Data Sources
Two eyes on every field: satellite overhead, sensors underground.
Prescription accuracy depends on the quality of the data going in. Soilynx combines ESA Sentinel-2 and USGS Landsat imagery with calibrated in-field soil probes — two independent data streams that cross-validate each other’s readings before a zone boundary is drawn. When they disagree, Soilynx flags the discrepancy rather than averaging it away.
Data Source One
Satellite multispectral imagery
Freely available ESA and USGS satellite imagery, processed on ingest into the band combinations that matter for corn and soybean production.
Sentinel-2 (ESA)
10-meter spatial resolution. 13 spectral bands including red-edge. 5-day revisit frequency at Iowa latitudes. Free, open-access data from the European Space Agency’s Copernicus program.
Landsat 8/9 (USGS)
30-meter resolution. 16-day revisit. Used as cloud-gap fill when Sentinel-2 passes are cloud-obscured. Thermal infrared band provides canopy temperature data for irrigation stress detection.
Cloud masking
Scene Classification Layer (SCL) masks clouds, cloud shadows, and water at pixel level. Masked pixels are filled from the nearest cloud-free pass. No cloud-contaminated pixels reach the zone model.
Band combinations computed per field
NDVI
Normalized Difference Vegetation Index. (NIR − Red) / (NIR + Red). General crop vigor proxy; correlates with leaf area index.
GNDVI
Green NDVI. Substitutes Green band for Red. More sensitive to chlorophyll concentration in dense-canopy, high-biomass corn at mid-season.
Red-Edge CI
Red-edge chlorophyll index. (NIR / Red-edge) − 1. Detects early nitrogen stress before visual symptoms appear in the crop canopy.
Data Source Two
In-field soil sensor integration
Soil sensor probes measure what satellite imagery can’t see — subsurface moisture content, temperature stratification, and bulk electrical conductivity at three depths.
Supported hardware
- Campbell Scientific — CS655, CS650, CS616 TDR probes
- Meter Group (Decagon) — TEROS 12, TEROS 21, 5TM sensors
- Stevens Water — Hydra Probe II, HydraPoint
- AquaSpy — Full-profile tube sensors (12-inch increment)
- Sentek — Drill & Drop profile probes
Don’t see your sensor brand? Contact us — we add hardware integrations on request for operators with existing installations.
Ask about your sensor hardwareWhat sensors measure
- Volumetric water content at 15 cm, 30 cm, and 60 cm depth
- Soil temperature at each depth (±0.5°C accuracy)
- Bulk electrical conductivity (proxy for clay content and salinity)
- Readings transmitted at 15-minute intervals
No sensors yet?
Soilynx also works with satellite data alone. Sensor integration improves prescription accuracy but is not required to get started.
Data Fusion
How satellite and sensor data combine
A single data source produces one view of field variability. Fusion of multiple co-registered sources produces a validated picture that is harder to distort from individual sensor failure or cloud gaps. Soilynx does not generate prescriptions from satellite data alone if sensor readings for that field show a VWC discrepancy of more than 15 percentage points versus the satellite NDWI estimate — the model flags the conflict and holds the export until you confirm.
Co-registration
Satellite raster layers are projected into the same coordinate reference system as sensor point locations. All data is spatially co-registered to the field boundary at 10m pixel resolution before zone generation begins.
Cross-validation weighting
Where sensor VWC readings and satellite NDWI disagree strongly, the model flags the discrepancy and requests confirmation before applying the prescription. Sensors and satellite serve as checks on each other.
User-supplied data integration
Soil test points from commercial labs (SSA, Midwest Labs, A&L), historical yield monitor data from combine, and fertilizer application history are all ingested and held as stable base-layer inputs to the zone model.
Questions about data
Not sure if your data setup qualifies?
We can review your current hardware and data situation in a 15-minute call. No sales pitch — just an honest answer about what Soilynx can and can’t do with your field data.