In order to evaluate human impacts on biodiversity at the species and ecosystem scales, my research focuses on identifying and predicting the occurrence of emerging land uses/processes. I have been particularly interested by habitat loss for agricultural production, including predicting deforestation and biodiversity impacts of oil palm plantation expansion globally and in a subsequent regional study in the Peruvian Amazon.

I am also the developer of a course and code repository on remote sensing and spatial analysis for SES systems in Google Earth Engine: GEEforSynthesis.

More recently, I have begun developing dynamic, predictive models of fire risk (currently focused on the Brazilian state of Mato Grosso), using a combination of biotic, abiotic, and socio-economic predictors.

My work at SESYNC examined potential tradeoffs between conservation and food production through the lens of the global protected area network. I use regression and machine learning models to understand and predict cropland occurrence in protected areas using remotely sensed data and coincident socio-ecological predictors, including determinants of food security. The first globabl analysis on this topic found here and forthcoming.

I work to develop techniques for assessing species populations and distributions by incorporating remotely sensed data on land use, citizen science observations, and population modeling. One example of this is work I co-led to assess the distribution and population of cheetah in Southern Africa. I am currently working to understand how differences in habitat utilization influence vertebrate responses to land use conversions for agricultural production.