We build spatial models that locate areas susceptible to wildfires, floods, erosion and other environmental processes, helping you prioritize studies, prevention measures and on-the-ground interventions.
01 / What changes in your work
Today
With spatial modeling
02 / What we can model
First we define the decision you need to support; then we choose the method. These are some of the questions we can answer.
Environmental risks
Susceptibility to floods, erosion, wildfires and droughts at the watershed or municipal scale, as a foundation for prevention and land-use planning.
Floods · Erosion · Wildfires · Droughts
Predictive modeling
High-resolution predictive maps that combine geostatistics and machine learning to estimate key variables across your entire study area.
Soil carbon · Species distribution · Crop yield
Climate change scenarios
Downscaling of hydroclimatic variables and bias correction of global climate models to obtain local-scale projections under different scenarios.
Bias correction · Downscaling · SSP scenarios
Hydrology & climate
Analysis of precipitation, temperature and evapotranspiration to assess droughts and water availability with high-resolution spatial data.
SPI / SPEI · Watershed delineation · Evapotranspiration
Geostatistics
Spatial interpolation and autocorrelation analysis to turn point data into continuous maps, with uncertainty estimates and hybrid models that incorporate machine learning.
Kriging · Variograms · Hybrid models
Custom statistics
From regressions and mixed models to time series and simulations: we choose the technique based on your data and your goals, not the other way around.
GLM · GAM · PCA · Clustering · Monte Carlo
03 / Rigor
Confusing them leads to the wrong decisions. That's why every map we deliver states explicitly what it represents, based on the information it actually incorporates.
01 · Susceptibility
Where could it happen?
How prone a place is to a given phenomenon, based on its conditions: slope, soil, vegetation or climate.
02 · Hazard
How likely, and how intense?
Adds the probability, frequency or magnitude with which the phenomenon may occur over a given period.
03 · Risk
What could we lose?
Combines the hazard with what is exposed—people, infrastructure, ecosystems—and how vulnerable it is.
A map without validation is just an opinion in color. Every model comes with the information you need to know how much to trust it.
We evaluate the model with data it wasn't trained on and, where appropriate, with spatial validation.
AUC, RMSE, R² or others, depending on the model type, reported alongside the map.
Ranges, uncertainty maps or model agreement, when the method allows it.
Documentation of each model's data, period, resolution, assumptions and limitations.
Methods and tools
04 / Models in action
Technical demo
Wildfire susceptibility across Chile at 1 km resolution, based on 2000–2023 fire records.
Explore the viewer
Technical demo
Flood susceptibility index at 30 m resolution, integrating topography, climate, vegetation and infiltration.
Explore the viewer
Technical demo
Species distribution model with current and future projections of climatically suitable areas for the mosquito, at 1 km resolution.
Explore the viewer05 / Let's talk
Tell us what decision you need to support and what data you have. We'll tell you what kind of model makes sense and what level of accuracy you can reasonably expect.
Explore other solutions
Web applications
Remote Sensing & GIS
Data infrastructure