Thursday, November 18, 2021 4:00pm-5:00pm https://nmsu.zoom.us/j/98051466351

Deep Echo State Neural Network for Light Propagation Through Atmospheric Turbulence

A machine learning model using a Deep Echo State Neural Network was developed and is being implemented for various types of air turbulence.  This network can also be used for other time dependent two-dimensional data (e.g., weather radar, seismography, fluid flow, etc.).

The work was done during two successive Summer Faculty Fellowships with the Air Force Research Laboratory (AFRL) both in Maui, HI and in Albuquerque, NM. During this presentation I will describe the network, the performance criteria and the optimization and validation processes. I will show examples with two types of turbulence: Long-distance atmospheric turbulence (>100km) and supersonic compressible turbulence. At the conclusion of the talk, I will go over the future work still needed in the field as well as how students from physics and engineering (both graduate and undergraduate) can be involved with this research.