement, respiration modeling, and probability analysis. The workflow transforms raw AmeriFlux BASE files into analysis-ready datasets with comprehensive soil physical and hydraulic properties.

Pipeline Sequence

The scripts must be run in the following order:

rawData/ (AmeriFlux files & site data)
    ↓
Script 1 → sensorMetadata.csv + stationData/*.csv
    ↓
Script 2 → siteMetadata.csv + enhanced_correlation.csv
    ↓
Script 3 → soilGrids data
    ↓
Script 3b → hydraulic_parameters.csv
    ↓
Script 4 → sensorMetadata_level2.csv + enhanced stationData/*.csv
    ↓
Script 5 → sensor_metadata_combo_resp.csv + stationData_level2/*.csv
    ↓
Scripts 6-8 → Final visualizations and analysis

Detailed Script Documentation

Script 1: Data Preprocessing - `1_preProcessData.R`

Purpose: Initial data cleaning, validation, and sensor metadata creation

Inputs:

rawData/BASE_2025/*.csv - Raw AmeriFlux station data files
rawData/BASE_MeasurementHeight_20250430.csv - Sensor height information

Processing:

Parses soil sensor variables (SWC and TS sensors)
Converts timestamps to standardized format
Replaces -9999 values with NA
Matches SWC sensors with co-located TS sensors
Extracts and standardizes sensor data with timestamps
Calculates data availability statistics and time ranges
Generates correlation analysis between co-located SWC/TS pairs

Outputs:

processedData/2025_AMF_files.csv - Inventory of processed files
processedData/sensorMetadata.csv - Comprehensive sensor information
processedData/stationData/*.csv - One file per sensor with columns:
- TIMESTAMP_START, TIMESTAMP_END, vwc, ts, precip, airtemp
processedData/summaries/site_summary.csv - Site-level statistics
processedData/summaries/profile_summary.csv - Profile-level statistics
processedData/summaries/correlation_summary.csv - SWC-TS correlations
processedData/summaries/overall_summary.csv - Dataset overview
processedData/summaries/swc_ts_correlation_by_depth.pdf - Visualization

Script 2: Site Characteristics - `2_siteCharacteristics.R`

Purpose: Enrich sensor data with site metadata and create site characterization maps

Inputs:

processedData/sensorMetadata.csv (from Script 1)
rawData/CC-BY-4-0_Sites.csv - Site characteristics
processedData/summaries/correlation_summary.csv (from Script 1)

Processing:

Matches sensor metadata with site characteristics (coordinates, climate, vegetation)
Identifies sites with CC-BY-4.0 vs Legacy licenses
Adds data availability flags for each measurement type
Creates enhanced correlation dataset with site metadata
Generates maps showing site distributions by various characteristics

Outputs:

processedData/siteMetadata.csv - Sites with characteristics
processedData/summaries/enhanced_correlation.csv - Correlations + site data
processedData/visuals/maps/sites_by_*.pdf - Site distribution maps
processedData/visuals/correlation_plots/correlation_by_*.pdf - Correlation plots

Script 3: Soil Properties from SoilGrids - `3_soilGrids.R`

Purpose: Extract soil texture, bulk density, and organic carbon from global SoilGrids database

Inputs:

processedData/siteMetadata.csv (from Script 2)

Processing:

Extracts site coordinates (lat/long)
Queries SoilGrids API for soil properties at each site
Retrieves soil texture (sand, silt, clay), bulk density, and SOC by soil layer
Processes multiple soil horizons per site

Outputs:

processedData/soilGrids/Complete_soilgriddata.csv - All variables including quantiles
processedData/soilGrids/Selected_soilgriddata.csv - Filtered for essential variables

Note: Requires internet connection to access SoilGrids API

Script 3b: Hydraulic Parameters - `3b_Rosetta.py`

Purpose: Calculate van Genuchten hydraulic parameters using Rosetta pedotransfer functions

Inputs:

processedData/soilGrids/Selected_soilgriddata.csv (from Script 3)

Processing:

Reads soil texture (sand, silt, clay) and bulk density data
Applies Rosetta pedotransfer functions to predict hydraulic parameters
Calculates van Genuchten parameters: θr, θs, α, n, Ksat
Provides both log-scale and linear-scale parameter values

Outputs:

processedData/soilGrids/soil_hydraulic_parameters.csv
processedData/soilGrids/rosetta_prediction.log

Script 4: Soil Physical Properties - `4_soilPhysicalProperties.R`

Purpose: Calculate porosity, match sensors to soil layers, and add saturation calculations

Dependencies: Sources _porosity_density.R for helper functions

Inputs:

processedData/sensorMetadata.csv (from Script 1)
processedData/soilGrids/Selected_soilgriddata.csv (from Script 3)
processedData/stationData/*.csv (from Script 1)

Processing:

Matches each sensor depth to appropriate SoilGrids soil layer
Calculates soil porosity using organic carbon, bulk density, and texture
Computes corrected porosity (max of calculated porosity and 98th percentile of soil moisture)
Adds saturation columns to all station data files (VWC/porosity)

Outputs:

processedData/sensorMetadata_level2.csv - Enhanced sensor metadata
processedData/stationData/*.csv - Modified in-place with sat column added
processedData/summaries/soil_moisture_quantiles.csv

Note: This script modifies station data files in-place

Script 5: Respiration Calculations - `5_calcMoistureRespiration.R`

Purpose: Calculate soil respiration based on moisture and temperature dependencies, and compute Jensen's inequality metrics

Dependencies: Sources _porosity_density.R for helper functions

Inputs:

processedData/sensorMetadata_level2.csv (from Script 4)
processedData/soilGrids/soil_hydraulic_parameters.csv (from Script 3b)
processedData/siteMetadata.csv (from Script 2)
processedData/stationData/*.csv (from Script 4)

Processing:

Calculates water potential using van Genuchten equations
Computes moisture-dependent respiration
Applies Q10 temperature response function
Combines moisture and temperature effects
Calculates Jensen's inequality metrics comparing response of mean vs mean of responses

Outputs:

processedData/sensor_metadata_combo_resp.csv - MOST CRITICAL FILE
- Contains Jensen's inequality metrics for 2,004 sensors
- Key columns: jensen_ineq_vwc, jensen_ineq_temp, jensen_ineq_vwc_temp
processedData/stationData_level2/*.csv - Enhanced time series with respiration values

