Sentinel-2 NDVI for Regen Operators: The Free 10-Metre Monitoring Layer Every Farm Already Has Access To

The Satellite That Changed the Arithmetic

The Sentinel-2 constellation images every point on Earth between 56 degrees south and 83 degrees north every five days. Two satellites, Sentinel-2A launched in June 2015 and Sentinel-2B in March 2017, carry identical Multi-Spectral Instruments producing Level-2A bottom-of-atmosphere surface reflectance at 10-metre resolution in the visible and near-infrared bands (ESA Copernicus Programme, 2015). The atmospheric correction is applied server-side. What arrives in the operator's GIS software is calibrated surface reflectance ready for index calculation, not raw radiance requiring specialist processing.

The instrument records 13 spectral bands. The four bands relevant to standard regen monitoring are Band 4 (Red, 665 nanometres), Band 8 (Near-Infrared, 842 nanometres), Band 8A (Red-Edge, 865 nanometres), and Bands 11 and 12 (Short-Wave Infrared at 1,610 and 2,190 nanometres). The visible and NIR bands resolve at 10 metres; SWIR runs at 20 metres but resamples cleanly to 10 metres for farm-scale workflows. The Level-2A product is the standard download for regen operators: it incorporates the Sen2Cor atmospheric correction algorithm and delivers imagery ready for NDVI, EVI, and NDMI computation without additional preprocessing (ESA Sentinel-2 Level-2A Algorithm Theoretical Basis Document, 2023).

The cost is zero. Before this infrastructure existed, 10-metre-resolution multispectral satellite imagery of comparable spectral quality required a commercial subscription at $50,000 or more per year for coverage over a single farm operation. The Copernicus programme, funded by the European Union and managed by the European Space Agency, was designed explicitly as open-access public infrastructure. Every scene ever acquired is archived and downloadable from the Copernicus Data Space Ecosystem under the Copernicus Open Licence, which imposes no restriction on commercial or non-commercial agricultural use. The satellite generates the data. The operator keeps it. There is no platform in the middle that owns the signal.

Three Indices, Four Regen Questions

NDVI, the Normalised Difference Vegetation Index, is computed from two bands: (NIR minus Red) divided by (NIR plus Red). The result scales from minus-one to plus-one. Bare soil runs 0.1 to 0.2. Sparse early-season cover sits around 0.2 to 0.35. Established cover crops with adequate biomass push above 0.4. Dense mid-season sward reaches 0.6 to 0.85. Tucker (1979) demonstrated the relationship between this ratio and photosynthetically active leaf area in his foundational work; the correlation holds across nearly every temperate crop and forage system within the 0.2 to 0.65 range that defines most regen monitoring windows.

EVI, the Enhanced Vegetation Index, corrects for soil-background noise and atmospheric aerosol effects by incorporating the blue band: 2.5 times (NIR minus Red), divided by (NIR plus 6 times Red, minus 7.5 times Blue, plus one). EVI degrades less in dense canopies where the NIR band begins to saturate, and performs more stably in high-residue no-till environments where soil reflectance confounds NDVI. The MODIS science team standardised EVI as part of their vegetation product suite from 2000 onwards; Sentinel-2's band set replicates it at 10-metre resolution rather than 250 metres. For regen operators running cover-cropped, high-residue systems in the spring establishment window, EVI often gives cleaner biomass estimates than NDVI at the transition moment before canopy closure.

NDMI, the Normalised Difference Moisture Index, uses the NIR and SWIR-1 bands: (NIR minus SWIR-1) divided by (NIR plus SWIR-1). Gao (1996) described the water-content sensitivity of this band combination in a foundational study of vegetation liquid water from space. In regen workflows, NDMI answers two distinct questions. First, residue cover: crop residue absorbs differently in the short-wave infrared than live vegetation, and NDMI values between minus 0.1 and positive 0.15 reliably indicate significant residue presence on the soil surface. Second, vegetation water stress: a sudden NDMI drop in July on a field that still shows healthy NDVI is typically the earliest measurable signal of drought stress, appearing five to ten days before NDVI itself begins to fall.

Four Regen Workflows in Practice

Cover crop establishment verification is the highest-value single use case for most regen operators planting into autumn. The workflow: download the November or early December Level-2A tile for the field, compute NDVI in QGIS or Google Earth Engine using the Band 8 and Band 4 layers, and compare against threshold. NDVI above 0.30 across the field polygon indicates adequate canopy closure for winter erosion control and nitrogen fixation targets. NDVI below 0.25 in patches signals thin establishment that may require intervention before spring termination. The USDA Natural Resources Conservation Service has begun accepting satellite-derived NDVI evidence as part of Conservation Stewardship Programme and Environmental Quality Incentives Programme practice verification in several states (USDA NRCS, 2024). An operator with a correctly documented NDVI record has a defensible, timestamped verification of establishment at zero cost per field.

Pasture biomass estimation is the second workflow. The correlation between NDVI and dry-matter biomass varies by forage species and region, but regression models calibrated to local conditions consistently achieve estimates within 15 to 20 percent of clipping samples across temperate grassland systems (Pullanagari et al., 2018, Remote Sensing journal, New Zealand dryland pastures). At 10-metre resolution and 5-day revisit, Sentinel-2 allows an operator running adaptive multi-paddock grazing to assess standing biomass across an entire rotation before making a shift decision. Gabe Brown of Brown's Ranch, Bismarck, North Dakota, has documented satellite monitoring as part of his adaptive grazing management toolkit alongside soil biology tests (Brown, "Dirt to Soil," 2018). The biomass map extends what the operator can see beyond any single field walk.

