Changing what we know (and do) about the weather

Vice President, Product and Analytics, The Weather Company


I have been in the weather and meteorological solutions industry for 15 years, but in the last five years I’ve witnessed an amazing transformation.

First, there are the incredible advancements in technology. As cloud and analytics technology have evolved, so have the accuracy and speed of predicting the weather. But, fundamentally, the core technique has remained unchanged over the years: atmospheric physicists have been driving weather prediction. Until recently, almost all weather research has centered on improving the physical equations that model the atmosphere.

A new shift

It’s no longer just physicists working on the problem. Applied mathematics has taken an equal stage in the field of weather prediction. At The Weather Company, we are hiring as many data scientists, machine learning experts and Hadoop and Cassandra engineers as we are hiring physicists. And when you put a PhD physicist in the same room with a PhD applied mathematician and ask them to solve the same weather problem, something magical happens—one plus one really does equal three.

Now we’re taking another significant step in the advancement of weather in business. WSI, the global B2B division of The Weather Company and the category leader in weather science, technology and content solutions, has formed a strategic partnership with IBM, and I can think of no better combination of companies to help enterprises use weather to their immediate advantage. This partnership brings together weather data and business analytics, creating a powerful new force for improving decision-making in industries such as insurance, energy, utilities, retail, logistics and government, among others, globally.

To provide a sense for how much weather has changed, here is what I’ve seen happen in just the past few years:

  • Deep machine learning adds bias correction to hurricane prediction and improves 6 to 10 day forecasts dramatically, where previous forecasts tended to stop at 5 days.
  • Marked improvements have been made in predicting electricity usage, which is driven by air conditioning and heating, which is driven by the weather (thanks to data scientists who think outside the weather sphere to gather information with predictive power, including how people use electricity differently on holidays).
  • The ability to use math and statistics, as well as the highest resolution weather data available, helps us better understand how weather impacts the sale of soup, beer, lotion, cars, houses and just about everything we buy.
  • Weather data can be used to predict taxi times at airports 8 hours from now with accuracy.
  • Customers can now get the most precise, resolute data available from anywhere in the world by delivering tens of billions of forecast requests per day, in real time.  

Bringing IBM and WSI together is like bringing the math and physics people together on an exponential scale. This partnership excites me in the same magical way that I was excited when I hired our first data scientists and put them in a room with physicists; we are bringing together the world’s best weather prediction and the world’s best analytics. And I predict that the results will be inspiring. 

Need further proof? Attend the April 9 Livestream event to learn more.