Extreme Weather Events: Why Energy Providers Need Hyper-Local, Ground Truth Weather Forecasting

Lightning strikes behind the lines designed to carry it in South Texas

As climate change accelerates, energy providers face a number of critical challenges. Of these, a notable increase in extreme weather events is causing particular disruption. High winds and lightning strikes can inflict significant structural damage to energy grid infrastructure, downing power lines and leaving large numbers of customers without power, sometimes for extended periods.

This issue affects many states and cities across the US, none more so than Michigan. Michigan had 71 major weather-related power outages between 2003 and 2012, with an average of 800,000 customers affected each year. This issue is compounded by the state of the national energy grid, much of which was designed and implemented up to 120 years ago, and is increasingly fragile and susceptible to adverse weather. Beyond thunderstorms and high winds, hurricanes, tornadoes, wildfires and ice storms pose a severe threat in various regions of the US.

Energy providers require accurate, localized weather forecasts to be able to proactively react to inclement weather conditions, deploying the maintenance personnel to prevent and respond to outages rapidly. However, traditional atmospheric forecast models predict conditions for the atmosphere above rather than the weather we experience on the ground and are often not granular enough to predict the neighborhood-level forecasts required for maintaining the grid.

Real-time, hyper-local road weather forecasts can close this gap and provide the highly accurate information of conditions on the ground that equip energy providers to proactively react to emergency situations. This is where Fathym’s WeatherCloud solution comes in.

WeatherCloud combines proprietary algorithms and machine learning with sensor and atmospheric weather observations to provide highly-accurate, Ground Truth forecasts. Using real-time data from WeatherMesh environmental monitoring stations, in conjunction with the WeatherCloud Ground Truth forecast engine and high winds/lightning prediction algorithms, energy providers can predict severe weather events and where outages are likely to occur.

WeatherCloud offers a statistically-based machine learning forecast system, providing more accurate and scalable ground weather information for end users than physical models and enabling sub-kilometer or neighborhood-level forecasts and alerts. Energy providers can monitor conditions on custom dashboards, developed through the Fathym Framework, and map weather elements – such as high winds, lightning strikes, outage zones and at-risk zones. Accurate, hyper-local road weather forecasts equip energy providers with the real-time, actionable information and alerts they need to more efficiently deploy resources, saving money and time by proactively responding.