Research group in Computing and Geotechnologies at the Federal University of Maranhão (UFMA), São Luís, Brazil. Registered in the CNPq Directory of Research Groups.
We work at the intersection of Computer Engineering, Geosciences and Open Science, building open, reproducible tools for spatial simulation, land-change modelling and semantic data integration.
"Do dado aberto ao modelo verificado, em código que qualquer um pode rodar."
From open data to a verified model, in code anyone can run.
🌐 lambdageo.ufma.br · ✉️ lambdageo@ufma.br
A Python ecosystem for discrete spatial simulation and land-change modelling, an independent alternative to INPE's TerraME/LuccME (which remain the reference). Its packages live in the DisSModel organization (documentation); the reference outputs, benchmarks, study cases and workflow live here. The repository map shows how they connect.
Verification against TerraME/LuccME
- luccme-goldens · terrame-goldens: reference outputs of TerraME/LuccME, pinned by SHA-256.
- disscube-benchmark · disslucc-benchmark: parity benchmarks of the DisSModel packages against those references.
Workflows and study cases
- open-lucc-workflow: a traceable land-use change experiment built from open data.
- disscube-grids: study-area examples that build gridded variables from open data.
- rescience-lulcc · lulcc-goldens: replication of a published land-change study for ReScience C.
- brmangue-terrame: cellular automata model of sea-level rise impacts on mangroves along the Brazilian coast.
- qgissparql: QGIS plugin to import Linked Data from SPARQL endpoints as spatial layers.
- rdfmapper: declarative Object-RDF mapper for Python.
- dbcells-ontology: RDF vocabulary for publishing land-change models as Linked Data.
- terrahs: a Haskell GIS toolkit, rewritten without its original TerraLib dependency.
- playreg: a regular-expression engine based on Brzozowski derivatives.
- geomosaic: a Python tool for building geospatial mosaics.
- fielddash: schema-driven Streamlit dashboards for field data collection.
- geospatial-modeling-python: open textbook on geospatial Python, from the basics to DisSModel (in Portuguese).
- Code of published studies: analise-cerrado-ma (deforestation in the Maranhão Cerrado) and acuracia-slz (land-cover accuracy around São Luís).
- Course textbooks: LambdaGeo Edu · lambdageo-edu.github.io.
Led by Prof. Sérgio Souza Costa. Students interested in environmental modelling, GIS and reproducible research software are welcome: write to lambdageo@ufma.br.