Note: Has test mode option (full_run = TRUE/FALSE)

Script 6: Respiration Visualization - `6_plotRespiration__FINAL.R`

Purpose: Create comprehensive visualizations of respiration analysis results

Inputs:

processedData/sensor_metadata_combo_resp.csv (from Script 5)
processedData/summaries/enhanced_correlation.csv (from Script 2)

Outputs:

forPaper/jensen_map.png - Spatial patterns of Jensen's inequality
forPaper/one2one_*.png - Response comparison plots
forPaper/netDiff_*.png - Difference plots
forPaper/veg_clim_*.png - Climate/vegetation analysis
forPaper/main_body/jensen_table.tex - LaTeX Table 3

Script 7: Probability Analysis - `7_probabilityAnalysis_V2.R`

Purpose: Calculate probability distributions and create profile visualizations

Inputs:

processedData/sensor_metadata_combo_resp.csv (from Script 5)
processedData/stationData_level2/*.csv (from Script 5)

Outputs:

processedData/SWC_prob/*_prob.csv - VWC distributions
processedData/TS_prob/*_prob.csv - Temperature distributions
processedData/RESP_*_prob/*_prob.csv - Respiration distributions
processedData/Joint_prob/*_joint_prob.csv - Joint distributions
processedData/profile_pdfs/*_PDFs.pdf - Profile visualizations

Note: Has test mode option (full_run = TRUE/FALSE)

Script 8: Site-Level Analysis - `8_generate_site_level_table.R`

Purpose: Generate supplementary Table S4 showing site-level Jensen's inequality statistics

Inputs:

processedData/sensor_metadata_combo_resp.csv (from Script 5)

Processing:

Calculates site-level median Jensen's inequality (242 sites)
Addresses spatial autocorrelation by aggregating sensors within sites
Computes summary statistics by climate group

Outputs:

forPaper/table_s4_site_level_jensen.tex - LaTeX supplementary table
CSV outputs with site-level data

Critical Intermediate Data Files

The following files are essential for pipeline continuity:

processedData/sensorMetadata.csv (Script 1) - Initial sensor metadata
processedData/siteMetadata.csv (Script 2) - Site characteristics
processedData/sensorMetadata_level2.csv (Script 4) - Enhanced with soil properties
processedData/soilGrids/soil_hydraulic_parameters.csv (Script 3b) - Hydraulic parameters
processedData/sensor_metadata_combo_resp.csv (Script 5) - MOST IMPORTANT
- Contains complete analysis results for 2,004 sensors
- Used by Scripts 6, 7, and 8
processedData/stationData_level2/*.csv (Script 5) - Time series with respiration

Computational Resources

Recommended Hardware

RAM: 16 GB minimum, 32 GB recommended
Storage: 100 GB free space for raw + processed data
CPU: Multi-core processor (scripts use parallel processing)

Parallel Processing

Several scripts utilize the parallel package:

Number of cores used: detectCores() - 1

Troubleshooting

Issue: Script cannot find input files
Solution: Ensure you're running scripts from the repo/ directory where they were extracted

Issue: Out of memory errors
Solution: Increase available RAM or use test mode where available

Issue: SoilGrids API timeout
Solution: Script includes retry logic; check internet connection

Issue: Missing R packages
Solution: Install all packages listed in Software Requirements

Issue: Python Rosetta problems
Solution: Ensure Python 3.7+ and try: pip install --upgrade rosetta-soil

Key Findings

Jensen's inequality applied to continental-scale soil data diagnoses when episodic moisture events dominate over mean conditions in controlling simulated respiration
Moisture distribution patterns, not absolute moisture levels, control the magnitude of temporal averaging differences between modeling approaches
Systematic moisture-temperature coupling in water-limited systems may constrain modeled climate sensitivity in ways current temporal averaging approaches do not capture

Citation

If you use this data or code, please cite:

Perez Rojas, Y.T., and Ghezzehei, T.A. (2025). Jensen's Inequality Reveals How Moisture Variability Controls Soil Respiration Across Continental Scales. JGR-Biogeosciences (under review).

AmeriFlux Site References (242 Sites)

