Ground Weather Radar Signal Characterization through Application of Convolutional Neural Networks
Abstract
The 45th Weather Squadron supports the space launch efforts out of the Kennedy Space Center and Cape Canaveral Air Force Station for the Department of Defense, NASA, and commercial customers through weather assessments. Their assessment of the Lightning Launch Commit Criteria (LLCC) for avoidance of natural and rocket triggered lightning to launch vehicles is critical in approving space shuttle and rocket launches. The LLCC includes standards for cloud formations, which requires proper cloud identification and characterization methods. Accurate reflectivity measurements for ground weather radar are important to meet the LLCC for rocket triggered lightning. Current linear interpolation methods for ground weather radar gaps result in over-smoothing of the vertical gradient and over-estimate the risk of rocket triggered lightning, potentially resulting in costly, unnecessarily delayed launches. This research explores the application of existing interpolation methods using convolutional neural networks to perform two-dimensional image interpolation, called inpainting, into the three-dimensional weather radar scan domain. Results demonstrate that convolutional neural networks can improve the accuracy of cloud characterization over current interpolation methods, potentially resulting in fewer launch delays with substantial associated cost savings due to increased capability to meet the LLCC.
Document Details
- Document Type
- Technical Report
- Publication Date
- Mar 26, 2020
- Accession Number
- AD1101491
Entities
People
- Stephen M. Lee
Organizations
- Air Force Institute of Technology