Using HPCMP Resources for AI/ML Weather Intelligence Product Development for Smart Installation Decision Support at Ft. Moore

The vision of the Army Installations Strategy (AIS) and the Installations of the Future (IotF) program is to enhance mission effectiveness and resilience in a prudent, efficient, and forward-thinking manner, including Army installations and contingency bases, energy, and environmental programs.

Severe weather events introduce uncertainty and may hinder the installations’ capabilities of supporting the needs of warfighters and civilian personnel. The integration of real-time meteorological data along with customized high-fidelity Numerical Weather Prediction (NWP) forecasts, enhanced information on road conditions, facility energy usage, personnel location, and training information will allow installations to better prepare for weather events. As part of the IoTF program, the Smart Base project looks to integrate a variety of disparate data sets together with AI/ML algorithms into a user-friendly dashboard with actionable insights for decision makers at many levels.

To accommodate the volume and variety of data collected, as well as provide the necessary compute power required for running NWP models, the team utilizes the Department of Defense High Performance Computing (DOD HPC) capabilities. Incorporating the use of the DOD HPC in this framework demonstrates the use of existing infrastructure to support systemic improvements to installation operations and procedures impacted by weather. The DOD HPC platform provides a common computational platform for development, making it easy to share code and large datasets with multiple project partners from a variety of military and civilian institutions. DOD HPC platform features such as the ability to parallelize code and store large datasets provide the key ingredients necessary for satisfying stakeholder needs and requirements on projects.

This presentation will provide the framework of how to utilize data, models, and analytics as guidance for decision makers to make highly informed weather calls at installations. An evaluation of the Weather Research Forecast (WRF) model configured over the Ft. Moore domain will also be discussed.

PRESENTER

Hoch, Brendon
brendon.hoch@erdc.dren.mil
603-646-4295

US Army Corps of Engineers Engineering Research Development Center

CO-AUTHORS

Hoch, Brendon
brendon.hoch@erdc.dren.mil

Bastian, Hyeyon
hyeyon.bastian@westpoint.edu

Obiako, Ifezue
Ifezue.V.Obiako@erdc.dren.mil

Gallarno, George E.
George.E.Gallarno@erdc.dren.mil

Buchanan, Randy
Randy.K.Buchanan@erdc.dren.mil

Richards, John
John.P.Richards@erdc.dren.mil

Rinaudo, Christina
Christina.H.Rinaudo@erdc.dren.mil

CATEGORY

Artificial Intelligence / Machine Learning usage for HPC Applications

SYSTEMS USED

Onyx, Carpenter

SECRET

No