gmoncrieff / bfast-explorer Goto Github PK

My name is Glenn Moncrieff and I am a Geospatial Data Scientist and ML Engineer based in Cape Town, South Africa.

My background is in geography and ecology, and nowadays I spend most of my time observing the earth with satellites and machine learning 🛰️

I train models and conduct research to analyse environmental data 🌏 Most of my effort is focussed on addressing the biodiversity and climate crises.

I am most comfortable coding in Python and R.

My research has looked at a range of environmental issues like mapping plant biomes, modelling water loss to invasive plants, rapidly detecting habitat loss in shurblands or forecasting post-fire vegetation recovery. I also wrote about climate change impacts on African ecosystems in the Africa chapter of the latest IPCC report.

I like to see science turned into real-world applications and software that the community can use. So I spend most of my time turning research into packages or operational products. Some fun software that I have created or contributed to:

• Global Renosterveld Watch: Deploys trained tensorflow models to GCP via Apache Beam to predict shrubland habitat loss.

• hyper-iap: Mapping alien invasive plants from hyperspectral imagery using deep learning.

• saeonobspy: An Python package to query and downloaded environmental data from the SAEON observations database

• Ecological Monitoring and Management Application: An environmental data processing pipeline for forecasting satellite observed postfire vegetation recovery

My time is currently primarily devoted to these projects

• 🔥 Ecosystem Monitoring for Management Application Combining Earth observations in situ observations, and Bayesian ecological forecasting models to characterise vegetation state and predict postfire recovery in a highly biodiverse shrublands. We are producing an operational system to support the decisions of land mangers and help them identify ecosystem degradation from a range of causes
• 🌈 Mapping species and simulating virtual plants as part of NASAs first biodiversity focussed field camping, the Biodiversity Survey of the Cape. BioScape is collecting high resolution hyperspectral, thermal and Lidar data of the Cape Floristic Region to test the limits of what we can learn about diversity from space.

Want to connect?

• 💻 Get in touch to chat about projects with a geospatial, earth observation or biodiversity focus: [email protected]

• 📫 Follow me on Twitter as @glennwithtwons, or read a post on my website