Satellite vs. Drone Imagery
Which imagery source to pick for a prescription map — costs, resolution, freshness, and field-by-field trade-offs.
The wizard lets you drive a prescription from satellite scenes or from your own drone (OTM) imagery. Both produce the same output (NDVI raster → zones → VRA shapefile). The trade-offs are about resolution, freshness, cost, and effort.
At a glance
| Satellite (Sentinel-2) | Drone (OTM orthomosaic) | |
|---|---|---|
| Resolution | 10 m / pixel | 2–10 cm / pixel |
| Freshness | Latest pass — up to 5 days old | As fresh as you fly |
| Cost to capture | Free | Drone flight time + processing |
| Effort | One click | Image capture flight + upload + processing wait |
| Cloud risk | Yes — affects many scenes | No |
| Field size practicality | Any size | Best up to a few hundred acres per flight |
| NDVI accuracy at field edges | Mixed pixels at boundary | Pixel-perfect at boundary |
| Pre-emergence usefulness | Good (bare soil pattern visible) | Excellent (full-resolution bare soil) |
When to pick satellite
Pick satellite when any of the following are true:
- You need a prescription today and you don't have time to fly.
- The field is large and an OTM mission would take hours.
- The NDVI signal you care about is field-wide (whole-field stress vs. healthy patches) — 10 m resolution is plenty.
- The agronomic problem is chronic (low spots, salinity, soil-type patterns) — last week's NDVI tells the same story as today's.
Satellite is also the right answer for the first prescription on a new field — see what's already visible from space before deciding whether a drone flight will add useful detail.
When to pick drone
Pick drone when any of the following are true:
- The decision is time-sensitive (e.g., responding to a developing stress within a 1-week window).
- The agronomic feature you care about is small — disease pockets, lodged corn strips, missed sprayer passes.
- The field has cloud cover that's been blocking satellite scenes for weeks.
- The field is small (< 100 ac) where the cost of a capture flight is low.
- You're doing research / trials and need true pixel-level resolution.
What's in a satellite scene
The portal pulls Sentinel-2 scenes (with future support for Planet and Landsat). Every scene includes:
- True Color for visual confirmation.
- NDVI computed from the red and near-infrared bands.
- Date/Time of capture.
- Cloud Cover percentage for that scene over your AOI.
- Platform identifier (Sentinel-2 A or B).
If no acceptable scene is available you'll see "No recent satellite image found for this area. Try adjusting your plot or try again later." Cloud cover and overflight schedules are the usual reasons. Try again in a day or two, or fall back to drone.
What's in a drone OTM dataset
OTM = orthomosaic. To produce one, you:
- Fly an OTM capture mission with the drone (vertical camera, fixed altitude, ~80% overlap).
- Upload the raw images to the portal.
- Wait for processing. Status transitions: Pending → Uploading → Queued → Processing → Completed.
- The portal stores both the true-color orthomosaic and a multi-band version, from which NDVI is computed.
Once processing completes, the dataset appears in the Processed Images dropdown of the Map Prescription wizard.
Mixing both
Some teams use satellite as the baseline and drone for spot-check when the satellite signal looks suspicious. The portal doesn't currently let you blend the two NDVI rasters into a single prescription, but nothing stops you from generating two prescriptions side by side and comparing.
Common mistakes
- Stale satellite scene: a scene from two weeks ago tells you about two weeks ago. Check the Date/Time in the info card before clicking Save.
- Cloud cover > 20%: even a moderately cloudy scene will produce wildly wrong NDVI in the cloudy parts. If cloud cover is high, wait or fall back to drone.
- Bare-soil drone flight after planting: NDVI on a fresh plant doesn't tell you much. Wait at least V3–V4 (corn) or R1 (soybean) before flying for prescription purposes.
- Wrong field-edge effect on satellite: the 10 m pixels at the field edge are a mix of crop and headland. Pull the plot polygon inwards by a swath width if you're seeing odd edge zones.
Related
Was this page helpful?
Last updated Oct 20, 2018