Darema F., Blasch E., Ravela S., Aved A. (eds) Dynamic Data Driven Applications Systems. DDDAS 2020. Lecture Notes in Computer Science, vol 12312. Springer, Cham. https://doi.org/10.1007/978-3-030-61725-7_36
Convective weather events pose a challenge to the burgeoning low altitude aviation industry. Small aircraft are sensitive to winds and precipitation, but the uncertainty associated with forecasting and the frequency with which impactful weather occurs require an active detect and response system. In this paper, we propose a dynamic, data-driven decision support system, with components of forecasting, realtime sensor observations, and route planning. We demonstrate our technology in the Dallas/Fort Worth metroplex, a large urban area with frequent thunderstorms which hosts the CASA Doppler radar network.
The high temporal and spatial resolution data provided by this network allows us to quickly and accurately identify ongoing meteorological hazards for flight planning purposes. Rapidly updating short term (0–90 min) forecast data are generated with features extracted as obstacles to avoid. A flight path generator submits requests for path routing which include randomized start and end locations and times, weather tolerance parameters, and buffer zones. A customized obstacle course is created and used as the basis for routing. Weather processing workflows are instantiated with Mobius, a multicloud provisioning system. The Pegasus Workflow Management System orchestrates processing via scalable workload distribution to compute resources. Sensor data is transmitted and processed in real time, and routes are periodically calculated for proposed flights. A Google Maps front-end interface displays the weather features and flight paths. Herein, we focus on the overall system design, with particular emphasis on the dynamic flexibility and interoperability that our architecture allows.