Sustainability &
Data science

current work

machine learning human footprint

Using existing data on global HFI, cloud-free satellite data from Landsat, and deep learning (with a convolutional neural network), we have developed the most up-to-date map of the human footprint index. The manuscript is published at Environmental Research Letters.

Topic modeling for Future scenarios

We use computational text analysis, including latent Dirichlet analysis (LDA) to identify latent themes and keywords across large corpuses of texts. These approaches can assist in distilling thematic content from natural language. We have used this approach for a corpus of Arctic news on the future, and on a corpus of scholarly Abstracts on human modification of the atmospheric water cycle.