5G Drones: Real Time Data Assimilation to Transform Wildfire Predictability

Under a DOE Office of Science funded project, scientists at Los Alamos National Laboratory (LANL) and New Mexico Technical college (NM Tech) are developing a new drone mounted 5G smoke sensor and a fast-running surrogate modeling framework to incorporate and use data from this sensor. The 5G sensor uses multiple wavelengths interference to back out black carbon (BC) amounts and will be tested in both the laboratory and the field. Laboratory experiments will be undertaken within LANL’s Center for Aerosol Forensic Experiments in which a known amount of BC is introduced into a chamber and compared against measurements from the sensor. Field experiments using the sensor mounted on a drone will initially involve measurements from controlled burns undertaken at a firefighting training school at NM Tech and later controlled burns planned by the Canadian Forest Service. In addition to the 5G sensor, the drone or drones will also have an optical sensor that measures aerosol and/or BC properties as well as a thermal imagining camera to quantify fire activity.

The data from the drone or drones from the planned Canadian controlled burns will be used in the development of a fast-running data-driven surrogate model for fire spread and behavior. The surrogate model is currently being developed using simulations from LANL’s fire behavior model, HIGRAD-FIRETEC. HIGRAD-FIRETEC is a physics-based CFD code developed over the past 30 years to model wildfire spread and smoke formation and transport for a variety of fuel conditions. And, unlike empirical based spread models, e.g., Rothermal, fire spread and particulate formation within FIRETEC is based on the combustion properties of the fuel, radiation transport, and spotting. The basic underlying dynamical properties of FIRETEC will be used indirectly to develop our surrogate model through the use of simulated data. Once drone data is obtained from the Canadian burns and HIGRAD-FIRETEC simulations undertaken, the impact of the drone data on improving surrogate model fidelity will be assessed. In this presentation, an overview of the new 5G sensor drone mounted sensor will be presented along with details concerning the surrogate model and how data (from HIGRAD-FIRETEC and drones) improves predictability of the fast-running tool.