Residue cover quantification via NDMI supports the no-till and cover-crop-to-cash-crop transition verification increasingly required by carbon markets and conservation programmes. NDMI values derived from late-summer or early-autumn Sentinel-2 imagery, before cash-crop establishment, correlate with residue cover percentage as measured by the standard line-transect method (Biard and Baret, 1997, Remote Sensing of Environment). An operator seeking RCPP payment for residue management or seeking to document year-over-year soil cover improvement can include Sentinel-2 NDMI evidence alongside field-measured transect counts.

Mid-season anomaly detection closes the set. Time-series NDVI analysis in Google Earth Engine, comparing the current season's week-by-week composite against a three-year historical baseline for the same field polygon, surfaces yield-limiting events three to four weeks before they become visible to ground-level scouting. Drought stress, disease progression, and in-season nitrogen deficiencies all leave NDVI signatures detectable at 10-metre granularity that a 30-metre Landsat image would blend into background variation. The naturalist's observation and the engineer's index meet at the farm map: the pattern the operator noticed on a Tuesday walk shows up in the satellite composite from the previous Friday, timestamped and georeferenced.

Access Paths: QGIS, SentinelHub, Google Earth Engine

QGIS, free and fully open-source, is the maximum-control option. The Semi-Automatic Classification Plugin (Congedo, 2021) handles Sentinel-2 tile download, band compositing, and index calculation inside a single workflow. Operators new to raster algebra should budget four to six hours for initial setup and first map generation. After that, the workflow repeats in under 30 minutes. QGIS connects directly to the Copernicus Data Space Ecosystem for tile search and download; no third-party account is required. Cost: zero, permanently.

SentinelHub's EO Browser provides the fastest path to a visual NDVI map without software installation. Register for a free account at the Copernicus Data Space EO Browser portal, navigate to the farm location, select the Sentinel-2 Level-2A source, choose the NDVI visualisation script, and the result renders in-browser within seconds, with a date-picker for historical comparison. The free tier supports visual inspection and limited statistical downloads. For programmatic access at scale, SentinelHub's Statistical API and Batch Processing API run from approximately $25 per month for light commercial use (SentinelHub, Sinergise/Planet Labs, 2024). This is the right option for operators who want speed and do not need to maintain local GIS infrastructure.

Google Earth Engine holds every Sentinel-2 scene ever acquired in a cloud analysis environment. Access is free for research, education, and non-commercial use (Google Earth Engine, 2024); farm management use by operators and their advisers typically qualifies. A 10-line JavaScript snippet produces a multi-year NDVI time series for a field polygon, rendered as an interactive chart, accessible from any browser without downloading raw data locally. For the anomaly detection workflow, Google Earth Engine is the only practical free option.

The data sovereignty note applies here. Both SentinelHub and Google Earth Engine process ESA public-domain pixels. The underlying Sentinel-2 data is permanently free under the Copernicus Open Licence. A proprietary precision-agriculture platform packaging Sentinel-2 NDVI into a $200 to $2,000 annual subscription adds processing convenience on top of publicly owned imagery. The field-level history that accumulates in a closed platform becomes the rent-stack problem the Data Sovereignty spoke names directly: the satellite pixels are public, but the aggregated farm intelligence is not, once it flows into a platform the operator does not control.

Named Operators and Documented Verification

The Savory Institute's Land to Market programme, which verifies ecological outcomes across its network of Savory Hubs and licensed ranching operations spanning North America, South America, Africa, Australia, and Europe, uses satellite-derived NDVI as one component of its Ecological Outcome Verification methodology (Savory Institute, "EOV Methodology v2.0," 2023). Participating ranchers provide geo-referenced field and pasture boundaries. The verification team runs Sentinel-2 NDVI time-series analysis across the agreed monitoring period, comparing trajectories against historical baselines to confirm that land under holistic planned grazing is trending towards improved ground cover and biomass production. The satellite layer does not replace annual in-person botanical and soil assessments, but it provides a temporally dense, spatially explicit verification layer at no cost to the ranching operation. No commercial satellite subscription is involved. The instrument is ESA's; the methodology is Savory's; the data stays with the operator.

The OpenTEAM consortium, the US-based farmer-owned data infrastructure network covering approximately 1.5 million acres across member farms including Stonyfield Organic, Wolfe's Neck Center for Agriculture and the Environment, and Clif Bar supply-chain producers (OpenTEAM, 2024), integrates satellite data through FarmOS API connections and partner tools. OpenTEAM's governing principle is that the farmer holds the data and that remote sensing enriches the agronomic record without creating a dependency on any platform's continued subscription. Sentinel-2's Copernicus Open Licence means the satellite layer within this architecture cannot be withdrawn, repriced, or controlled by a third party. The NDVI composite for a cover crop field at Wolfe's Neck in November is generated from the same free ESA pixels as the NDVI for a Bayer FieldView subscriber in Iowa: the difference is not the signal. The difference is where the field-level history lives and who profits from it.

The broader comparison of satellite monitoring alongside drone imagery, ground sensors, and third-party verification platforms is developed in the satellite and drone monitoring overview spoke. This spoke is narrowed to Sentinel-2 specifically: the instrument, the indices, the four regen use cases, and the three access paths that cost nothing. Sentinel-2 makes every hectare visible from orbit, every five days, for free. What the operator does with that visibility is the management question, not the access question. The access question was answered in 2015.