Antonio Antonino (2025). Caatinga Serra Talhada (BR-CST). AmeriFlux AMP. https://doi.org/10.17190/AMF/1562386
Mike Goulden (2025). Santarem-Km83-Logged Forest (BR-Sa3). AmeriFlux AMP. https://doi.org/10.17190/AMF/1245995
Manuel Helbig (2025). Acadia Research Forest (CA-AF1). AmeriFlux AMP. https://doi.org/10.17190/AMF/2479623
Elyn Humphreys (2025). Attawapiskat River Bog (CA-ARB). AmeriFlux AMP. https://doi.org/10.17190/AMF/1480319
Elyn Humphreys (2025). Attawapiskat River Fen (CA-ARF). AmeriFlux AMP. https://doi.org/10.17190/AMF/1480318
T. Andrew Black (2025). British Columbia - 1949 Douglas-fir stand (CA-Ca1). AmeriFlux AMP. https://doi.org/10.17190/AMF/1480300
T. Andrew Black (2025). British Columbia - Clearcut Douglas-fir stand (harvested winter 1999/2000) (CA-Ca2). AmeriFlux AMP. https://doi.org/10.17190/AMF/1480301
T. Andrew Black (2025). British Columbia - Pole sapling Douglas-fir stand (CA-Ca3). AmeriFlux AMP. https://doi.org/10.17190/AMF/1480302
Ralf Staebler (2025). Ontario - Mixed Deciduous, Borden Forest Site (CA-Cbo). AmeriFlux AMP. https://doi.org/10.17190/AMF/1498755
Susan Natali (2025). Churchill Fen Site 3 (CA-CF3). AmeriFlux AMP. https://doi.org/10.17190/AMF/2335573
Charles Bourque (2025). New Brunswick - Charlie Lake site 01 (immature balsam fir forest to be thinned in year 3) (CA-Cha). AmeriFlux AMP. https://doi.org/10.17190/AMF/1436317
Harry McCaughey (2025). Ontario - Groundhog River, Boreal Mixedwood Forest (CA-Gro). AmeriFlux AMP. https://doi.org/10.17190/AMF/1245996
Thomas Andrew Black (2025). British Columbia - Mountain pine beetle-attacked lodgepole pine stand (CA-LP1). AmeriFlux AMP. https://doi.org/10.17190/AMF/1660337
Brian Amiro (2025). Manitoba Agricultural Site 1 (CA-MA1). AmeriFlux AMP. https://doi.org/10.17190/AMF/1617701
Brian Amiro (2025). Manitoba Agricultural Site 2 (CA-MA2). AmeriFlux AMP. https://doi.org/10.17190/AMF/1617702
Brian Amiro (2025). Manitoba Agricultural Site 3 (CA-MA3). AmeriFlux AMP. https://doi.org/10.17190/AMF/1617703
Peter Lafleur (2025). Ontario - Eastern Peatland, Mer Bleue (CA-Mer). AmeriFlux AMP. https://doi.org/10.17190/AMF/2448423
Susan Natali (2025). Mittimatalik (Pond Inlet) Tundra (CA-Mtk). AmeriFlux AMP. https://doi.org/10.17190/AMF/2531141
Charles P.-A. Bourque (2025). New Brunswick - 1967 Balsam Fir - Nashwaak Lake Site 01 (Mature balsam fir forest) (CA-Na1). AmeriFlux AMP. https://doi.org/10.17190/AMF/1436319
Mike Goulden (2025). UCI-1850 burn site (CA-NS1). AmeriFlux AMP. https://doi.org/10.17190/AMF/1245998
Mike Goulden (2025). UCI-1930 burn site (CA-NS2). AmeriFlux AMP. https://doi.org/10.17190/AMF/1245999
Mike Goulden (2025). UCI-1964 burn site (CA-NS3). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246000
Mike Goulden (2025). UCI-1964 burn site wet (CA-NS4). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246001
Mike Goulden (2025). UCI-1981 burn site (CA-NS5). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246002
Mike Goulden (2025). UCI-1989 burn site (CA-NS6). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246003
Mike Goulden (2025). UCI-1998 burn site (CA-NS7). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246004
Warren Helgason (2025). Saskatchewan - Western Boreal, Mature Jack Pine (CA-Ojp). AmeriFlux AMP. https://doi.org/10.17190/AMF/1375199
Hank Margolis (2025). Quebec - 1975 Harvested Black Spruce (HBS75) (CA-Qc2). AmeriFlux AMP. https://doi.org/10.17190/AMF/1419514
Hank A. Margolis (2025). Quebec - Eastern Boreal, Mature Black Spruce (CA-Qfo). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246829
Oliver Sonnentag (2025). Scotty Creek Landscape (CA-SCC). AmeriFlux AMP. https://doi.org/10.17190/AMF/1480303
Brian Amiro (2025). Saskatchewan - Western Boreal, forest burned in 1977 (CA-SF1). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246006
Brian Amiro (2025). Saskatchewan - Western Boreal, forest burned in 1989 (CA-SF2). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246007
Brian Amiro (2025). Saskatchewan - Western Boreal, forest burned in 1998 (CA-SF3). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246008
Alan Barr (2025). Saskatchewan - Western Boreal, Jack Pine forest harvested in 1994 (CA-SJ1). AmeriFlux AMP. https://doi.org/10.17190/AMF/1436320
Alan Barr (2025). Saskatchewan - Western Boreal, Jack Pine forest harvested in 2002 (CA-SJ2). AmeriFlux AMP. https://doi.org/10.17190/AMF/1436321
M. Altaf Arain (2025). Ontario - Turkey Point 2002 Plantation White Pine (CA-TP1). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246009
M. Altaf Arain (2025). Ontario - Turkey Point 1989 Plantation White Pine (CA-TP2). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246010
M. Altaf Arain (2025). Ontario - Turkey Point 1974 Plantation White Pine (CA-TP3). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246011
M. Altaf Arain (2025). Ontario - Turkey Point 1939 Plantation White Pine (CA-TP4). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246012
M. Altaf Arain (2025). Ontario - Turkey Point Mature Deciduous (CA-TPD). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246152
Juan J. Armesto (2025). Senda Darwin Forest (CL-SDF). AmeriFlux AMP. https://doi.org/10.17190/AMF/1902273
Jorge Perez-Quezada (2025). Senda Darwin Peatland (CL-SDP). AmeriFlux AMP. https://doi.org/10.17190/AMF/1902274
Mark Johnson (2025). Filadelfia sugar cane cropland (CR-Fsc). AmeriFlux AMP. https://doi.org/10.17190/AMF/1829504
National Ecological Observatory Network (2025). NEON Guanica Forest (GUAN) (PR-xGU). AmeriFlux AMP. https://doi.org/10.17190/AMF/1773393
National Ecological Observatory Network (2025). NEON Lajas Experimental Station (LAJA) (PR-xLA). AmeriFlux AMP. https://doi.org/10.17190/AMF/1773394
Ryan Sullivan (2025). ARM-AMF3-Oliktok (US-A03). AmeriFlux AMP. https://doi.org/10.17190/AMF/1498752
Ryan Sullivan (2025). ARM-NSA-Barrow (US-A10). AmeriFlux AMP. https://doi.org/10.17190/AMF/1498753
Sebastien Biraud (2025). ARM-SGP Medford hay pasture (US-A32). AmeriFlux AMP. https://doi.org/10.17190/AMF/1436327
