Satellite monitoring for farms and land: what it can and cannot tell you

Earth observation satellites photograph the same places again and again. For farms, estates and large landholdings, that regular record makes it possible to see change without visiting every field.

What satellite imagery can show

• Crop condition — differences in vegetation health across and between fields, often shown with indices such as NDVI.
• Land cover — cropland, bush, bare ground, water and built areas, and how they change.
• Change over time — clearing, new structures, flooding or fire scars between two dates.
• Field boundaries and areas — for planning and record keeping.

How the analysis works

1. Choose imagery — the right satellite and dates for the question and the size of the area.
2. Prepare — remove cloud-covered areas and line images up on the map.
3. Calculate — indices and classifications from the different light bands the satellite records.
4. Check on the ground — field visits confirm what the images suggest.
5. Report — maps and summaries that point to where attention is needed.

Machine learning can help sort large areas into land-cover classes or flag unusual changes, but it needs examples checked on the ground to be reliable.

Limits to keep in mind

• Clouds hide the ground, especially in the rainy season.
• Resolution — each pixel covers an area of ground, so small features may not be visible.
• Satellite data shows symptoms, not causes: a stressed patch could be water, pests, disease or soil, and only a field check tells which.
• Timing — an image shows one moment; regular images show trends.

When drones are the better tool

Drones capture much finer detail over smaller areas and can fly under cloud. Satellites cover large areas regularly at lower cost per hectare. Many projects use both: satellites to find where to look, drones and field visits to look closely.

Used well, satellite monitoring helps decide where to send people and resources first — it does not replace the people on the ground.