Which satellite data inputs are most useful for predicting malaria risk?

The core trio is land surface temperature (LST) from MODIS or VIIRS, normalised difference vegetation index (NDVI) from Landsat or Sentinel-2, and surface water extent from SAR (Sentinel-1 or ICEYE). These proxies capture mosquito development temperature thresholds, breeding habitat greenness, and standing-water persistence — the three dominant environmental drivers of Anopheles suitability. Rainfall estimates from GPM IMERG improve the model further by predicting ephemeral pool formation.

Why shouldn't a government simply buy risk maps from a commercial provider like Planet or Spire?

A commercially procured map is a snapshot licensed under terms the vendor can change, restrict, or price up at contract renewal. When a government owns the downstream processing pipeline — even if feeding on open Sentinel or Landsat data — it retains the analytical capability, the historical archive, and the freedom to integrate classified population or military-movement data that no commercial vendor should hold. During a declared public health emergency, speed and control of the data product matter more than cost savings on the imagery itself.

How often do risk maps need to be refreshed to be operationally useful?

Entomological evidence suggests that Anopheles habitat changes significantly on a 5–10 day cycle following rainfall events, so weekly updates are the minimum meaningful cadence for intervention targeting. Daily revisit constellations (Planet SuperDove, Satellogic) make this achievable, while the free Sentinel-2 constellation offers 5-day revisit at 10 m resolution as a cost-free baseline. Risk indices for slower-moving vectors like sandflies (Leishmania) can tolerate monthly updates.

Can a low-income country realistically build this capability indigenously?

Yes, at the analytics layer — which is the highest-value part. Open data from USGS Landsat, ESA Sentinel, and NASA MODIS is free; processing using Google Earth Engine or open-source tools (SNAP, QGIS, R) costs nothing in licensing. A nation needs perhaps 4–6 trained earth-observation analysts embedded in the national public health institute, supported by a sovereign cloud or government data centre. Owning the satellite constellation itself is a second-order ambition; controlling the analysis pipeline is the immediate sovereign priority.

How does this differ from the outbreak hotspot detection application?

Vector-borne disease risk mapping is prospective and environmental — it identifies where conditions favour vector survival and reproduction before human cases occur. Outbreak hotspot detection is retrospective and case-based — it clusters confirmed or suspected cases to find active transmission foci. The two are complementary: risk maps guide pre-emptive vector control, hotspot detection triggers rapid case response. Ideally, both feed the same national health dashboard.

What accuracy should a government expect from a satellite-based risk model?

Peer-reviewed studies in sub-Saharan Africa report area-under-curve (AUC) values of 0.75–0.88 for satellite-driven malaria risk models, meaning they correctly rank high-risk areas significantly better than chance but are not perfect. Accuracy degrades during cloud-contaminated seasons, in data-sparse regions, and when applied outside the geographic area used for model training. Governments should treat outputs as decision-support tools to prioritise field investigation, not as definitive maps that replace entomological surveillance.

What role does the WMO and EUMETSAT data infrastructure play?

WMO's Global Telecommunication System and EUMETSAT's data services provide near-real-time meteorological inputs — particularly precipitation, humidity, and surface temperature — that are essential driver variables for vector-habitat models. EUMETSAT's Meteosat Third Generation (MTG) satellites, launched from 2023, deliver 15-minute full-disk thermal imagery over Africa and Europe that substantially improves the temporal resolution of LST inputs to disease models.

Are there international agreements that obligate data sharing for disease surveillance?

The WHO International Health Regulations (IHR 2005) oblige member states to notify WHO of potential public health emergencies of international concern (PHEIC) and to share relevant epidemiological data, which increasingly includes environmental and geospatial datasets. However, the IHR contains no specific mandate to share satellite imagery, leaving data-sharing largely voluntary and subject to bilateral negotiation. This gap is one reason sovereign analytical capability — rather than reliance on a neighbour's goodwill — is essential.

13.2.1Live

Vector-Borne Disease Risk Mapping

Health worker crouching beside stagnant water in a rural tropical village, examining mosquito breeding sites — A child holding two white containers illustrates everyday exposure risks that satellite disease mapping helps communities identify and reduce.

Satellite Mapping of Mosquito-Borne Disease Risk

1
Earth Observation
Multispectral sensors map vegetation, water bodies
2
Downlink & Ingestion
Daily imagery streamed to ground processing hubs
3
Risk Modelling
ML fuses climate, land-use, incidence data
4
Health Alert
District maps trigger targeted vector-control response

Satellite land-cover and rainfall data feed disease models that deliver actionable vector-risk maps to local health authorities.

Page in the build queue

This application has its place in the atlas. The full page — short and long descriptions, what-matters bullets, sovereignty rationale, and the eight-field architecture block — is being written next.

Maturity reading: Operational systems exist with paying customers. Phase 1 built §4.1 (Maritime Intelligence) end-to-end as the depth-and-format reference. Section 4 is the next section to graduate from scaffold to full content; the rest follow on the roadmap.

Related applications

All in disease intelligence

13.2.2 · Next

Outbreak Hotspot Detection

Bring this into your sovereign programme.

All solutions Speak with engineering