Climate change has already had significant detrimental effects across the United States and globally. On the West Coast of the USA, we are experiencing less winter snowpack accumulation, longer springs and hotter summers, resulting in the largest, hottest fires ever seen. Nationally, a recent Nature paper concludes that our national flood risk mapping system is underestimating current risk by 60%, and that overall flood risk will increase by 25% in the next few decades. Meanwhile Texas and Europe have seen ice damage driving significant outages.
It is becoming increasingly clear that statistical risk models built against historic climatologies are not meeting current needs. Utilities need to include nowcasting, multiple short term forecasts and IPCC climate change scenarios directly into utility asset management practices. They also need to continuously monitor vegetation structure, fuel loads and fuel moisture in detail over large areas. Operationally, this is needed to improve routine vegetation and equipment management, public safety power shutoffs (PSPS) and disaster response capabilities. As an industry based on fixed assets with high capital expenditures and sizable
maintenance costs, it is economically critical that utilities make large long term investments with a clearer understanding of both risks and opportunities.
These are individually and collectively hard topics to address. However, there are already some important best practices emerging, as well as significant improvements in supporting solutions. The combination of GPS, microsatellites and LIDAR technologies is simultaneously driving down remote sensing costs, increasing resolution and increasing timeliness. Together, these allow us to routinely and economically gather environmental data describing conditions around utility networks, as well as influencing demand. For example, we can continuously assess the drought stress and wind exposure of potential strike trees, even miles from the nearest road.
The challenge most utilities now face is not the availability or cost of the primary required datasets. It is the development of data pipelines, analytic tools and models which can automate the integration of such vast, variable and high velocity data and generate reliable risk assessments and related analytics. In a highly-regulated industry, the organizational changes are more difficult than any simple model of technology adoption.
We explore here three innovative methods used in adapting commercial “off the shelf” technologies to support real world utilities use cases. We begin by looking at weather and climate models, and their connection to user-driven visual analytics. Next, we consider fire risk, and the advent of building and individual tree-level models supported by continuous monitoring and ML. Such ‘digital twins’ allow assessment of strike tree risk anywhere, as well as critical ‘fire safe’ practices in wildland urban interface (WUI) environments. Lastly, we look at how conventional vegetation management contracting and auditing practices can be enhanced using high-resolution, high-cadence tasked imagery. We believe that the combination of broad-scale risk assessment with detailed follow-up capabilities can simultaneously lower costs and improve performance over common single-sensor single-scale approaches.