Honey Locust

Honey Locust is an ongoing side project exploring how machine learning and spatial data can support better understanding of tree health — starting with the Honey Locust (Gleditsia triacanthos).

The idea began as a small visualization exercise, mapping occurrence records from the GBIF API to see where the species appears across Argentina. From there, the focus has shifted toward understanding what canopy- or soil-related variables might correlate with stress or decline.

Distribution Map

Below is an interactive map showing Honey Locust occurrence records across Argentina, based on GBIF data. The heatmap visualization reveals regional patterns and helps identify areas of higher species density.

What’s inside

• Spatial data exploration. Interactive Folium maps display Honey Locust occurrences and density heatmaps, making it easier to visualize regional patterns and sampling bias.
• Data alignment. Early scripts normalize canopy imagery, parcel data, and inspection notes to prepare for feature extraction.
• Baseline modeling. The next phase will test simple tabular and multimodal ML models (starting with PyTorch) to explore whether canopy-derived features can predict stress indicators.

While still at an early stage, the long-term goal is to develop open, reproducible baselines that link ecological field data with public imagery and help researchers or city planners detect potential canopy issues earlier.

If you have forestry expertise, environmental data, or interest in similar ecological ML projects, feel free to reach out.