Reviewed August 2026 against USDA NASS Cropland Data Layer documentation, NASA’s Landsat and OpenET program pages, and USDA ERS farm-technology survey data.

Try it: Run your own numbers →

A satellite data API is a programmatic feed: you send coordinates, a date range, and a product type, and it returns imagery, a vegetation index, or a water-use estimate instead of a map you’d click through by hand. In US agriculture, the core sources are already free — USDA’s Cropland Data Layer, NASA’s Landsat missions, and NASA’s OpenET water-use tool all run on public satellite data and expose it through an API in agriculture use cases that range from crop scouting to irrigation-compliance reporting. This piece breaks down what these agriculture satellite data sources deliver in meters, days, and dollars, so you can pick one instead of guessing.

Table of Contents

• What a Satellite Data API Actually Returns
• Agriculture Satellite Data: Free Sources vs Paid APIs
• How Farms, Foresters, and Explorers Use the Feed
  • Crop Health Monitoring & Land-Use Classification
  • Deforestation & Forest Health Tracking
  • Soil Moisture, Evapotranspiration & Irrigation
  • Pest & Disease Outbreak Prediction
  • Forest Inventory & Silviculture Planning
  • Mining Site Surveillance & Mineral Exploration
  • Try it: Run your own numbers
• Choosing and Implementing a Satellite Data API
• How to Verify These Figures Yourself
• Try It: Field Pixel & Revisit Planner
• FAQ
• Conclusion

What a Satellite Data API Actually Returns

Call it a satellite API, an API satellite feed, or a satellite data API — the mechanics are the same across providers. You submit an area of interest (a bounding box, a polygon, or field-boundary coordinates), a date range, and a product type; the service authenticates your key, pulls matching scenes from its archive or tasks a new capture, and returns a GeoTIFF, PNG, or JSON payload: a raw image, a vegetation-index layer, or a water-use estimate, depending on which product you asked for.

Three attributes decide whether a given source fits your use case:

• Spatial resolution — the ground distance one pixel represents. USDA’s Cropland Data Layer moved from 30-meter to 10-meter resolution starting with the 2024 crop year, built from Landsat 8/9 OLI/TIRS and Sentinel-2A/2B imagery, per USDA NASS’s Cropland Data Layer documentation.
• Revisit cadence — how often a fresh image is available for the same spot. A lone Landsat 9 pass repeats every 16 days; flown alongside Landsat 8, the pair covers the same ground every 8 days, per NASA’s Landsat 9 mission page.
• Coverage and cost — global or regional, free or metered. NASA credits Landsat’s free-and-open data policy with contributing $25.6 billion to the US economy in 2023 — public evidence that free-tier satellite data already carries real economic weight, not just novelty value.

Availability is not the same as use. In the 2023 Agricultural Resource Management Survey (ARMS), guidance and autosteering systems — a more mature precision-ag technology than most satellite feeds — were reported on 52% of midsize farms and 70% of large-scale crop farms, while yield monitors, yield maps, and soil maps combined reached 68% of large-scale farms, according to USDA’s Economic Research Service. Satellite-fed layers sit downstream of that same adoption curve: the data is free and available now; the gap is integration, not access.

Agriculture Satellite Data: Free Sources vs Paid APIs

The phrase “satellite data for agriculture” covers a wider range of sources than most searches assume. Below are the six that matter for US farms, forests, and exploration sites, with the specifications each publishes.

The Cropland Data Layer row shows why “resolution” is a moving target rather than a fixed spec. From the 2008 crop year through 2023, the CDL held at 30-meter resolution built primarily from Landsat; the 2024 crop year jumped to 10 meters as Sentinel-2 became a primary input, per USDA NASS. Producer accuracy for major crop categories on the CDL runs 85–95%, validated against USDA Farm Service Agency ground-truth data — the layer is not raw imagery, it’s a classified product checked against what farmers actually reported planting.

How Farms, Foresters, and Explorers Use the Feed

The same underlying imagery feeds six distinct workflows, each pulling different bands and update cadences from the sources above.

