Projects
From continent to pond: five projects, from 5 km to 1 m.
Explore GeoVita's research on interactive maps: continent-wide malaria forecasting, the dengue mosquito under future climate across Mexico, habitat suitability for a venomous scorpion in south-western Iran, province-wide screening of larval habitats at 10 m, and a 1 m habitat model for a single town.
Africa · 5 km
Malaria endemicity on a living globe.
Annual Plasmodium falciparum parasite rate in children aged 2–10 (PfPR₂–₁₀) across 1.46 million grid cells, observed by the Malaria Atlas Project and predicted by GeoVita's ConvLSTM2D model from nine years of environmental history.
- Region
- Mainland Africa and islands · 1,457,085 cells
- Period
- 2000–2024 · forecast test 2022–2024
- Data
- MAP PfPR₂–₁₀, NASA GLDAS-2.1 Noah (9 variables), MODIS NDVI
- Method
- ConvLSTM2D vs 5 trained models and 2 baselines · 250 km spatial-block CV + untouched temporal holdout
- Resolution
- ≈ 5 km (0.0417°)
- Lead researcher
- Fahimeh Youssefi et al.
- Status
- Peer-reviewed article
PfPR₂–₁₀ measures endemicity (prevalence), not clinical incidence. The MAP surfaces are model-based references; forecasts complement, but never replace, epidemiological surveillance.
Model benchmark
Six trained models, two honest baselines.
Every approach was tested twice: on geographically separated 250 km blocks (spatial transfer) and on an untouched 2022–2024 period (future-year forecasting).
01
Best spatial transfer
ConvLSTM2D ranked first on unseen regions: RMSE 0.022, r 0.993, NSE 0.986.
02
A demanding benchmark
On 2022–2024, simple persistence ranked first overall; ConvLSTM2D was the best trained model, with no clear RMSE difference between them.
03
Humidity, wind, canopy water
Specific humidity, wind speed, canopy interception and air temperature carried the most predictive signal; the previous year mattered most.
RMSE by model — Spatial CV
Lower is better
- Model
- Baseline
Forecast error by horizon (RMSE)
ConvLSTM2D leads one year ahead; the persistence baseline catches up as the horizon grows.
- ConvLSTM2D
- Persistence
- Other models
What drives the ConvLSTM2D forecast
Increase in RMSE when each predictor is shuffled (permutation importance). Predictive dependence, not causation. ΔRMSE ×10⁻³.
Which years matter most
Increase in RMSE when all predictors in one lag year are shuffled. ΔRMSE ×10⁻³.
Want this for your region?
We co-design pilots with local partners, data and validation.