Please join CPAESS for our monthly virtual seminar series with October presenter Graham Taylor, former CPAESS Postdoctoral Researcher at NOAA Geophysical Fluid Dynamics Laboratory (GFDL)
WHEN: Wednesday, October 15, 2025 at 11:00am MT (Virtual)
PRESENTER: Graham Taylor, former CPAESS Postdoctoral Researcher at NOAA Geophysical Fluid Dynamics Laboratory (GFDL)
TITLE: The Role of Observation-Based Data in Local Climate Projection Uncertainty
WATCH LIVE WEBCAST
View The Seminar Flyer
BIO
Graham Taylor is a former CPAESS Postdoctoral Scholar at NOAA GFDL, where he worked on evaluating methods for translating coarse-resolution climate models into actionable data, focused on the question "What methods are fit for what purpose?". Prior to that, Graham received his PhD from Portland State University, focused on atmospheric circulation, climate change, and wildfire weather.
SUMMARY
High-resolution climate data and downscaling methods provide local-scale projections, but their precision can mask hidden uncertainties. This work assesses how uncertainties in commonly used observation-based gridded datasets propagate through three widely adopted statistical downscaling workflows (LOCA2, STAR-ESDM, and NEX-GDDP), using the Puget Sound region of Washington State as a case study. First, we compare observation-based data to five long-term station records of winter "Frost Days" and find evidence of substantial biases at individual sites. Next, we demonstrate that these biases are inherited by downscaled climate model historical simulations. Finally, we extend our analysis to projections under SSP5‑8.5 through 2100 and map the spread in signal across downscaling methods.
The talk will be recorded and published on the CPAESS YouTube Channel.
Questions? Contact Dawn Mullally
Find out more at the CPAESS Discovery Seminar Series
------------------------------
Alex Meyer
Graphic Artist II, CPAESS
UCAR, UCP and NSF NCAR
------------------------------