Crop Health Monitoring & Land-Use Classification

NDVI and related vegetation indices, computed from the red and near-infrared bands common to Landsat and Sentinel-2, flag water stress, nutrient deficiency, and disease before they’re visible from the field edge. Published studies on corn yield prediction using NDVI-based models report average accuracies in the 93–95% range against final harvest figures, and a USDA Agricultural Research Service regression approach using MODIS-derived indices kept predicted yields within 20% of official estimates, with most falling within 10%. Land-use classification — confirming what’s actually growing on a parcel, as opposed to what’s zoned or permitted — runs on the same imagery; see our guide to land-use classification types, methods, and global standards for how classification schemes differ by country and agency.

How fresh that NDVI layer can be depends entirely on which source feeds it — and the gap between options is large:

Deforestation & Forest Health Tracking

Change detection on temporal image stacks — comparing the same pixel across successive Landsat or Sentinel-2 passes — flags canopy loss, burn scars, and regrowth. The Forest Service’s Forest Inventory and Analysis (FIA) program pairs this kind of imagery with ground plots to stratify its national forest census; its BIGMAP tool explicitly combines FIA’s plot data with remote sensing to produce forest-resource maps at national scale, per USDA Forest Service FIA program documentation. For a landowner or regulator, that means canopy-health trend lines and illegal-logging alerts don’t require a ground crew to walk every acre first.

Soil Moisture, Evapotranspiration & Irrigation

Thermal and shortwave bands feed soil-moisture proxies, but the most concrete US example is NASA’s OpenET, which converts Landsat-scale imagery into a direct evapotranspiration — actual crop water use — estimate at quarter-acre pixels, updated daily, monthly, and annually, across the 23 westernmost continental states. NASA launched OpenET’s public API on October 3, 2023, specifically so water managers and farmers could pull the data into their own systems instead of using the web viewer by hand. The clearest measured outcome so far: in California’s Sacramento–San Joaquin Delta, reporting water use — a process that used to take farmers half a day to a full day — now takes as little as ten minutes, per NASA’s account of the OpenET API launch, which also credits the tool with saving thousands of dollars a year in flow-meter deployment and maintenance costs for some operations.

Pest & Disease Outbreak Prediction

Correlating multispectral imagery with climate and soil data lets a model flag pest and disease risk before a visible outbreak, rather than after. This is the least standardized use case of the six — there is no single US government API that outputs “pest risk” the way OpenET outputs evapotranspiration — so most operational systems here are built on top of the raw NDVI/thermal layers already described, tuned against a specific crop and region’s historical outbreak data.

Forest Inventory & Silviculture Planning

Manual timber-cruise surveys across remote forest tracts are slow and, on steep or roadless terrain, genuinely hazardous. The FIA program’s national plot network is remeasured on a five-to-ten-year cycle depending on the region, with aerial photography and classified satellite imagery used to stratify the population between plot visits and extend estimates across acreage that hasn’t been walked recently. That combination — sparse ground truth plus dense satellite coverage — is the same architecture agriculture uses with the CDL and FSA field reports; forestry just remeasures its plots on a longer clock.

Mining Site Surveillance & Mineral Exploration

Mineral exploration uses the same API pattern as crop and forest monitoring — submit an area of interest, get back an analysis — pointed at a different target. At Farmonaut, clients upload an area of interest (coordinates, polygons, or KML/KMZ) and a target mineral; multispectral and hyperspectral imagery, run through custom analytics, comes back as an intelligence report instead of a vegetation index. For a deeper technical look, see our Satellite-Driven 3D Mineral Prospectivity Mapping writeup, which covers how subsurface models and drilling-intelligence layers combine to reduce exploration risk.

Choosing and Implementing a Satellite Data API

Whichever of the six sources above fits, the implementation questions are the same:

• Define the decision before the data. “Monitor crop health” is not a spec; “flag any 10-hectare block where NDVI drops more than 15% week over week” is.
• Match cadence to the decision window. An annual layer like the CDL confirms what was planted; an irrigation call needs the multi-day cadence of Sentinel-2, Landsat, or HLS.
• Check the band list, not just the resolution number. Red-edge bands matter for crop stress, thermal bands for moisture and water use, SWIR for mineral alteration — a source with fine resolution but the wrong bands still won’t answer the question.
• Budget for integration, not just for the imagery. The satellite data itself is free across every US government source above; the recurring cost is compute, storage, and the pipeline that turns scenes into a usable API in agriculture or forestry software.
• Ground-truth on a schedule. USDA’s own CDL accuracy figures (85–95% by crop category) come from continuous comparison against FSA-reported ground data — build the same habit into any pilot before scaling it.

