Satellite monitoring of oil palm in Colombia: indices, alerts and costs
How satellite monitoring detects stress, bud rot and yield drops in oil palm before the human eye can. Indices, resolution and costs for Colombia.
Colombia is the world's fourth-largest palm oil producer and the largest in the Americas. With more than 600,000 hectares planted across four producing zones, oil palm is one of the crops where satellite monitoring delivers the clearest return: large blocks, 25-to-30-year cycles, and a silent enemy —bud rot (PC)— that spreads before it is visible from the ground.
This guide explains what can be detected from space, with which indices, at what resolution and at what cost, in the specific context of Colombian plantations.
Why oil palm needs satellite monitoring
A field walk across a 2,000-hectare plantation takes days and only covers what the agronomist can reach. A satellite, by contrast, delivers a complete, comparable reading every few days. The three reasons that most justify the investment:
- Early stress detection. A stressed palm changes its spectral response weeks before showing obvious symptoms. The satellite sees that signal in the near-infrared.
- Bud rot surveillance. PC spreads in foci. Catching a focus early lets you fence it off, eradicate and halt its advance across the block.
- Per-block yield traceability. Comparing vigor between blocks and across seasons reveals which blocks underperform their potential, and why.
What can be detected from space
Not everything shows up equally well. These are the indices we use and what each is good for in palm:
NDVI, MSAVI and GNDVI
NDVI is the starting point, but it saturates in mature palm: the canopy is so dense that the index stops distinguishing "healthy" from "very healthy". So we combine:
- MSAVI for young palm, where exposed soil distorts NDVI.
- GNDVI, more sensitive to chlorophyll content, useful for separating stress levels under a closed canopy.
- NDWI / NDMI for moisture content, key in zones with seasonal water deficit.
Bud rot foci
PC isn't "seen" directly from space, but its consequence is: the progressive loss of vigor in a group of palms. A multi-temporal analysis —comparing the same block month over month— makes the foci obvious, where vigor falls steadily while the rest of the block holds. That turns an image into an actionable alert.
Resolution and satellites: what to choose
| Source | Resolution | Frequency | Cost | Typical palm use | |---|---|---|---|---| | Sentinel-2 | 10 m | ~5 days | Free | Block-level vigor monitoring | | Landsat 8/9 | 30 m | ~16 days | Free | Long historical series | | PlanetScope | 3 m | Daily | Commercial | PC foci, palm-group detail | | Drone (multispectral) | <10 cm | On demand | Commercial | Field confirmation and counting |
The strategy that works best is tiered: Sentinel-2 to sweep the whole plantation continuously and for free, and high resolution (Planet or drone) only over the foci the sweep flags as suspicious. That controls cost without losing detail where it matters.
How we do it at GeoSAT
Our precision-agriculture platform, Geobristol, is built on this logic. As an EOS Data Analytics (EOSDA) partner, we process Sentinel-2 series and generate:
- Per-block vigor maps, normalized to compare across seasons.
- Multi-temporal alerts when a sector falls below its baseline.
- Reports for the agronomy team —not raw files to interpret.
It's the same approach we use to monitor close to 2,800 farmers within the FAO food-security project in Medellín, adapted to the scale and timing of a palm plantation.
When it makes sense (and when it doesn't)
Satellite monitoring pays off when there is enough area and decisions that depend on the data: plantations from a few hundred hectares, agronomy teams that act on alerts, and a multi-season horizon. On very small blocks or without field response capacity, the cost of high-resolution tracking isn't recovered.
Its limits are real too: cloud cover in humid zones reduces usable optical imagery (where Sentinel-1 radar helps as a complement), and the satellite tells you where to look, but confirming the diagnosis —PC, pests, nutrition— is still field work.
How to start
The most cost-effective path is a free Sentinel-2 baseline over the whole plantation plus a multi-temporal analysis of recent seasons. That reveals the problem blocks with no upfront investment and defines where high-resolution tracking is worth it.
To evaluate your plantation, see our satellite monitoring service or get in touch for a trial on your blocks. And for the agronomic side, read why NDVI alone is no longer enough to understand which index to use at each crop stage.