Agroecosystem Datasets
Quality-controlled, harmonized soil and management datasets documented for discovery and reuse.
Explore Soil Carbon and Greenhouse Gas Measurements
Each point marks a field location with compiled measurements across five core variables: soil carbon (C), soil nitrogen (N), CO₂, CH₄, and N₂O, colored by how many variables are present. Clustered pins indicate multiple land uses at the same site; zoom in to separate them. Hover for a quick preview, or click to pin the full data window.
Loading map…
Reading the Site Summary
Clicking any point opens a window detailing measurements at that location. A green check (✓) confirms data availability, while soil carbon encompasses total, organic, inorganic, and fractionated pools (POM and MAOM). Where data exist, the table summarizes collection parameters:
- Sites
- The number of distinct field sites or experiment installations at this location.
- Treatments
- The number of distinct management practices compared (e.g., tillage systems, nitrogen rates, or crop rotations).
- Plots
- The count of individual field plots or replicates. A dash (–) indicates the dataset reports treatment-level averages rather than individual plot data.
- Dates
- The number of distinct sampling or measurement dates recorded.
- Depth & Layers (cm)
- The sampled depth range and, in brackets, the maximum number of soil profile layers sampled. Trace gas measurements do not have depth.
- Samples
- One distinct measurement from one plot, on one date, at one depth (or one treatment average). The count reflects actual measurements taken, not every theoretical combination of plots, dates, and depths.
Where multiple datasets contribute data on the same variable, indented rows (↳) detail samples by source.
Interpretation Notes
- Sampling intensity vs. study duration: Sample counts reflect study design. High-frequency gas monitoring or intensive depth-slicing generates far more sample rows than an annual surface core over the same number of years.
- Differing levels of aggregation: Datasets vary in reporting precision, ranging from individual plot replicates to published treatment means or single site-level averages.
- Overlapping compilations: Landmark or long-term agricultural experiments may appear in more than one compiled dataset, meaning some sample counts across datasets describe the same underlying trials.
- Interpreting dashes (–): A dash indicates that a field does not apply (such as depth layers for greenhouse gas fluxes) or was not reported by the original research team.
Data Products
High-quality datasets for ecosystem and soil carbon science, each with metadata, citations, and download access for consortium members.
- Global Time-Series Soil Carbon for DayCent® and MEMSCurated collection49 sites · 7,423 observations
- GRACEnet57 sites · 517,727 observations
- Global N₂O Database141 sites · 555,502 observations
- Global Soil Carbon Fractions: Regenerative + Conventional Croplands163 sites · 1,250 observations
- Global Soil Carbon Fractions: Managed and Unmanaged Ecosystems808 sites · 2,998 observations
- Global SOC: Grassland Management284 sites · 665 observations
External Databases
Beyond our own data products, we have compiled more than 30 external databases and repositories spanning soil, carbon, and environmental science. From global soil profile archives to flux networks and national observatories, these resources are gathered in one place to help you find the data that supports your work.
Explore DatabasesReady to explore models?
Explore DayCent & MEMS Models
