From Miles Above, Satellite Data Helps Spot Hazardous Trees and Reduce Wildfire Risk

Preventing ignitions that can cause catastrophic wildfires requires focus, vigilance and the integration of ever-evolving technology. And as PG&E continues to make progress – including what the company has calculated to be a 72% reduction of ignitions in high fire-risk areas in 2023 compared to the three-year average – satellite technology plays an increasing role.

That’s true with weather-forecasting models and it’s true with vegetation management, how a utility with approximately 70,000-square-mile service area in Northern and Central California works to keep millions of trees and limbs away from powerlines.

One example is PG&E’s partnership with Planet Labs PBC (NYSE: PL), a San Francisco-based provider of daily data and insights about Earth. PG&E leverages Planet’s satellite-derived data on vegetation including canopy height, cover and proximity to electric-system infrastructure to prioritize the mitigation of vegetation associated risks. The companies recently expanded their work together with a year-long contract that will give PG&E access to weekly Basemaps from Planet, to be used alongside the vegetation data already being leveraged.

“Planet’s tree data plays an important role as we analyze conditions and changes across our service area,” said Andy Abranches, PG&E’s senior director for Wildfire Preparedness and Operations. “We’ve used it to improve our data set of tree density around powerlines, and it has been beneficial as we’ve studied which trees in which locations caused outages during the winter storms last year and this year."

Going forward, Abranches said, data showing which trees have the potential to strike powerlines is being added to the next version of PG&E’s own vegetation management risk-modeling.

“This data provides a more comprehensive layer of information, highlighting the relative vegetation situation and condition within designated areas. We’re currently proving the concept and assessing the potential benefits of machine-learning-based tree species identification and dead or dying trees detection models,” he said.

Since 2019, PG&E has used data from Salo Sciences, acquired by Planet in 2023, as a feeder data set into its wildfire risk models. And, since 2023, PG&E has incorporated two additional uses for Planet data: measurements of Canopy Density as the thickness of a tree’s canopy can forecast how much material might strike a powerline, and Tree Mortality which can identify areas of dead trees, which pose a considerable risk to electric infrastructure and may indicate areas of high pest infestation and drought impacts.

Planet’s proprietary Vegetation Encroachment Planetary Variable combines multiple data sources together to deliver analysis-ready data on vegetation canopy height and cover, potential strike tree locations and dead tree cover. With that information, PG&E can evaluate areas with high potential for vegetation-driven outages and ignitions. The innovative technology helps PG&E optimize their operations and ensure workers are spending their valuable time investigating and fixing high-risk areas first, helping to reduce the probability of power outages and wildfire ignition from their electric grid assets.

“It’s a huge undertaking to manually observe and de-risk the over 70,000 square miles that PG&E has under management in California,” said Will Marshall, co-founder and CEO of Planet. “Planet’s satellite data and artificial intelligence supports broad area management to help reduce the risk of wildfires. We’re excited to be partners in this effort.”