Joseph Kelly

Old-Growth Labs

Biodiversity Conservation
Spatial Ecology
Remote Sensing

Curriculum Vitae
GitHub
Google Scholar
ORCID

I work at the intersection of biodiversity conservation, spatial ecology and remote sensing, building on sixteen years of experience in fieldwork and data analysis. I apply geographic information systems (GIS) and machine learning to produce maps and statistics for effective decision-making.

My past work has included bird diversity projects in Madagascar and Indonesia, great ape habituation in Gabon, human–elephant conflict in Cameroon and protected area delineation and management in Malaysia. In this case study, currently under review, I used satellite data and machine learning to predict alternative scenarios of forest intactness for a new tiger reserve.

Joe with members of Association Mitsinjo, Madagascar.

Joe stands at a high mountain pass, Ladakh, India.

Joe with a colleague in Loango National Park, Gabon.

If you would like to discuss research collaborations, professional opportunities or consultancy, please message me at: joe@oldgrowthlabs.com

Case study

How intact is a forest?

Figure 1. Aboveground biomass density (AGBD) and relative intactness on Mount Tapis, Pahang, Peninsular Malaysia. The upper slopes and ridges have lower AGBD (b), but score highly in relative intactness (c). Mg ha⁻¹ means tonnes per hectare.

Figure 2. From satellite data and machine learning regression to alternative forest scenarios. Changing forest cover and disturbance while holding terrain and location fixed allows the model to estimate how much biomass each site could support. Comparing this with its present state reveals the shortfall. TMF is tropical moist forest. Figure numbers within the diagram refer to the research article (Kelly et al., 2026b).