Predicting Future Temperatures and Precipitation Just Became Easier

Posted by InnoCentive on Mar 14, 2019 5:06:53 PM

Everyone can understand the effects of having inclement weather ruin their day. They also know the feeling of having a local TV station’s meteorologist steer them wrong on a weather prediction. This is because predicting weather patterns is an incredibly nuanced and tricky business. For farmers and water reclamation specialists, predicting precipitation is all the more important. That’s why the Bureau of Reclamation reached out to InnoCentive. They were looking for a better method of predicting temperatures and precipitation rates in both short and long-term forecasts.  

InnoCentive is happy to report that their call for a better solution was answered by our open innovation marketplace. Our Solvers got to work and delivered not one or two, but three, successful solutions. 

David Raff, PhD, the Reclamation Science Advisor, said: 

"We are very excited with the outcome of this competition The innovative American public is demonstrating that we can improve sub-seasonal forecasts for temperature and precipitation which will allow water managers to better prepare for shifts in hydrologic regimes, such as the onset of drought or occurrence of wet weather extremes." 

InnoCentive’s Chief Innovation Officer, Jon Fredrickson, said:  

“We are thrilled for Reclamation that they found material and important advances in forecasting in Rodeo I.  Our Solvers showed up, they understood the problem well and they offered a diversity of Solutions that we now see Reclamation also found valuable.” 

About the ChallengeSub-Seasonal Climate Forecast Rodeo

The Challenge put forth by the Bureau of Reclamation, and created in conjunction with InnoCentive, sought to improve on the existing sub-seasonal forecasts (spanning approximately 15 to 45 days in the future). It asked Solvers to develop a system that would perform demonstrably better than the existing baseline forecast for temperature and precipitation. 

After having three months of developing their systems, Solvers were given 13 months to initiate forecasts every two weeks in order to demonstrate the efficacy of their forecasts.  

See the full details of the challenge here. 

The Finalists  

The Bureau of Reclamation selected three winning teams to share the $525,000 prize, stating each of the three teams “demonstrated novel approaches and outperformed the baseline forecasts.” The three finalists were:  

  1. Salient, a family team from Massachusetts consisting of Raymond Schmitt (an oceanographer) and his two sons, Eric and Stephen. They were awarded $250,000 for an artificial intelligence model that used ocean data to make forecasts for rainfall. 
  2. StillLearning, a team comprised of Lester Mackey of Arlington, Massachusetts, Judah Cohen and Ernest Fraenkel of Newton, Massachusetts, and Jessica Hwang and Paulo Orenstein of Stanford, California. They were awarded $200,000 for combining two methods that led to numerous awards for forecasting temperature and precipitation.  
  3. Lupoa13, a team made up of Anthony Lupo and Joseph Renken of Columbia, Missouri, and Joshua Herman of Oshkosh, Wisconsin. They were awarded $75,000 for using global pressures anomalies to identify analog periods from historical records.   

Two other teams also outperformed the benchmarks of the year-long competition. See the Bureau of Reclamation’s full report on the Challenge and winners here.

The Bureau of Reclamation considers the Challenge to be a huge success and is currently planning to launch another Challenge, the Sub-Seasonal Forecast Rodeo II this summer. 

Follow InnoCentive

Search Blog

Join 5,000+ Subscribers

Recent Posts