Ryan Sullivan (2025). ARM-SGP-Waukomis (US-A37). AmeriFlux AMP. https://doi.org/10.17190/AMF/2531140
Sebastien Biraud (2025). ARM SGP milo field (US-A74). AmeriFlux AMP. https://doi.org/10.17190/AMF/1436328
Michael Moreo (2025). Amargosa Desert Research Site (ADRS) (US-ADR). AmeriFlux AMP. https://doi.org/10.17190/AMF/1418680
Roser Matamala (2025). Argonne Testbed for Multiscale Observational Science (ATMOS) (US-AMS). AmeriFlux AMP. https://doi.org/10.17190/AMF/2531142
Margaret Torn (2025). ARM USDA UNL OSU Woodward Switchgrass 1 (US-AR1). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246137
Margaret Torn (2025). ARM USDA UNL OSU Woodward Switchgrass 2 (US-AR2). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246138
Sebastien Biraud (2025). ARM Southern Great Plains site- Lamont (US-ARM). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246027
Andrew Richardson (2025). Bartlett Experimental Forest (US-Bar). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246030
Dennis Baldocchi (2025). Bouldin Island Alfalfa (US-Bi1). AmeriFlux AMP. https://doi.org/10.17190/AMF/1480317
Dennis Baldocchi (2025). Bouldin Island corn (US-Bi2). AmeriFlux AMP. https://doi.org/10.17190/AMF/1419513
Kimberly Novick (2025). Bayles Road Grassland Tower (US-BRG). AmeriFlux AMP. https://doi.org/10.17190/AMF/1756416
Eugenie Euskirchen (2025). Bonanza Creek Thermokarst Bog (US-BZB). AmeriFlux AMP. https://doi.org/10.17190/AMF/1773401
Eugenie Euskirchen (2025). Bonanza Creek Rich Fen (US-BZF). AmeriFlux AMP. https://doi.org/10.17190/AMF/1756433
Eugenie Euskirchen (2025). Bonanza Creek Old Thermokarst Bog (US-BZo). AmeriFlux AMP. https://doi.org/10.17190/AMF/1846662
Eugenie Euskirchen (2025). Bonanza Creek Black Spruce (US-BZS). AmeriFlux AMP. https://doi.org/10.17190/AMF/1756434
Steve Kannenberg (2025). Cedar Mesa (US-CdM). AmeriFlux AMP. https://doi.org/10.17190/AMF/1865477
Dave Huggins (2025). CAF-LTAR Cook East (US-CF1). AmeriFlux AMP. https://doi.org/10.17190/AMF/1543382
Dave Huggins (2025). CAF-LTAR Cook West (US-CF2). AmeriFlux AMP. https://doi.org/10.17190/AMF/1543383
Dave Huggins (2025). CAF-LTAR Boyd North (US-CF3). AmeriFlux AMP. https://doi.org/10.17190/AMF/1543385
Dave Huggins (2025). CAF-LTAR Boyd South (US-CF4). AmeriFlux AMP. https://doi.org/10.17190/AMF/1543384
Patty Oikawa (2025). Concord Grazed Grassland (US-CGG). AmeriFlux AMP. https://doi.org/10.17190/AMF/1987600
Kenneth J. Davis (2025). Cole Farm (US-CLF). AmeriFlux AMP. https://doi.org/10.17190/AMF/1987601
Russell Scott (2025). Charleston Mesquite Woodland (US-CMW). AmeriFlux AMP. https://doi.org/10.17190/AMF/1660339
David Bowling (2025). Corral Pocket (US-Cop). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246129
Asko Noormets (2025). Davy Crockett National Forest (US-CRK). AmeriFlux AMP. https://doi.org/10.17190/AMF/2204055
Jiquan Chen (2025). Curtice Walter-Berger cropland (US-CRT). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246156
Ankur Desai (2025). Tri county school Pine Forest (US-CS2). AmeriFlux AMP. https://doi.org/10.17190/AMF/1617711
Ankur Desai (2025). Central Sands Irrigated Agricultural Field (US-CS5). AmeriFlux AMP. https://doi.org/10.17190/AMF/1846663
Ankur Desai (2025). Central Sands Irrigated Agricultural Field (US-CS6). AmeriFlux AMP. https://doi.org/10.17190/AMF/2001297
Kim Novick (2025). Crossett Experimental Forest (US-Cst). AmeriFlux AMP. https://doi.org/10.17190/AMF/1902275
Ankur Desai (2025). US Dairy Forage Research Center, Prairie du Sac (US-DFC). AmeriFlux AMP. https://doi.org/10.17190/AMF/1660340
Valentin Picasso Risso (2025). Dairy Forage Research Center - Kernza (US-DFK). AmeriFlux AMP. https://doi.org/10.17190/AMF/1825937
Paul Stoy (2025). Duke Forest-hardwoods (US-Dk2). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246047
Paul Stoy (2025). Duke Forest - loblolly pine (US-Dk3). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246048
Nick Christen (2025). Staten Corn 1 (US-DS1). AmeriFlux AMP. https://doi.org/10.17190/AMF/2407202
Steve Deverel (2025). Staten Corn 2 (US-DS2). AmeriFlux AMP. https://doi.org/10.17190/AMF/2407203
Steve Deverel (2025). Staten Rice 1 (US-DS3). AmeriFlux AMP. https://doi.org/10.17190/AMF/1890490
Preston Fleck (2025). EAA Field Research Park Woodland (US-EA4). AmeriFlux AMP. https://doi.org/10.17190/AMF/2315767
Marcus Gary (2025). Uvalde Ranch Mesquite Woodland (US-EA5). AmeriFlux AMP. https://doi.org/10.17190/AMF/2204056
Marcus Gary (2025). Camp Wood Shield Ranch Oak Savannah (US-EA6). AmeriFlux AMP. https://doi.org/10.17190/AMF/2315768
Steve Oberbauer (2025). Everglades (long hydroperiod marsh) (US-Elm). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246118
Ted Schuur (2025). Eight Mile Lake Permafrost thaw gradient, Healy Alaska. (US-EML). AmeriFlux AMP. https://doi.org/10.17190/AMF/1418678
Steve Oberbauer (2025). Everglades (short hydroperiod marsh) (US-Esm). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246119
Yoshinobu Harazono (2025). Cascaden Ridge Fire Scar (US-Fcr). AmeriFlux AMP. https://doi.org/10.17190/AMF/1562388
Thomas Kolb (2025). Flagstaff - Managed Forest (US-Fmf). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246050
Praveena Krishnan (2025). Flux Observations of Carbon from an Airborne Laboratory (FOCAL) Campaign Site 2 (US-Fo2). AmeriFlux AMP. https://doi.org/10.17190/AMF/2531146
Thomas Kolb (2025). Flagstaff - Unmanaged Forest (US-Fuf). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246051
Thomas Kolb (2025). Flagstaff - Wildfire (US-Fwf). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246052
Rob Hubbard (2025). GLEES (US-GLE). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246056
J. William Munger (2025). Harvard Forest EMS Tower (HFR1) (US-Ha1). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246059
J. William Munger (2025). Harvard Forest Hemlock Site (US-Ha2). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246060
Tom O'Halloran (2025). Hobcaw Barony Mature Longleaf Pine (US-HB2). AmeriFlux AMP. https://doi.org/10.17190/AMF/1660342