How to Verify These Figures Yourself

Every figure above carries a publication date because each of these programs updates on its own schedule. Before acting on a number from this page, check it against the primary source directly:

• CDL resolution and accuracy — USDA NASS republishes Cropland Data Layer documentation each release cycle at the page cited above; check it for the resolution and accuracy figures attached to the current crop year.
• Precision-ag adoption rates — USDA ERS updates its ARMS-based adoption charts after each survey cycle; the chart page cited above will carry newer percentages once the next ARMS wave publishes.
• Satellite specs and revisit times — NASA’s Landsat 9 mission page and NASA Earthdata’s HLS page are the sources of record for resolution and revisit; both change if a satellite is added, decommissioned, or repositioned.
• OpenET coverage — the 23-state western footprint is a stage in an announced rollout; check NASA’s OpenET reporting for whether coverage has expanded beyond that footprint.

Try It: Field Pixel & Revisit Planner

Enter a field size and pick a resolution and revisit cadence from the sources compared above to see how many pixels actually cover the field and how many passes to expect across a season.

FAQ: Satellite Data API for Agriculture, Forestry & Mining

What is a satellite data API?

An internet-based interface that returns satellite imagery or derived data — vegetation indices, land-use classification, water-use estimates — for a requested area and date range, in a machine-readable format like GeoTIFF or JSON, instead of a viewer you'd browse by hand.

Is agriculture satellite data actually free, or do I need to pay for an API?

The core US sources are free: Landsat, Sentinel-2, the harmonized HLS product, USDA's Cropland Data Layer, and NASA's OpenET all require no license fee. Commercial providers charge for tasked, sub-5-meter imagery or for hosted processing — request current rates directly from the vendor, since those are metered and change by plan.

What's the difference between a "satellite API" and "API satellite" imagery tasking?

They describe the same category from two angles. "Satellite API" usually means pulling from an existing archive (Landsat, Sentinel-2); "tasking" means requesting a fresh, on-demand capture over a specific area, which is a paid, commercial-only capability since no free government source tasks new imagery on request.

How current can the data actually get?

It depends on the source, not on how the request is worded. HLS's harmonized Landsat-Sentinel-2 feed reaches a 1.4-day combined revisit; Sentinel-2 alone reaches 5 days at the equator; Landsat 8+9 combined reaches 8 days. None of these numbers account for cloud cover, which can push the actual usable-image gap wider in any given week.

Can the same API cover forestry and mineral exploration, not just crop fields?

Yes — the request pattern (area of interest, date range, band selection) is identical across sectors. What changes is the analysis layered on top: NDVI and land-use classification for crops, canopy change detection for forestry, and multispectral/hyperspectral alteration-mapping for mineral exploration.

Do I need a GIS background to use one of these APIs?

To query Landsat, Sentinel-2, or OpenET directly, yes — you need at least enough GIS literacy to define an area of interest and handle GeoTIFF output. Farm-management platforms and specialist providers exist specifically to remove that requirement by delivering the same underlying data as a dashboard or report instead of a raw file.

Where do I start for a mining or mineral-exploration read on a site?

Start with an area of interest and a target mineral. Request a quote or contact the team directly to scope a project against a specific field, forest, or mining zone.

Conclusion

The technical case for a satellite data API in agriculture, forestry, or mining is no longer in question — Landsat, Sentinel-2, HLS, USDA's Cropland Data Layer, and NASA's OpenET already cover the resolution, cadence, and cost points most US operations need, at no license cost. The open question is integration: which resolution and revisit actually match your decision, and whether your pipeline turns that free feed into something a field team acts on. Both are answerable with the comparison table, the planner, and the verification checklist above — not with a fresh API call, but with the same public documentation cited throughout this page, checked again whenever a figure needs a newer date on it.