Ross Mower

Ross Mower — Machine Learning Engineer & Earth Systems Researcher

Building scalable AI systems for Earth observation, scientific discovery, and environmental prediction.

I specialize in machine learning, scientific software, high-performance computing, remote sensing, and open source — building systems that turn satellite and sensor data into operational environmental predictions.

View ProjectsResume GitHubPublicationsContact
A subalpine ridge in the North Cascades with receding snowfields, a hiking trail, and blooming glacier lilies in the foreground

Featured Projects

All projects
Parallel SnowModel
A distributed-memory rewrite of a snow-evolution model, scaled to 1,800 HPC cores to produce a 21-year, 100m-resolution snow record for the continental United States.
HPCFortranMPICoarray FortranDistributed SystemsGeospatial
Learn more Code Paper
Distributed Snow Prediction Platform
An operational ML platform fusing in-situ observations, remote sensing, and physically based models into near-real-time snow forecasts for federal water managers.
Production MLData PipelinesRemote SensingFull-StackAWS
Learn more Live
Station Optimization
Multimodal Vision Transformer architectures fusing high-resolution terrain, lidar, and satellite data to optimize where environmental monitoring stations should go.
Vision TransformersMultimodal MLPyTorchDistributed TrainingGeospatial
Learn more
Scalable Sentinel-1 Processing Framework
An automated pipeline for Sentinel-1 SAR acquisition, quality control, and time-series analysis, built to characterize snowmelt progression across space and time.
Remote SensingSARData PipelinesPython
Learn more

Scalable AI Systems

Distributed training and production ML pipelines, from MPI/Coarray Fortran at 1,800-core scale to automated daily inference in operational systems.

Earth Observation & Geospatial ML

Multimodal fusion of SAR, lidar, MODIS, and physically based models into architectures that predict environmental conditions at continental scale.

Software Engineering First

Owns platforms end-to-end — ingestion, training, inference, monitoring, and deployment — not just the modeling code.

HPC cores scaled to
1,800

HPC cores scaled to

snow simulation output
150 TB

snow simulation output

continental-scale record
21 yrs

continental-scale record

federal agencies in production
3

federal agencies in production

Operational systems supporting NASA · Bureau of Reclamation · Natural Resources Conservation Service

Recent Publications

All publications
Evaluating Scalable Sentinel-1 Melt Progression Across Space and Time in a Warm Alpine Snowpack
In review
Mower, R., Pflug, J., Gagliano, E., Gutmann, E., Cristea, N., Lundquist, J.

Water Resources Research

Tests how well an automated Sentinel-1 SAR processing pipeline tracks snowmelt timing across a warm, high-elevation alpine snowpack.

Snowscapes of fear: Snow conditions mediate interactions between apex- and meso-predators
In review
Ganz, T., Mower, R., Sullender, B., Lundquist, D., Cunningham, C., Kertson, B., Wirsing, A., Prugh, L.

Journal of Animal Ecology

Examines how snow conditions shape encounter and avoidance dynamics between apex and mesopredators.

Apex predators exploit advantageous snow conditions across hunting modes
October 2025
Sullender, B., Ganz, T., Mower, R., Lundquist, J.D., Kertson, B., Newsome, T., Wirsing, A., Prugh, L.

Journal of Animal Ecology

Shows how apex predators use varying snow conditions to their advantage across different hunting strategies.

Let's talk

Open to machine learning engineering, research engineering, and scientific software roles.

ContactEmail