Tom O'Halloran (2025). Hobcaw Barony Longleaf Pine Restoration (US-HB3). AmeriFlux AMP. https://doi.org/10.17190/AMF/1660343
Xingyuan Chen (2025). Hanford 100H grassland (US-Hn2). AmeriFlux AMP. https://doi.org/10.17190/AMF/1562389
Andrew Ouimette (2025). Howland Forest (main tower) (US-Ho1). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246061
Andrew Ouimette (2025). Howland Forest (west tower) (US-Ho2). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246062
Andrew Ouimette (2025). Howland Forest (harvest site) (US-Ho3). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246063
Benjamin R. K. Runkle (2025). Humnoke Farm Rice Field ‚Äì Field A (US-HRA). AmeriFlux AMP. https://doi.org/10.17190/AMF/1543376
Michele L. Reba (2025). Humnoke Farm Rice Field ‚Äì Field C (US-HRC). AmeriFlux AMP. https://doi.org/10.17190/AMF/1543375
Dennis Baldocchi (2025). Hill Slough Marsh (US-Hsm). AmeriFlux AMP. https://doi.org/10.17190/AMF/1890483
Sarah Goslee (2025). USDA ARS Pasture Sytems and Watershed Management Research Unit- Hawbecker Site (US-HWB). AmeriFlux AMP. https://doi.org/10.17190/AMF/1811363
Roser Matamala (2025). Fermi National Accelerator Laboratory- Batavia (Agricultural site) (US-IB1). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246065
Roser Matamala (2025). Fermi National Accelerator Laboratory- Batavia (Prairie site) (US-IB2). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246066
Syndonia Bret-Harte (2025). Imnavait Creek Watershed Heath Tundra (US-ICh). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246133
Syndonia Bret-Harte (2025). Imnavait Creek Watershed Wet Sedge Tundra (US-ICs). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246130
Syndonia Bret-Harte (2025). Imnavait Creek Watershed Tussock Tundra (US-ICt). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246131
Craig Tweedie (2025). Jornada Experimental Range Bajada Site (US-Jo1). AmeriFlux AMP. https://doi.org/10.17190/AMF/1767833
Enrique R. Vivoni (2025). Jornada Experimental Range Mixed Shrubland (US-Jo2). AmeriFlux AMP. https://doi.org/10.17190/AMF/1617696
Nathaniel Brunsell (2025). Kansas Field Station (US-KFS). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246132
Nathaniel Brunsell (2025). Kansas Land Institute (US-KLS). AmeriFlux AMP. https://doi.org/10.17190/AMF/1498745
Nathaniel Brunsell (2025). Konza Prairie LTER (KNZ) (US-Kon). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246068
Silvano Fares (2025). Lindcove Orange Orchard (US-Lin). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246830
Gregory Starr (2025). Longleaf Pine - Baker (Mesic site) (US-LL1). AmeriFlux AMP. https://doi.org/10.17190/AMF/1773395
Gregory Starr (2025). Longleaf Pine - Dubignion (Intermediate site) (US-LL2). AmeriFlux AMP. https://doi.org/10.17190/AMF/1773396
Gregory Starr (2025). Longleaf Pine - Red Dirt (Xeric site) (US-LL3). AmeriFlux AMP. https://doi.org/10.17190/AMF/1773397
Russell Scott (2025). San Pedro River Lewis Springs Sacaton Grassland (US-LS1). AmeriFlux AMP. https://doi.org/10.17190/AMF/1660346
Russell Scott (2025). San Pedro River Lewis Springs Savanna (US-LS2). AmeriFlux AMP. https://doi.org/10.17190/AMF/1660347
Chris Still (2025). Metolius mature ponderosa pine (US-Me2). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246076
Bev Law (2025). Metolius-second young aged pine (US-Me3). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246077
Bev Law (2025). Metolius-first young aged pine (US-Me5). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246079
Chris Still (2025). Metolius Young Pine Burn (US-Me6). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246128
Rich Phillips (2025). Morgan Monroe State Forest (US-MMS). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246080
Sharon Weyers (2025). Morris: Corn-Soybean with Cover Crops, Strip Tillage (US-MN1). AmeriFlux AMP. https://doi.org/10.17190/AMF/2407204
Sharon Weyers (2025). Morris: Corn-Soybean-Wheat with Cover Crops, Minimal Tillage (US-MN2). AmeriFlux AMP. https://doi.org/10.17190/AMF/2407205
Sharon Weyers (2025). Morris: Corn-Soybean, Conventional Tillage (US-MN3). AmeriFlux AMP. https://doi.org/10.17190/AMF/2407206
Lianhong Gu (2025). Missouri Ozark Site (US-MOz). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246081
Ankur Desai (2025). NW2 Aspen-1 CHEESEHEAD 2019 (US-PFc). AmeriFlux AMP. https://doi.org/10.17190/AMF/1717851
Ankur Desai (2025). NW3 Tussock-1 CHEESEHEAD 2019 (US-PFd). AmeriFlux AMP. https://doi.org/10.17190/AMF/1717852
Ankur Desai (2025). NW4 Lake-1 CHEESEHEAD 2019 (US-PFe). AmeriFlux AMP. https://doi.org/10.17190/AMF/1717853
Ankur Desai (2025). NE1 Pine-2 CHEESEHEAD 2019 (US-PFg). AmeriFlux AMP. https://doi.org/10.17190/AMF/1717854
Ankur Desai (2025). NE2 Pine-3 CHEESEHEAD 2019 (US-PFh). AmeriFlux AMP. https://doi.org/10.17190/AMF/1717855
Ankur Desai (2025). NE3 Hardwood-1 CHEESEHEAD 2019 (US-PFi). AmeriFlux AMP. https://doi.org/10.17190/AMF/1717856
Ankur Desai (2025). NE4 Maple-1 CHEESEHEAD 2019 (US-PFj). AmeriFlux AMP. https://doi.org/10.17190/AMF/1717857
Ankur Desai (2025). SW1 Aspen-2 CHEESEHEAD 2019 (US-PFk). AmeriFlux AMP. https://doi.org/10.17190/AMF/1717858
Ankur Desai (2025). SW3 Hardwood-2 CHEESEHEAD 2019 (US-PFm). AmeriFlux AMP. https://doi.org/10.17190/AMF/1717860
Ankur Desai (2025). SW4 Hardwood-3 CHEESEHEAD 2019 (US-PFn). AmeriFlux AMP. https://doi.org/10.17190/AMF/1717861
Ankur Desai (2025). SE2 Hardwood-4 CHEESEHEAD 2019 (US-PFp). AmeriFlux AMP. https://doi.org/10.17190/AMF/1717862
Ankur Desai (2025). SE3 Aspen-4 CHEESEHEAD 2019 (US-PFq). AmeriFlux AMP. https://doi.org/10.17190/AMF/1717863
Ankur Desai (2025). SE4 Tussock-2 CHEESEHEAD 2019 (US-PFr). AmeriFlux AMP. https://doi.org/10.17190/AMF/1717864
Ankur Desai (2025). SE5 Aspen-5 CHEESEHEAD 2019 (US-PFs). AmeriFlux AMP. https://doi.org/10.17190/AMF/1717865
Ankur Desai (2025). SE6 Pine-4 CHEESEHEAD 2019 (US-PFt). AmeriFlux AMP. https://doi.org/10.17190/AMF/1717866
Hiroki Ikawa (2025). Poker Flat Research Range Black Spruce Forest (US-Prr). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246153
