Postdoctoral Scholar, Terrain Analysis and Cryosphere Observation Lab
University of Washington
Remote sensing · Cryosphere · SAR
I'm a postdoctoral researcher at UW CEE's Terrain Analysis and Cryosphere Observation Lab, with a research focus on understanding when and where mountain snowmelt occurs — a critical question for the more than one billion people who depend on seasonal snowpack for freshwater. I graduated from UW with my PhD in December 2025, and my dissertation work combined satellite radar remote sensing with large-scale cloud computing to produce the first global, high-resolution record of snowmelt runoff onset timing, processing ~3.9 million Sentinel-1 C-band SAR scenes at 80-meter resolution from 2015 to 2024. This dataset enabled a systematic analysis of elevation, aspect, and temperature controls on snowmelt timing patterns across ~150 mountain ranges.
As a postdoc, I'm continuing to work on snowmelt phase delineation and wet snow detection, and I'm developing and evaluating snowmelt methods that leverage data from NASA's newly launched NISAR L-band SAR mission. I'm also building scalable, automatically-updating geospatial data pipelines using emerging tools like Icechunk and GitHub Actions. Alongside research, I develop and maintain open-source Python tools to make snow-related geospatial datasets more accessible to the broader community. I'm passionate about teaching — I've taught UW's graduate Geospatial Data Analysis in Python (CEE 467/CEWA 567) twice and TA'd often — and hope to eventually transition to a public high school teaching role.
A 10-year (2015–2024), 80-meter resolution global dataset of annual mountain snowmelt onset timing derived from ~3.9 million Sentinel-1 SAR scenes. Covers ~150 mountain ranges and is validated against 900+ SNOTEL/CCSS weather stations. The dataset reveals global patterns in snowmelt timing and their links to climate variability.
A global, 500-meter resolution snow phenology dataset derived from a decade (2015–2024) of MODIS 8-day maximum snow extent observations. Explore snow appearance, disappearance, and duration for any location and year, with cloud-gap filling and polar-region corrections applied.
A station archive and Python toolkit for snow point observations — snow water equivalent and snow depth from automated pillows, manual snow courses, aerial markers, and mirrored climate stations across nine networks in the United States, Canada, and Norway. CSV-first and automatically updating, with a live map of every daily-reporting station.
The research prototype behind the global dataset: Sentinel-1 C-band backscatter change-point analysis to detect snowmelt runoff onset across Western U.S. mountain ranges, and the first systematic evaluation of SAR-derived snowmelt timing against in-situ records.
Building a global-scale raster dataset with GitHub Actions for parallel compute and Icechunk for versioned, conflict-free Zarr storage. Hundreds of concurrent runners process MODIS land surface temperature across 376 tiles worldwide — 753 GB in about 40 minutes — visualized in an interactive web map.
A Python package that makes it easy to access snow-related geospatial datasets — including SNOTEL/CCSS station records, MODIS snow products, SAR-derived snowmelt timing, and more. Designed for researchers and practitioners working with snowpack data at scale.
A curated archive of SNOTEL and CCSS station metadata and daily records, with convenient Python access utilities for snow water equivalent and snowmelt validation studies. Superseded by Global snow networks, which extends the same approach to nine networks.
Tool for generating per-pixel local incidence angle maps for Sentinel-1 SAR imagery, useful for terrain correction in mountainous areas.
A collaborative SnowEx project for snow depth estimation from Sentinel-1 cross-polarization SAR. Developed during SnowEx HackWeek 2022 and continuing as an active open-source project.
Open-access course materials for CEE 467/CEWA 567 at UW, which I designed and teach. Covers vector and raster data processing, remote sensing, and cloud-based geospatial workflows in Python. New cohort each winter quarter.
Global patterns and controls on mountain snowmelt runoff onset from a decade of high-resolution observations
Identifying wet pixels for SAR-based SWE retrieval using model output and Sentinel-1 backscatter signals
Remotely sensed melt fraction enhances streamflow modeling in snow-dominated ungauged basins with long short-term memory networks
Hydrological Processes · preprint: 10.2139/ssrn.6675822
Accurate snow depth predictions across the Western U.S. using a deep learning model trained on 7 years of airborne lidar snow depth measurements
A global high-resolution dataset of snowmelt runoff onset timing from Sentinel-1 SAR, 2015–2024
Earth System Science Data, 18(8), 5871–5894 · 10.5194/essd-18-5871-2026
Breaker of images: Synthetic aperture radar and satellite iconoclasm
Environment and Planning F (online first) · 10.1177/26349825261456297
Feasibility mapping of L-band InSAR for SWE retrievals across the Western United States
Geophysical Research Letters, 53(10), e2025GL120162 · 10.1029/2025GL120162
When and where does water originate? Leveraging stable water isotopes and synthetic aperture radar to assess the hydrology of a snow-dominated watershed in southwestern Montana
Remote Sensing Applications: Society and Environment, 41, 101887 · 10.1016/j.rsase.2026.101887
Sentinel-1 SAR estimates of snowmelt onset coincide with SNOTEL soil moisture pulses across the Western United States
Hydrological Processes, 39(12), e70341 · 10.1002/hyp.70341
Evaluating snow depth retrievals from Sentinel-1 volume scattering over NASA SnowEx sites
The Cryosphere, 18(11), 5407–5430 · 10.5194/tc-18-5407-2024
Capturing the onset of mountain snowmelt runoff using satellite synthetic aperture radar
Geophysical Research Letters, 50(21), e2023GL105303 · 10.1029/2023GL105303
Characterizing and mapping volcanic flow deposits on Mount St. Helens via dual-band SAR imagery
Remote Sensing, 15(11), 2791 · 10.3390/rs15112791