Long-range hurricane forecasts defended

Colorado State University scientists
Phil Klotzbach and William Gray will issue their second forecast for the Atlantic
hurricane season Wednesday, calling for well above average tropical cyclone activity
this year.

In an interview during the National
Hurricane Conference in Orlando, Florida last week, the two scientists defended
the issuance of long-range forecasts.

After badly missing their forecasts
in 2006 and 2007 – which Klotzbach admits “stunk” – some in the media have
abandoned publicising their forecasts. 
Jim Poling of the Florida News Network, another of the many speakers at
last week’s hurricane conference, said his network steers clear of seasonal
forecasts because he believes they confuse the public.

“Everyone was bailing [on the
forecasts],” Gray said.

Although Klotzbach said he didn’t
want to put out a December forecast for the 2010 season, Gray believes they are
useful. A regular speaker at the National Hurricane conferences, Gray said
there’s great curiosity about what can be expected from hurricane season.

“For years, I’ve been going to [hurricane
conferences] and everyone always wants to know ‘is this going to be an active
season’?” he said.

Gray noted that the forecasts he
and Klotzbach issue are based on “hindcast”, a method where various climate
conditions existing before hurricane seasons in the past are compared to the
actual activity that occurred during the subsequent hurricane season.

“We would never put out a forecast
if we didn’t have hindcast skill,” Gray said, adding that he can Klotzbach now
have 60 years of past data to use for their forecasts.

“What we can say is that if the
atmosphere behaves this season as it has in has the past 60 years, the odds
favour an active or inactive season,” he said. “That’s all we can say, and we
give rough numbers on it.”

Gray believes this is information
the public should want to know.

“Nobody knows what the season is
going to be,” he said. “However what are the odds? How has the climate set
itself up now to say something about the coming season?”

As far as Gray knows, he and
Klotzbach are the only people really using hindcast as a way of forecasting
hurricane activity.

“I think it’s a worthy venture as
[compared] to having nothing,” he said.

Gray said people should not judge
the forecasts on one or two years.

“These forecasts are based on long
period hindcasts and we’re going to fail on a season or two, or maybe on two in
a row as in ‘06 and ‘07,” he said.

However, the forecast failures led
to methodology improvements and the duo had accurate forecasts in 2008 and

“You learn much more when you
bust,” Gray noted.

Klotzbach agreed, adding that after
2007, he re-did all of the statistical models to get better results from the

However, even when the Colorado
State forecasts are accurate, Klotzbach pointed out that there is no way to
predict if a particular strong hurricane might develop and affect a particular

“Sometimes the media doesn’t really
interpret [the forecasts] correctly,” he said.”We’re just trying to predict how
active the hurricane season is going to be.”

But there are also other benefits
to the forecasts, Klotzbach said.

“I think one of the primary good
things about these forecasts is that they’ve taught people a lot about the
kinds of physical factors you look at for an active season.”

Gray agreed, noting that he was the
first person to draw a correlation between the effects of El Nino in the
Tropical Pacific on hurricane seasons in the Atlantic.  That correlation, which has been a discussed
part of his forecasts for years, is now widely accepted by scientists.

Based on the current climatology, Klotzbach thinks 2010 will be stormy
in the Atlantic basin.

“At this point we’re expecting the
hurricane season to be pretty active,” he said.

One reason is that the El Nino
phenomenon in the Pacific, which helped reduce hurricane activity last year, is
expected to diminish this year.

El Nino, which is a warming of the
sea surface temperatures in the tropical eastern Pacific Ocean, creates greater
wind shear in the Atlantic basin that inhibits hurricane formation and

Klotzbach thinks the current El
Nino is probably going to end.

“That’s… what we’re expecting,”
he said.  “Most of the forecast models
are calling for a weakening to either a neutral condition, which is kind of the
average, or La Nina, by summer and fall.”

Because El Nino can affect weather
over a large portion of the world, forecasting models have improved in recent
years. Klotzbach said one European model, known as ECMWF, is particularly good
a forecasting El Nino.  That model uses
an ensemble approach with 41 separate members. In a recent run of all 41 members,
only one showed a continuance of El Nino by September, the peak of the Atlantic
hurricane season.

Another reason the two scientists
are predicting a very active hurricane season is the unusually high sea surface
temperatures currently in the part of the tropical Atlantic Ocean where most
hurricanes form.

“There is a very significant
warming in the Tropical Atlantic. The anomalies are very high right now,”
Klotzbach said, adding that the temperatures are currently at record levels for
this time of the year.

“Just because it is warm now
doesn’t guarantee it will be warm during hurricane season,” he said. “But there
is certainly a positive correlation.

“At this point, we expect the
waters to probably at least be significantly above average in the summer and

As part of the improvements to
their forecast product, Klotzbach and Gray now include probabilities of having
a tropical cyclone impacting all locations in the Caribbean.

Klotzbach said that this year there
is a 27 per cent chance of a tropical cyclone coming within 50 miles of the
Cayman Islands, up from the statistical 100-year average of 20 per cent.  He also said there was 10 per cent chance of
a major Category 3, 4 or 5 hurricane coming within 50 miles of the Cayman Islands,
up from the long-term statistical average of seven per cent.