Sebastien C. Biraud (2025). Arkansas Corn Farm (US-RGA). AmeriFlux AMP. https://doi.org/10.17190/AMF/1880913
Sebastien C. Biraud (2025). Butte County Rice Farm (US-RGB). AmeriFlux AMP. https://doi.org/10.17190/AMF/1870591
Sebastien C. Biraud (2025). Stanislaus County Forage Farm (US-RGF). AmeriFlux AMP. https://doi.org/10.17190/AMF/2001310
Sebastien C. Biraud (2025). Glenn County Organic Rice Farm (US-RGo). AmeriFlux AMP. https://doi.org/10.17190/AMF/1880914
Sebastien C. Biraud (2025). Desha County Rice Farm (US-RGW). AmeriFlux AMP. https://doi.org/10.17190/AMF/1880915
Gerald Flerchinger (2025). RCEW Low Sagebrush (US-Rls). AmeriFlux AMP. https://doi.org/10.17190/AMF/1418682
Gerald Flerchinger (2025). RCEW Mountain Big Sagebrush (US-Rms). AmeriFlux AMP. https://doi.org/10.17190/AMF/1375202
Timothy Griffis (2025). Rosemount- G21 (US-Ro1). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246092
Timothy Griffis (2025). Rosemount Prairie (US-Ro4). AmeriFlux AMP. https://doi.org/10.17190/AMF/1419507
Timothy Griffis (2025). Rosemount I18_South (US-Ro5). AmeriFlux AMP. https://doi.org/10.17190/AMF/1419508
Timothy Griffis (2025). Rosemount I18_North (US-Ro6). AmeriFlux AMP. https://doi.org/10.17190/AMF/1419509
Yoshinobu Harazono (2025). Poker Flat Research Range: Succession from fire scar to deciduous forest (US-Rpf). AmeriFlux AMP. https://doi.org/10.17190/AMF/1579540
Scott Neubauer (2025). Rice Rivers Center Marsh (US-RRC). AmeriFlux AMP. https://doi.org/10.17190/AMF/2331382
Michele L. Reba (2025). RCEW Reynolds Mountain East (US-Rwe). AmeriFlux AMP. https://doi.org/10.17190/AMF/1617721
Gerald Flerchinger (2025). RCEW Upper Sheep Prescibed Fire (US-Rwf). AmeriFlux AMP. https://doi.org/10.17190/AMF/1617724
Gerald Flerchinger (2025). Reynolds Creek Wyoming big sagebrush (US-Rws). AmeriFlux AMP. https://doi.org/10.17190/AMF/1375201
Dennis Baldocchi (2025). Sherman Barn (US-Snf). AmeriFlux AMP. https://doi.org/10.17190/AMF/1579718
Shirley Papuga (2025). Santa Rita Creosote (US-SRC). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246127
Russell Scott (2025). Santa Rita Grassland (US-SRG). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246154
Russell Scott (2025). Santa Rita Mesquite (US-SRM). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246104
Enrique R. Vivoni (2025). Santa Rita Savanna (US-SRS). AmeriFlux AMP. https://doi.org/10.17190/AMF/1660351
Kenneth J. Davis (2025). Susquehanna Shale Hills Critical Zone Observatory (US-SSH). AmeriFlux AMP. https://doi.org/10.17190/AMF/1880916
Elise Pendall (2025). Saratoga (US-Sta). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246831
Ankur Desai (2025). Sylvania Wilderness Area (US-Syv). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246106
Camilo Rey-Sanchez (2025). Timberlake Observatory for Wetland Restoration (TOWeR) (US-TLR). AmeriFlux AMP. https://doi.org/10.17190/AMF/2531147
Dennis Baldocchi (2025). Tonzi Ranch (US-Ton). AmeriFlux AMP. https://doi.org/10.17190/AMF/1245971
Dennis Baldocchi (2025). Twitchell Corn (US-Tw2). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246148
Dennis Baldocchi (2025). Twitchell Alfalfa (US-Tw3). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246149
Yoshinobu Harazono (2025). University of Alaska, Fairbanks (US-Uaf). AmeriFlux AMP. https://doi.org/10.17190/AMF/1480322
Sarah Goslee (2025). LTAR UCB (Upper Chesapeake Bay) EC1 (US-UC1). AmeriFlux AMP. https://doi.org/10.17190/AMF/1865482
Sarah Goslee (2025). LTAR UCB (Upper Chesapeake Bay) EC2 (US-UC2). AmeriFlux AMP. https://doi.org/10.17190/AMF/1865483
Carl J Bernacchi (2025). University of Illinois Switchgrass (US-UiA). AmeriFlux AMP. https://doi.org/10.17190/AMF/1617725
Carl J Bernacchi (2025). University of Illinois Maize-Soy (US-UiC). AmeriFlux AMP. https://doi.org/10.17190/AMF/1846665
Carl J Bernacchi (2025). University of Illinois Restored Native Prairie (US-UiD). AmeriFlux AMP. https://doi.org/10.17190/AMF/1987605
Gil Bohrer (2025). Univ. of Mich. Biological Station (US-UMB). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246107
Gil Bohrer (2025). UMBS Disturbance (US-UMd). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246134
Paul Inkenbrandt (2025). UFLUX Dugout Ranch (US-UTD). AmeriFlux AMP. https://doi.org/10.17190/AMF/2531148
Paul Inkenbrandt (2025). UFLUX Escalante (US-UTE). AmeriFlux AMP. https://doi.org/10.17190/AMF/2531149
Paul Inkenbrandt (2025). UFLUX Bluff (US-UTJ). AmeriFlux AMP. https://doi.org/10.17190/AMF/2531150
Paul Inkenbrandt (2025). UFLUX Desert View (US-UTV). AmeriFlux AMP. https://doi.org/10.17190/AMF/2531152
Paul Inkenbrandt (2025). UFLUX Wellington (US-UTW). AmeriFlux AMP. https://doi.org/10.17190/AMF/2507522
Dennis Baldocchi (2025). Vaira Ranch- Ione (US-Var). AmeriFlux AMP. https://doi.org/10.17190/AMF/1245984
Ankur Desai (2025). Willow Creek (US-WCr). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246111
Russell Scott (2025). Walnut Gulch Lucky Hills Shrub (US-Whs). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246113
Russell Scott (2025). Walnut Gulch Kendall Grasslands (US-Wkg). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246112
Richard L. Coulter (2025). Walnut River Watershed (Smileyburg) (US-Wlr). AmeriFlux AMP. https://doi.org/10.17190/AMF/1246115
National Ecological Observatory Network (2025). NEON Abby Road (ABBY) (US-xAB). AmeriFlux AMP. https://doi.org/10.17190/AMF/1617726
National Ecological Observatory Network (2025). NEON Klemme Range Research Station (OAES) (US-xAE). AmeriFlux AMP. https://doi.org/10.17190/AMF/1671891
National Ecological Observatory Network (2025). NEON Barrow Environmental Observatory (BARR) (US-xBA). AmeriFlux AMP. https://doi.org/10.17190/AMF/1671892
National Ecological Observatory Network (2025). NEON Blandy Experimental Farm (BLAN) (US-xBL). AmeriFlux AMP. https://doi.org/10.17190/AMF/1671893
National Ecological Observatory Network (2025). NEON Caribou Creek - Poker Flats Watershed (BONA) (US-xBN). AmeriFlux AMP. https://doi.org/10.17190/AMF/1617727
