One of the reasons why there is business interest in the science of weather forecasting is because there exists a strong relationship between financial variable like sales, volume, profits/margins, and the weather. This type of digital-led information gives companies a competitive edge, as well as mitigates any weather-related risk. An example of a company working in this area is DTN , the largest private weather company in the world.
To learn more, Digital Journal caught up with Mike Eilts, who is one of the leading practitioners in the world when it comes to incorporating weather technology in business. Eilts worked at the National Severe Storms Laboratory (NSSL) for 18 years before forming the company WDT. Eilts works as the senior vice president at DTN.
Digital Journal: Why are weather data so valuable?
Mike Eilts: Weather impacts decision making in almost every industry, from operations and safety, to supply chain logistics and demand. When companies use weather information to improve or secure their operations, they can improve their bottom line. Studies show that weather disrupts up to 70% of the operating and financial performance of companies globally and is estimated to cost U.S. businesses alone nearly $630 billion annually.
DJ: How are companies using weather data as part of their digital transformation strategies?
Eilts: As businesses become more savvy with their digital information, they can quantify the effects of weather on their business much more than they could before. Sophisticated businesses are using that knowledge to incorporate weather data into their decision support systems and operational intelligence. As businesses engage in digital transformation, they evolve from using weather information to react to emerging or past weather events to incorporating weather data with other business data to drive long-term strategic decisions. We are seeing this with our customers across multiple sectors. For example, our shipping customers optimize their routing by incorporating weather data with other seakeeping data, such as vessel type and marine conditions. Doing so helps them improve performance and profit margins while reducing fuel consumption.
Once a company is tying weather and other operational data together, they then move to combining weather data and artificial intelligence that results in recommended actions that drive business outcomes before weather happens. For our customers in the transportation sector, relying on hyperlocal forecasts using road sensors, map data and predictive weather models help managers schedule crews and road maintenance work, as well as prepare for potential emergency responses.
DJ: How important is digitally captured and processed weather data?
Eilts: Most weather data observed are natively digital data: surface observations, radar data, weather satellites, to name a few. These discrete data create enormous amounts of information about the existing state of the atmosphere. Much of the preprocessing of these data is for quality control and to combine the data into a coherent picture about the atmosphere. Then these data become inputs or starting points for creating a forecast using sophisticated numerical weather prediction models run on high-performance computing clusters or supercomputers. Using an ensemble of individual models, including our own proprietary models, as well as a combination of models results in better forecasts and empowers sophisticated users to model potential risks in a probabilistic manner. Even though the techniques applied to produce forecasts continue to evolve, the starting point is a quality set of observations that accurately describe the state of the atmosphere.
DJ: How does the DTN technology work?
Eilts: DTN is a leading data, analytics and technology company; one that truly understands how weather can impact operations across a number of industries. Our operational intelligence delivers what customers need most, whether that is long-range forecasts, hyperlocal forecasts, lightning data, up-to-the-second updates or even direct connections with our meteorologists. We focus on making sure that our data, forecasts and intelligence are available in a variety of ways: web portals, mobile apps, APIs, etc.
To do that, DTN handles terabytes of weather data produced daily around the globe to create critical, accurate, precise and timely forecasts. Weather industry providers need to have a certain scale to achieve the level of accuracy, detail, and frequency required by global businesses that need to make smart, confident decisions based on those forecasts and insights.
With our size and global scale, DTN is able to invest time and resources in bringing our customers machine learning solutions that leverage weather and customer-specific data. We use an ensemble of weather models (proprietary, government and research agencies) to produce a single forecast solution and then apply machine learning to validate and adjust forecasts, increasing accuracy. Then, we combine the forecasts with customer and industry-specific data to better predict weather-related impacts that can disrupt a customer’s business. This brings actionable insights for our customers so they can be confident in their weather-related decisions ahead of any major weather event.
DJ: How accurate are the forecasts?
Eilts: Currently, the most accurate temperature forecasts can vary from actual to three degrees Fahrenheit within 24 hours. A longer forecast (6-9 days out) can vary by up to six degrees. DTN uses multi-models, ensemble techniques, machine learning, and statistical methods to create our global forecasts and are on the leading edge of science – meeting or beating the most accurate information available and consistently being recognized as the most accurate forecasts available.
DJ: How powerful are meteorological predictive analytics?
Eilts: Predictive analytics can improve operational efficiencies, save time, cost and lives. It removes the guess work for decision makers when weather threatens their business, such as major winter storms, landfall hurricanes or prolonged heat waves. It is just as powerful in marginal weather conditions. For example, knowing if there are ice patches on certain stretches of the road, if a rescue helicopter can safely take off, or an outdoor event should be delayed.
DJ: Who are DTN’s main clients? What do they do with the data?
Eilts: DTN works with many of the Fortune 500 companies in the energy, transportation, agriculture and other weather-dependent industries. Here are a few examples of how our clients are using weather data to make smart weather-related decisions.
Eversource is a utility company that relies on our outage prediction solution to make complex decisions during the days leading up to a forecasted storm that may produce power outages in their service territory. The company turns the weather intelligence into plans, prepares for outages and deploys crews often in advance so they can restore power sooner.
Boeing is using one of our more recent innovations that predicts clear air turbulence (CAT). CAT causes violent buffering of aircraft and can result in serious injury to crew and passengers. While it is virtually impossible to detect, our sophisticated weather models and expert systems allow us to accurately forecast turbulence at major flight cruising levels. Our turbulence forecasts allow airlines and their pilots to choose routes that are safe while not deviating far from the originally planned flight path, thereby mitigating risk and retaining operational efficiency.
