Sources/Usage: Public Domain. View Media Details

Data documenting the planet’s health now stream to receiving stations every hour of every day, from optical, radar, lidar, and thermal sensors that grow more sophisticated by the year. Even the breadbox-sized cubesats sometimes launched by the hundreds now collect data across multiple bands of the electromagnetic spectrum—a capability once reserved for large-scale legacy satellite systems like Landsat.

All that data can help us understand the Earth system, detect change in near real time, and forecast future change. First, however, those data need to be trustworthy, interoperable and optimized for the machine learning algorithms now used to scan for and forecast change. 

That’s why gatherings such as next week’s Joint Agency Commercial Imagery Evaluation (JACIE) workshop are so critical to the study and understanding of our planet.

The 2022 JACIE workshop, a free event set for Jan. 10-13, 2022, will see a host of experts in satellite data calibration and validation presenting their findings and working through important questions about how to best realize the potential of the modern remote sensing community’s data riches.

It’s a discussion space quite familiar to teams at the USGS Earth Resources Observation and Science (EROS) Center. The USGS was a part of the first JACIE 20 years ago, and took the lead of the group in 2003. Through workshops like JACIE and contributions to similar gatherings throughout each year, the USGS works to set the bar for data quality and to guide the development of cross-platform standards for interoperable Analysis Ready Data (ARD).

Only a handful of commercial Earth observation systems existed at the turn of the century. The first JACIE conference, in fact, saw attendees tearing into every detail of just two commercial satellites. Today, according to Greg Stensaas of EROS, “we can’t even touch the surface of all the data that’s coming in. We have a hard time just trying to hit the highlights in these workshops.”

Sources/Usage: Public Domain. View Media Details

“The good thing is that there are systems fulfilling the requirements of the user groups in a way we’ve never been able to do before,” Stensaas said, referring to the emergence of daily high-resolution data and expanded efforts to use them alongside data from satellites like Landsat to improve landscape monitoring across the board.

JACIE is now one of three global assemblies on ARD, data quality, and interoperability that Stensaas and his colleagues look forward to each year. EROS representatives were among the organizers of the ARD21 Satellite Data Interoperability Workshop, for example, a conference led by the commercial sector that marked its fourth year in 2021 with a 4-day virtual conference in late October. Each JACIE workshop now dovetails with those ARD conferences, and many of the same players meet again each spring for the European Space Agency’s Very High-resolution Radar & Optical Data Assessment (VH RODA) workshop.

“These conferences are all trying to get at the same question: ‘how do we use all the datasets that are coming in now?’” said Jon Christopherson, a contractor at EROS who leads JACIE efforts and co-authors the annual JACIE “Land Remote Sensing Satellite Compendium.”

“We’re working to try and coordinate efforts across agencies and internationally,” he said, leaning on and learning from the ARD efforts of the commercial and civilian sectors to work toward a set of interoperability guidelines that will advance the science of Earth observation.

It’s no small task. One presenter at ARD21 led his talk by pointing out that there are

“trillions of pixels” of data available now, making the development of reliable algorithms to scan for change a necessity. The fruits of that labor are already beginning to emerge. A few weeks before ARD21, United Nations University unveiled its World Flood Mapping Tool, which uses satellite data and cloud computing resources to produce street-level maps of flooding nearly anywhere in the world, from 1985 to the present, with the click of mouse.

Gatherings like JACIE, ARD21, and VH RODA, Stensaas said, help data managers stay responsive to the needs of the scientists building such applications, and to optimize their systems for the machine learning approaches that underpin them.

“We’ve established a set of actions and paths forward,” he said. “Things that came out of ARD will be discussed at JACIE, things from JACIE will be picked up at VH RODA.”

Click here to learn more about the 2022 JACIE workshop. Click here to register.