National Ecological Observatory Network (2025). NEON Bartlett Experimental Forest (BART) (US-xBR). AmeriFlux AMP. https://doi.org/10.17190/AMF/1579542
National Ecological Observatory Network (2025). NEON LBJ National Grassland (CLBJ) (US-xCL). AmeriFlux AMP. https://doi.org/10.17190/AMF/1671894
National Ecological Observatory Network (2025). NEON Central Plains Experimental Range (CPER) (US-xCP). AmeriFlux AMP. https://doi.org/10.17190/AMF/1579720
National Ecological Observatory Network (2025). NEON Dakota Coteau Field School (DCFS) (US-xDC). AmeriFlux AMP. https://doi.org/10.17190/AMF/1617728
National Ecological Observatory Network (2025). NEON Delta Junction (DEJU) (US-xDJ). AmeriFlux AMP. https://doi.org/10.17190/AMF/1634884
National Ecological Observatory Network (2025). NEON Dead Lake (DELA) (US-xDL). AmeriFlux AMP. https://doi.org/10.17190/AMF/1579721
National Ecological Observatory Network (2025). NEON Disney Wilderness Preserve (DSNY) (US-xDS). AmeriFlux AMP. https://doi.org/10.17190/AMF/1671895
National Ecological Observatory Network (2025). NEON Great Smoky Mountains National Park, Twin Creeks (GRSM) (US-xGR). AmeriFlux AMP. https://doi.org/10.17190/AMF/1634885
National Ecological Observatory Network (2025). NEON Harvard Forest (HARV) (US-xHA). AmeriFlux AMP. https://doi.org/10.17190/AMF/1562391
National Ecological Observatory Network (2025). NEON Healy (HEAL) (US-xHE). AmeriFlux AMP. https://doi.org/10.17190/AMF/1617729
National Ecological Observatory Network (2025). NEON Jones Ecological Research Center (JERC) (US-xJE). AmeriFlux AMP. https://doi.org/10.17190/AMF/1617730
National Ecological Observatory Network (2025). NEON Jornada LTER (JORN) (US-xJR). AmeriFlux AMP. https://doi.org/10.17190/AMF/1617731
National Ecological Observatory Network (2025). NEON Konza Prairie Biological Station - Relocatable (KONA) (US-xKA). AmeriFlux AMP. https://doi.org/10.17190/AMF/1579722
National Ecological Observatory Network (2025). NEON Konza Prairie Biological Station (KONZ) (US-xKZ). AmeriFlux AMP. https://doi.org/10.17190/AMF/1562392
National Ecological Observatory Network (2025). NEON Lenoir Landing (LENO) (US-xLE). AmeriFlux AMP. https://doi.org/10.17190/AMF/1773398
National Ecological Observatory Network (2025). NEON Moab (MOAB) (US-xMB). AmeriFlux AMP. https://doi.org/10.17190/AMF/1671896
National Ecological Observatory Network (2025). NEON Mountain Lake Biological Station (MLBS) (US-xML). AmeriFlux AMP. https://doi.org/10.17190/AMF/1671897
National Ecological Observatory Network (2025). NEON Northern Great Plains Research Laboratory (NOGP) (US-xNG). AmeriFlux AMP. https://doi.org/10.17190/AMF/1617732
National Ecological Observatory Network (2025). NEON Onaqui-Ault (ONAQ) (US-xNQ). AmeriFlux AMP. https://doi.org/10.17190/AMF/1617733
National Ecological Observatory Network (2025). NEON Niwot Ridge Mountain Research Station (NIWO) (US-xNW). AmeriFlux AMP. https://doi.org/10.17190/AMF/1671898
National Ecological Observatory Network (2025). NEON Rocky Mountain National Park, CASTNET (RMNP) (US-xRM). AmeriFlux AMP. https://doi.org/10.17190/AMF/1579723
National Ecological Observatory Network (2025). NEON Oak Ridge National Lab (ORNL) (US-xRN). AmeriFlux AMP. https://doi.org/10.17190/AMF/1773400
National Ecological Observatory Network (2025). NEON Ordway-Swisher Biological Station (OSBS) (US-xSB). AmeriFlux AMP. https://doi.org/10.17190/AMF/1671899
National Ecological Observatory Network (2025). NEON Smithsonian Conservation Biology Institute (SCBI) (US-xSC). AmeriFlux AMP. https://doi.org/10.17190/AMF/1671900
National Ecological Observatory Network (2025). NEON Smithsonian Environmental Research Center (SERC) (US-xSE). AmeriFlux AMP. https://doi.org/10.17190/AMF/1617734
National Ecological Observatory Network (2025). NEON San Joaquin Experimental Range (SJER) (US-xSJ). AmeriFlux AMP. https://doi.org/10.17190/AMF/1671901
National Ecological Observatory Network (2025). NEON North Sterling, CO (STER) (US-xSL). AmeriFlux AMP. https://doi.org/10.17190/AMF/1617735
National Ecological Observatory Network (2025). NEON Soaproot Saddle (SOAP) (US-xSP). AmeriFlux AMP. https://doi.org/10.17190/AMF/1617736
National Ecological Observatory Network (2025). NEON Santa Rita Experimental Range (SRER) (US-xSR). AmeriFlux AMP. https://doi.org/10.17190/AMF/1579543
National Ecological Observatory Network (2025). NEON Steigerwaldt Land Services (STEI) (US-xST). AmeriFlux AMP. https://doi.org/10.17190/AMF/1617737
National Ecological Observatory Network (2025). NEON Talladega National Forest (TALL) (US-xTA). AmeriFlux AMP. https://doi.org/10.17190/AMF/1671902
National Ecological Observatory Network (2025). NEON Lower Teakettle (TEAK) (US-xTE). AmeriFlux AMP. https://doi.org/10.17190/AMF/1617738
National Ecological Observatory Network (2025). NEON Toolik (TOOL) (US-xTL). AmeriFlux AMP. https://doi.org/10.17190/AMF/1617739
National Ecological Observatory Network (2025). NEON Treehaven (TREE) (US-xTR). AmeriFlux AMP. https://doi.org/10.17190/AMF/1634886
National Ecological Observatory Network (2025). NEON The University of Kansas Field Station (UKFS) (US-xUK). AmeriFlux AMP. https://doi.org/10.17190/AMF/1617740
National Ecological Observatory Network (2025). NEON University of Notre Dame Environmental Research Center (UNDE) (US-xUN). AmeriFlux AMP. https://doi.org/10.17190/AMF/1617741
National Ecological Observatory Network (2025). NEON Woodworth (WOOD) (US-xWD). AmeriFlux AMP. https://doi.org/10.17190/AMF/1579724
National Ecological Observatory Network (2025). NEON Wind River Experimental Forest (WREF) (US-xWR). AmeriFlux AMP. https://doi.org/10.17190/AMF/1617742
National Ecological Observatory Network (2025). NEON Yellowstone Northern Range (Frog Rock) (YELL) (US-xYE). AmeriFlux AMP. https://doi.org/10.17190/AMF/1617743
Susan Natali (2025). Yukon-Kuskokwim Delta, Izaviknek-Kingaglia uplands, Burned 2015 (US-YK1). AmeriFlux AMP. https://doi.org/10.17190/AMF/2331384
Sue Natali (2025). Yukon-Kuskokwim Delta, Izaviknek-Kingaglia uplands, Unburned (US-YK2). AmeriFlux AMP. https://doi.org/10.17190/AMF/2331385

