Jace Innis

Marine data science portfolio focused on harmful algal bloom (HAB) forecasting, computer vision, and oceanographic analysis.

Neural network training animation for harmful algal bloom prediction
Deep Learning • Time Series

Harmful Algal Bloom Early Warning System

An LSTM sequential neural network pipeline predicting shellfish toxicity and red tide outbreaks in the Philippines using multi-decade monitoring bulletins paired with Google Earth Engine oceanographic data.

Classified microscope image of ocean particles from a 2013 sample
Computer Vision • Segmentation

Ocean Flux Particle Classification

An instance-segmentation pipeline using YOLOv8 to classify sinking marine particles in the Oceanic Flux Program archive, turning microscope images into reproducible measurements of particle abundance and composition.

Explore Data Visualizations

View animated ocean current trajectories, Google Earth Engine environmental rasters, and spatial maps.

Open Visualization Gallery →