Create a project
Open BR-MANGUE Studio, create a project folder, and choose where the simulation outputs will be saved.
Practical guide
From your data to a clear result, step by step.
Prepare your inputs, map the classes, set a scenario, and inspect annual changes without leaving BR-MANGUE Studio.
Quick start4 stepsData · Classes · Scenario · Results
Load aligned rasters
Select a categorical land-cover raster and an elevation raster with the same grid, extent, and coordinate reference.
Map the classes
Inspect the numeric codes and assign each one to mangrove, natural vegetation, water, anthropized or blocked, or another model role.
Run and inspect
Set the years and coastal parameters, choose an engine, run the model, and explore annual maps, animation, trajectories, and CSV tables.
Workflow
Prepare and run each scenario with confidence.
Every screen is designed to keep the modelling decisions visible. Save the project before running so that input paths, class mapping, and parameters remain documented.
Project dataChoose land-cover, elevation, optional study-area mask, and optional suitability raster.
Class mappingConnect source codes to model roles. Excluded codes are treated as NoData.
SimulationDefine initial and final years, tidal influence, sea-level rise, accretion, block size, and engine.
ResultsReview annual figures, animation, trajectories, state counts, and transition tables.
Inputs
Give the model a clean grid.
Land-cover and elevation rasters must be aligned. The land-cover file is categorical; the elevation file supplies the relative surface used by the transition rules. A soil suitability layer is optional.
Land coverGeoTIFF · numeric classes
ElevationGeoTIFF · same grid
SuitabilityOptional GeoTIFF
Coordinate systemSame CRS or reproject first
Processing engines
Choose the scale that fits your study.
Fastest when it fits
Keeps the active grid in RAM and avoids block-to-block disk I/O. Choose it when the raster fits comfortably in memory. The same state arrays are reused at every step, so adding years increases time, not RAM in direct proportion.
Capacity and recovery
Copies the input state to persistent files, releases the materialized grid while the steps run, and processes one chunk at a time. Choose it for very large coastal rasters or long runs where capacity and recovery matter more than raw speed. The grid is reopened after the run when maps and figures need it.
Audit checklist
Make the result traceable.
Before interpreting a map, record the raster sources, class mapping, coordinate reference, years, tidal influence, sea-level rise, block size, and processing engine.
User manual
Learn the model from the first cell to the final map.
The illustrated Portuguese manual explains cellular automata, neighbourhoods, raster preparation, every interface field, processing engines, outputs, troubleshooting, and reproducibility.
Download the illustrated PDFComplete guideBR-MANGUE Studio · Manual 1.0.0Captures of the interface, original maps and graphs, and plain-language interpretation.
Demonstration run
See one complete run before you start.
This animation was generated directly by BR-MANGUE Studio from the Ilha de São Luís demonstration run (2025–2100; 1,083,556 cells). It is an illustrative example of the workflow, not a predictive result for the island.
The complete figures and CSV outputs are documented in the manual and generated in the results folder after each run.
