Skip to page content

Offshore Precipitation Capability

The FAA's non-radar weather radar


Offshore Precipitation Capability is a radar-like severe-weather presentation using satellite imagery and lightning sensors.

At the heart of the software is a machine-learning that analyzes huge volumes of weather data.

Forecasters in Puerto Rico used the tool in fall 2017 after Hurricane Maria destroyed the island's Next Generation Radar site.

The FAA has discovered a way to provide controllers actionable weather intelligence without radar. The breakthrough, made possible by passive satellite sensors, big data, machine learning, and automation, will enable controllers for the first time to deploy weather constraints and no-go zones in oceanic and remote areas where weather radar is not available.

Called the Offshore Precipitation Capability (OPC), the weather radar proxy was developed by the FAA's Aviation Weather Division and MIT Lincoln Laboratory (MITLL) to give controllers, airline dispatchers and planners a view of the internal makeup of weather systems in locations where Next Generation Radar (NEXRAD) is not available. The capability will eventually be embedded into trajectory tools as part of the FAA's NextGen Weather Processor.

The tool estimates precipitation location and intensity for storms outside of radar coverage by using machine-learning techniques and data from nearby NEXRAD sites, visual and infrared imagery from geostationary weather satellites, and global lightning-detection system data.

satellite to NEXRAD to OPC comparison

Hurricane Harvey in the Gulf of Mexico on August 24, 2017

Machine learning is used to create weather "models" by analyzing huge volumes of historical data from radar and non-radar sources, and identifying the links between the two. "We learn the relationship and use it to create a weather picture outside of the radar areas," said Haig Iskenderian, an MITLL technical staffer on the program. Along with a snapshot of current weather conditions, OPC in the future may be able to provide a 12-hour forecast as well.

The FAA has been operating OPC in demonstration mode as a situational awareness aid for controllers at the Houston, Miami, Puerto Rico and New York en route centers, and at the Command Center in Warrenton, Virginia. While not a substitute for NEXRAD, OPC provides a weather picture that goes beyond the visual and infrared imagery of satellites by generating six color-coded precipitation intensity levels within clouds — light green for the most benign to purple for the most severe.

Drawbacks of using a modeled version of radar include uncertainty in the precise location and intensity of precipitation. Air traffic controllers, however, say it is operationally accurate enough to help, and it is a major upgrade to what they had before to watch oceanic storms — simple satellite imagery of clouds.

An unanticipated proving run occurred in fall 2017. The National Weather Service pressed OPC into early service when Hurricane Maria's winds destroyed the NEXRAD site in San Juan, Puerto Rico. Forecasters there tapped into OPC over the internet and used a combination of other satellite-based and ground-sensor systems to continue providing weather advisories to residents.

During hurricanes Irma and Maria, Federal Emergency Management Agency officials in the National Response Coordination Center and regional offices used OPC to gain an awareness of offshore precipitation and weather system movements.

Randy Bass in front of weather data on a computer screen.

Randy Bass, a meteorologist in the FAA's Aviation Weather Research Program, in front of an Offshore Precipitation Capability image of hurricanes Irma (center) and Jose (right).

Along with emergency responders, the U.S. Air Force has taken a keen interest in the program for worldwide operations. Officials are funding the development of a global version of the tool as well as the 12-hour forecast capability, an enhancement that also will be available to the FAA and other users.

Back to top

Page last modified:

This page was originally published at: