• •Thesis: "Propagating observational uncertainty through land surface classification at regional scale".
• •Examined without corrections. Funded by a competitive doctoral training partnership studentship.

• •Lead the uncertainty quantification workstream of a multi-institution project on land use change detection, working with two partner universities and a national agency.
• •Developed an open Bayesian hierarchical framework for propagating sensor uncertainty into classification outputs, released as an R package with documentation and test data.
• •Initiated a collaboration with a hydrology group at a partner institution that has produced two co-authored papers outside the original project scope.
• •Contributed the methods section and data management plan to a successful £480,000 council grant application, and am named co-investigator.
• •Review for three journals in the field, typically eight to ten manuscripts a year.

• •Led practical classes in spatial analysis and statistical computing for undergraduate cohorts of 40 to 70.
• •Marked coursework and examinations across three modules, with moderation responsibilities in the final year.

Accepted, in press.

• •£480,000 over three years, funded by a national research council; three partner institutions.
• •Contributed the methods section and data management plan to the successful application.
• •Lead the uncertainty quantification workstream and line-manage one research assistant.

• •£18,000 internal seed award to test how land managers interpret probabilistic maps.
• •Wrote the application, obtained ethics approval and ran the study; produced one accepted publication.

Invited talk on communicating classification uncertainty to non-specialists.

Session on remote sensing uncertainty.

Invited methods tutorial on hierarchical models.

• •Convene and teach a master’s module of approximately 60 students, including lecture design, assessment and moderation.
• •Redesigned the assessment from a written examination to a reproducible analysis portfolio, which raised submission of working code from around half the cohort to almost all of it.

• •Co-supervised two doctoral students to completion, both examined without major corrections.
• •Primary supervisor for nine master’s dissertations across three cohorts; two were developed into conference papers.

Competitive fully funded studentship covering fees and stipend.