Exploiting Heterogeneous Monitoring Data for Spatiotemporal Algal Bloom Prediction

Harmful algal blooms need to be mitigated because they can cause significant negative effects to humans and other organisms. If such algal blooms can be predicted in advance by monitoring water quality, they can be suppressed at an early stage by making decisions to take actions. We describe our ongoing work on integrating the heterogeneous water quality monitoring data and on recovering the missing data using tensor completion techniques. We also discuss the challenges in carrying out this study.