Sunday, July 22, 2018

New DARPA Program Aims to Bring Precision of Terrestrial Weather Forecasting to Near-Earth Space Environment

An aurora, visible at the Earth's poles as a result of a geomagnetic storm, can interfere with satellite operations. DARPA’s Space Environment Exploitation program aims to develop precision forecasting of a host of disturbances in the near-Earth space environment.

Models for providing hourly terrestrial weather forecasts anywhere in the world have become increasingly precise—our smartphones buzz or chirp with local alerts of approaching thunderstorms, heavy snow, flash floods, and big events like tornadoes and hurricanes. The military relies on accurate weather forecasts for planning complex operations in the air, on ground, and at sea.

But when it comes to predicting environmental conditions in specific locations within the vast volume of space, no similar forecasting exists. As space launch companies make access to space more affordable and constellations of low-Earth orbit small satellites continue to grow, military and commercial space operators need new tools to track space environmental conditions and their potential impact.

DARPA’s new Space Environment Exploitation (SEE) program aims to accurately predict near-Earth space environment disturbances and perturbations (scales as small as 100 kilometers in size) in one-hour increments extending out 72 hours. DARPA is hosting a Proposers Day in Arlington, Virginia, on July 31, 2018. The event also will be webcast. More details are available here:

“We currently have capability to predict and track big space weather events like sun spots, coronal mass ejections, or solar winds that can wreak havoc on critical space assets in higher space orbits,” said Air Force Maj. Charlton “David” Lewis, II, DARPA program manager. “But we can’t make one-hour to 72-hour predictions of local, smaller space environment disturbances close to the Earth, from magnetic substorms to auroral-E, which can interfere with a host of ground and space-based assets.”

Lewis likened the capabilities sought in the SEE program to the current 72-hour operational planning cycle for an Air Force Air Tasking Order.

“Current operational preparation of the battlespace includes terrestrial weather forecasts for the next three days,” Lewis said. “Our goal is to provide the same fidelity in space environment forecasts that would be operationally and tactically relevant to a commander.”

The program includes two technical challenge areas. The first is to create a holistic space environment physics model using modern theory and leading-edge computational standards. Current models do not consider all of the physics we currently have the capability of exploring, Lewis said. With the advent of highly capable graphics processing units (GPUs) and tensor processing units (TPUs), a much higher-resolution physics model could be developed and computed.

The second area is increasing the number of measurement samples in the large volumetric near-Earth space environment. Currently, this space environment is severely under sampled.

“One way to get measurements is looking up into space from Earth, but ideally you want to be taking your measurements directly in the space environment,” Lewis said. “Right now that requires expensive satellite launches or piggybacking off other satellites.”

To address the challenge, SEE will explore novel, affordable concepts to unify and leverage elements of existing ground- and satellite-based data to make space environment predictions, while developing non-traditional sensors and sensing concepts. Relying on existing ground sensors and satellites alone, however, cannot yield an adequate number of samples for forecasting precision, Lewis said. The program will look at advanced machine learning and machine-training methods to extend the life of satellite measurements virtually, for example, and even increase the number of satellites virtually to greatly increase sampling size.

DARPA is seeking expertise in the following technical areas for the program: magnetospheric, ionospheric, and thermospheric physicists/chemists; machine learning/machine-training specialists; scalar, vector, and tensor processing computer scientists and engineers; and system integrators.

If SEE is successful, it may become routine for space planners and operators to get timely updates like this: “Satellite Weather Alert! Severe geomagnetic storm expected in one hour for the next 12 hours over the polar region producing ionospheric disturbances affecting LEO satellites between 100 – 175km above the earth.”


No comments:

Post a Comment