Acknowledgments

We thank the AmeriFlux network and individual site PIs for making their data publicly available. Soil property data were obtained from ISRIC SoilGrids. Hydraulic parameter estimation used the Rosetta pedotransfer functions.

Contact

For questions about this dataset or analysis pipeline, please contact:

Yulissa T. Perez Rojas: yperezrojas@ucmerced.edu
Teamrat A. Ghezzehei: taghezzehei@ucmerced.edu

Last Updated: May 2026
Repository Version: 1.3 (Dryad: processedData.zip CC0; Zenodo Software: scripts MIT; Zenodo Supplemental: data_zenodo.zip CC BY 4.0)

Data from: Jensen's inequality reveals how moisture variability controls soil respiration across continental scales

Data files

Abstract

README: Data from: Jensen's inequality reveals how moisture variability controls soil respiration across continental scales

Publication Information

Repository Structure

Dryad: processedData.zip (CC0)

Zenodo (Software, MIT): individual scripts

Zenodo (Supplemental, CC BY 4.0): data_zenodo.zip

Not redistributed

Assembling a working directory

Data Availability

Raw AmeriFlux Data

Data Provenance and Third-Party Licensing

SoilGrids Data

Processed Data

Software Requirements

R Environment

Python Environment

Quick Start Guide

Option A: Reproduce Visualizations and Analysis (No Raw Data Needed)

Option B: Run Complete Pipeline from Raw Data

Analysis Pipeline

Pipeline Overview

Pipeline Sequence

Detailed Script Documentation

Script 1: Data Preprocessing - 1_preProcessData.R

Script 2: Site Characteristics - 2_siteCharacteristics.R

Script 3: Soil Properties from SoilGrids - 3_soilGrids.R

Script 3b: Hydraulic Parameters - 3b_Rosetta.py

Script 4: Soil Physical Properties - 4_soilPhysicalProperties.R

Script 5: Respiration Calculations - 5_calcMoistureRespiration.R

Script 6: Respiration Visualization - 6_plotRespiration__FINAL.R

Script 7: Probability Analysis - 7_probabilityAnalysis_V2.R

Script 8: Site-Level Analysis - 8_generate_site_level_table.R

Critical Intermediate Data Files

Computational Resources

Recommended Hardware

Parallel Processing

Troubleshooting

Key Findings

Citation

AmeriFlux Site References (242 Sites)

Acknowledgments

Contact

Dryad: `processedData.zip` (CC0)

Zenodo (Supplemental, CC BY 4.0): `data_zenodo.zip`

Script 1: Data Preprocessing - `1_preProcessData.R`

Script 2: Site Characteristics - `2_siteCharacteristics.R`

Script 3: Soil Properties from SoilGrids - `3_soilGrids.R`

Script 3b: Hydraulic Parameters - `3b_Rosetta.py`

Script 4: Soil Physical Properties - `4_soilPhysicalProperties.R`

Script 5: Respiration Calculations - `5_calcMoistureRespiration.R`

Script 6: Respiration Visualization - `6_plotRespiration__FINAL.R`

Script 7: Probability Analysis - `7_probabilityAnalysis_V2.R`

Script 8: Site-Level Analysis - `8_generate_site_level_table.R`