Atmospheric Science & Computer Science Student at Cornell University
I am an atmospheric science and computer science student at Cornell University, specializing in mountain meteorology and data science. With a strong passion for leveraging advanced machine learning techniques to tackle complex meteorological challenges, I aim to bridge the gap between data science and atmospheric phenomena.
WindBorne Systems · Contract · May 2024 - Present (1 yr)
Developed and operationalized Pointy, a machine learning-based point forecast system. Led point-forecast training data proliferation, optimized model training & inference scripts, constructed end-to-end ML data pipeline generating Pointy forecasts using WeatherMesh inputs, wrote API to access Pointy data.
PowderChasers · Part-time · Jun 2022 - Present (2 yrs 11 mos)
Analyzed and interpreted complex model data from various global and mesoscale deterministic and ensemble forecast models. Developed forecast algorithms and wrote forecasts for ski resorts in North and South America several times per week. Negotiated and managing partnership with Powder Magazine since January 2024.
University of Washington · Part-time · May 2022 - Present (3 yrs)
Analyzed climate change’s impact on seasonal snowfall and snowpack during different synoptic regimes in the Pacific Northwest. Investigating the effect of evaporative cooling processes from the Eastern Pacific on mid-latitude cyclone temperatures and precipitation type on a climatological timescale.
University of Colorado Boulder · Part-time · May 2022 - Oct 2022 (6 mos)
Analyzed Doppler radar data from the 2022 WINTRE-MIX field campaign in Quebec to study freezing rain processes, integrating surface observations and model data for improved precipitation forecasting. Presented findings at the Mountain Meteorology Conference 2022 and co-authored an upcoming paper. Assessed cloud seeding efficacy during the 2017 SNOWIE field campaign in Idaho through vertical profile analysis of radar and surface data. Collaborated closely with Dr. Katja Friedrich and her research team on both projects.
OpenSnow · Part-time · Aug 2021 - Present (3 yrs 9 mos)
Lead developer for the HARPNET machine learning-based precipitation downscaling system, expected to integrate into METEOS before the 2024-2025 season.
SnowBrains · Part-time · May 2020 - Present (5 yrs)
Write articles and ski forecasts for mountainous regions around the globe.
HARPNET is a next-generation machine learning precipitation downscaling system designed to improve the resolution of weather models that struggle to capture fine terrain influences. Utilizing an attention-gated residual convolution UNet architecture, HARPNET dynamically downscales precipitation with high efficiency and spatial continuity, opening doors for more accurate weather forecasting and future enhancements in mesoscale modeling.
HARPNET Project Page View HARPNET on GitHubThis project is a tool to access data from an 80-member super ensemble comprised of 30 GEFS members and 50 EPS members with a focus on high-precision snowfall prediction.
View on GitHub