About Geoneon
Geoneon builds scalable, science-backed climate risk mapping products using AI and Earth observation. We turn satellite and geospatial data into decision-ready layers for governments, insurers, utilities and communities — with a strong focus on wildfire preparedness and vegetation intelligence.
Why we are hiring
We are looking for a Wildfire & Forest Scientist who can lead scientific development for our wildfire and vegetation products — and who is equally comfortable writing strong Python as they are designing rigorous validation and interpreting results in real landscapes.
The role
You will work across research, product and delivery to improve and extend Geoneon's wildfire and vegetation modelling capability and take models from concept → validated outputs → scalable pipelines.
What you'll do
Develop and refine wildfire-related layers (e.g., severity/susceptibility, exposure-relevant vegetation drivers, burn scar / recovery where relevant).
Build and improve vegetation / forest structure products (cover, density, canopy metrics, height/structure where feasible with available data).
Design robust evaluation and validation (spatial splits, error/bias analysis, uncertainty; calibration with field/LiDAR where available).
Build scalable Python pipelines for large-area geospatial inference (tiling, mosaicking, QA, reproducibility).
Integrate multi-source data: Sentinel-2/Landsat, DEM/topography, fuel/vegetation proxies, climate normals, and high-resolution commercial/aerial imagery when needed.
Communicate results clearly: model behaviour, limitations, and what outputs mean for real-world decisions.
You might be a great fit if you…
Have deep expertise in wildfire science, forestry/vegetation ecology, or remote sensing for vegetation, with an applied, quantitative mindset.
Are a strong Python developer and can ship maintainable code (not just notebooks).
Enjoy solving hard applied problems like generalisation across regions/biomes, and data limitations.
Essential skills & experience
Strong Python: scientific computing, geospatial processing, clean code practices.
Raster + geospatial capability: working with large AOIs and gridded datasets (e.g., rasterio/rioxarray/xarray; GDAL concepts; COGs; chunking).
Scientific rigour: experimental design, validation/metrics, clear interpretation of results.
Demonstrated experience in at least one of:
wildfire fuels / severity drivers / fire effects and recovery
forestry / vegetation structure mapping
EO-based vegetation classification, change detection, canopy metrics
Highly regarded (nice-to-have)
ML / deep learning for EO (PyTorch; segmentation/regression; domain shift/generalisation).
Vegetation height/structure estimation using high-res imagery and/or LiDAR calibration.
Cloud-scale geospatial workflows (Dask, containers, CI, basic MLOps).
Experience delivering products for government/industry stakeholders.
What success looks like (first 3–6 months)
You have taken ownership of a core modelling thread (wildfire severity drivers and/or forest structure mapping improvements).
Validation + QA is stronger and more defensible, with clear performance narratives and failure modes.
Our Python workflows are more scalable and repeatable for national-scale production runs.
How to apply
Send a short note explaining your fit, plus:
A CV (or LinkedIn profile)
A link to code (GitHub, publications with code, or sample repo/notebook)
Anything showing applied impact (maps, products, reports, case studies)