Building on our work linking harmful algal blooms (HABs) to health outcomes, this project uses satellite imagery to improve the early detection of HABs and enable targeted early warning systems for coastal communities. We exploit patterns in remotely sensed variables, such as sea surface temperature, chlorophyll levels, and surface reflectance, to train a convolutional neural network (CNN) to identify blooms as they emerge. This approach allows for more timely, location-specific alerts that go beyond the broad seasonal warnings currently in place. Our aim is to support faster public health and environmental responses in coastal areas where monitoring resources are limited, with a system that can be adapted to similar regions worldwide.