Seminar: Earth Observation Foundation Models

This is a joint hybrid event by the IEEE Computer Society (CS) ACT Chapter, the IEEE Geoscience and Remote Sensing Society (GRSS) ACT&NSW Joint Chapter, and the Canberra Data Scientists (CDS) Meetup. After the talk, there will be free pizzas and soft drinks provided to encourage people to stay after the presentation and socialise with others. RSVP is required, please following instructions below for registration.

Title: Advancing Spatiotemporal Ecohydrological Monitoring with Earth Observation Foundation Models: Concepts, Progress and Challenges
Speaker: Dr Yi Yu, Associate Lecturer, The University of Sydney
Hosts:
- Yiqing Guo (IEEE GRSS)
- Warren Jin (IEEE CS)
- Yanchang Zhao (CDS)
Date: Thursday 9 July
Times:

• 3:30-4:30pm - Presentation
• 4:30-5:30pm - Food/Networking

Venue: Stringybark Room, Ground Floor of Synergy Building (B801, corner of North Science Road and Dickson way), CSIRO Black Mountain Science and Innovation Park, Acton ACT
Event Sponsor: SHURA

RSVP for In-Person Attendees: Please register here if you are attending in person: Attendance Sheet for In-Person Attendees. To assist in catering, please register by 6pm Tuesday 7 July 2026. Parking information is also provided at the link.
Sign in at the Synergy Building: When you arrive at the Synergy Building foyer, please sign in using the iPad at the counter. Enter 'Yiqing Guo,' 'Warren Jin' or 'Yanchang Zhao' as the person you are visiting. Attach the printed name tag to your chest and wait to be collected. NO tailgating, please.
Sign out: Please ask one of the event organizers or a CSIRO helper to assist you with leaving the building and signing yourself off at the counter.
RSVP for Online Attendees: Please register here if you are attending online: Attendance Sheet for Online Attendees. A Teams meeting link will be sent to your email before the event.

Abstract: Ecohydrology seeks to understand how water, energy, vegetation and carbon processes interact across space and time. With the rapid growth of satellite archives, environmental sensor networks and data-driven modelling, deep learning (DL) has become increasingly useful for estimating ecohydrological variables such as soil moisture, land surface temperature, evapotranspiration, vegetation water content, gross primary productivity and solar-induced chlorophyll fluorescence. These developments have made large-scale ecohydrological monitoring increasingly feasible, especially where direct observations are sparse. However, most DL applications remain strongly dependent on pattern recognition from the training data. This poses risks for ecohydrology, where target processes are highly dynamic, physically constrained and often non-stationary under climate variability, land-use change and extreme events. A purely data-driven model may perform well within the statistical envelope of its training data but become unreliable under unseen conditions. More importantly, ecohydrology requires not only prediction of what is happening, but also understanding of why it is happening through attribution, physical consistency and causal interpretation. Earth observation foundation models (EOFMs) offer a promising opportunity to refine spatiotemporal ecohydrological estimation by learning reusable representations from large, heterogeneous EO archives. Yet their role should not be viewed simply as larger-scale prediction. A key challenge is how EOFMs can be elegantly constrained by physical principles and integrated with process understanding. This talk will discuss recent progress in spatial, multimodal and temporal EOFMs, and examine how they may support physics-constrained estimation, uncertainty propagation, cross-sensor transfer and domain-specific benchmarking. Particular attention will be given to the need to distinguish lower-level EO-derived variables with stronger physical links from higher-level ecohydrological products built from multiple layers of proxies, assumptions and model constraints.

Bio: Dr Yi Yu received a Ph.D. degree in hydrology and remote sensing from The Australian National University, Canberra, ACT, Australia. He is currently an Associate Lecturer with The University of Sydney, Sydney, NSW, Australia. His research interests focus on developing hybrid approaches that integrate physical and data-driven methods to better understand land–atmosphere interactions across various spatiotemporal scales, particularly in the context of climate extremes such as drought. Yi was a recipient of the Commonwealth Scientific and Industrial Research Organisation (CSIRO) Agriculture and Food Director’s Award—Next Generation Science in 2023. He was among the Student Paper Competition Finalists at the 2024 IEEE International Geoscience and Remote Sensing Symposium (IGARSS).