Data Assimilation
Accuracy and utility of models are often greatly improved by incorporating real-time observations. Assimilation of sensor data can reduce error, increase resolution, and improve timeliness, particularly when observations are continually updated. However, accounting for multi-scale dynamics, measurement variability, and model error require expertise of underlying mathematical models and approaches that preserve uncertainty along with these gains. GDA has leveraged our expertise in data assimilation in localized wind forecasts, ocean current and wave height measurements, and agricultural spray drift applications.
Combining ocean current information with wind and wave data, this diagram shows a comparison of predicted ocean float drift vs. actual float trajectories (DARPA Ocean of Things Phase II - 2022)