How well did last spring’s Atlantic hurricane season’s forecasts perform?

Hurricane season starts on June 1 and runs through November. Despite our man-made timeline for storms, which is based on typical start and end dates over the years, mother nature occasionally starts the year early as she did this year with Tropical Storm Arlene which developed on April 19.

Every spring, Colorado State University and North Carolina State University use statistics, global models, and other information to create their own seasonal forecasts for the upcoming hurricane season. The predictions garner a good deal of attention for a few days, and then are typically forgotten by the general public.

What they predicted:

Colorado State’s initial forecast was issued on April 6, 2017. For the purpose of this blog, I am only looking at the initial forecasts of each school. Revisions are usually transmitted in August, after the season has begun and the forecasters have a chance to see how the conditions may differ from what the models predicted in April.

CSU predicted slightly below average activity (citing 12 to be the median number of named storms from 1981-2010) with 11 named storms, four becoming hurricanes, and two major hurricanes (category 3 or higher on the Saffir-Simpson scale).  One key factor they considered was their expectation for the development of El Nino with its warmer-than-normal sea surface temperatures in the Pacific. Yes, water temperatures in the Pacific affect the number of storms we have in the Atlantic. In weather, everything is connected.

NCSU’s Dr. Lian Xie – one of my professors when I earned my degree – saw things slightly differently and predicted an above-average year. He and his cohorts predicted 11 to 15 named storms (citing 11 to be the average number of named storms in the Atlantic from 1950-2014). They thought four to six of those may become hurricanes, and one to three of those could become major hurricanes.

2017 map

The National Hurricane Center’s 2017 Preliminary Hurricane Season Summary map

How they did:

The reality of the season was even greater than either group expected. The summary on the National Hurricane Center’s website lists 16 named storms, 10 of which became hurricanes, and six of those became major hurricanes. While the season isn’t technically over, it looks like Tropical Storm Philippe was our last named storm of the season. You might not have even noticed him. He spent his time far out in the Atlantic on October 28 and 29.

Why it matters:

Whether meteorologists are forecasting for a season or just the next three days, looking at predictions versus reality after the period ends helps us learn from our mistakes and confirm what we got right. This summer, El Nino never really formed as the CSU team thought it would. In fact, in August, the sea surface temperatures were slightly cooler than average in the equatorial Pacific. That fact might have been a contributing factor to the final tally for the Atlantic basin.