End of Seasonal Forecast Press Releases

[HurrMark]

One of my big irritants is when a forecast of, say, 9 hurricanes is made, the media's headline often says something like "Nine hurricanes to menace the coast this year", like all of them will make landfall.

I think that while for the most part (this year as a major exception), the forecasts have been good in terms of numbers, it means nothing in terms of where these storms will end up. As wxman57 said, you can have an Alicia or Andrew-like year with few storms but that one memorable storm, or a year like 2010 with plenty of bark but really no bite for the US.

Klotzbach and Gray have been forecasting landfalls for a while...since we will likely go an eighth consecutive year without a major hurricane, let us look back at what they have forecasted since 2006 in terms of the probability of a major hurricane (by June forecast):

2013 - 72%
2012 - 48%
2011 - 72%
2010 - 76%
2009 - 48%
2008 - 69%
2007 - 74%
2006 - 82%

So except for 2009 and 2012, every single year had forecasts that were far above climatology (52%), and even 2009 and 2012 were close to climatology. The Brier score is .47, which is roughly equivalent to forecasting a 30% chance of rain for eight consecutive days, and it rains each day. Not that great of an outcome. So there really is a lot of work that has to be done. I personally think the climate signals are changing where it is becoming more and more difficult for a strong tropical cyclone to make landfall in the US...but we need a few more years of data to confirm this. I also think that we need to take more objectivity - we may have 15 storms each year, but if half of them are 40-50 mph tropical storms, how is that any busier than a 10 storm season where the weakest storm is a 60-65 mph tropical storm? I have noticed a recent trend of the average ACE/storm going downwards...whether it continues remains to be seen, but I think the 2006-2013 pattern is much different from the 1995-2005 pattern in many regards.

[jeff]

I've occasionally discussed the seasonal forecasts with Phil Klotzbach. My main point has been that a prediction of a number of named storms is of little value to anyone. People will make incorrect assumptions about the level of risk based on the numbers alone. We've seen many seasons with fewer than 10 named storms but with major impacts. Andrew (1992), Betsy (1965), Alicia (1983) and Audrey (1957) come to mind. And there have been quite "active" seasons with little or no U.S. or Caribbean threat.

I've said that what we really need to attempt to forecast is where the active regions will be across the basin. Basically, what areas may have the greatest risk of an impact? Could the concentration of hurricanes in the vicinity of Florida been predicted in the spring of 2004? Until we can do that, then seasonal predictions don't really have much value. I've attempted to do this with at least some level of success in recent years (not much this year).

The bottom line is that I think seasonal predictions of numbers alone often do more harm than good. The general public will likely misinterpret what the numbers mean. They should be prepared for a possible hit regardless of the prediction. Businesses may incorrectly assume an increased or decreased level of risk based upon the numbers.

However, as a meteorologist I love to hear what my colleagues are thinking about upcoming seasonal activity. I don't think that such forecasts can pass between forecast groups without the media wanting to make a story out of it. It's unfortunate, but I don't think we (forecast groups) can just say that we're not issuing any more forecasts. I think the forecasts, and the attention they garnish, force us to find the flaws in our methods. Why was this year so quiet in all basins? (I think I know)

I think this topic becomes an exercise in understanding and point of view which are all different depending on who is looking at the numbers. Those on this board have way more understanding than nearly all in the general public. The media and especially social media is changing the way forecasting is done and in many cases if you are not saying it is raining at location X and it is...then you are wrong...good luck with that during the convective season. Warnings are even issued now based off of Facebook and Twitter posts of tornadoes and flooding...maybe that is a good thing...at least the warning is getting out, but the event is already in progress making the forecaster look somewhat silly. As far as hurricane seasonal forecast, I don't think they have much value and don't think they ever will. As 57 stated, where storms will tend to affect could be more helpful, but suggesting in May that an area could see increased activity (even if it were a TS) could cause its own set of problems. I do countless public events on hurricane education and am amazed at the amount of people who just don't care until it is 3 days away. They surely don't care in May of what MAY happen in September!

[vbhoutex]

I've occasionally discussed the seasonal forecasts with Phil Klotzbach. My main point has been that a prediction of a number of named storms is of little value to anyone. People will make incorrect assumptions about the level of risk based on the numbers alone. We've seen many seasons with fewer than 10 named storms but with major impacts. Andrew (1992), Betsy (1965), Alicia (1983) and Audrey (1957) come to mind. And there have been quite "active" seasons with little or no U.S. or Caribbean threat.

I've said that what we really need to attempt to forecast is where the active regions will be across the basin. Basically, what areas may have the greatest risk of an impact? Could the concentration of hurricanes in the vicinity of Florida been predicted in the spring of 2004? Until we can do that, then seasonal predictions don't really have much value. I've attempted to do this with at least some level of success in recent years (not much this year).

The bottom line is that I think seasonal predictions of numbers alone often do more harm than good. The general public will likely misinterpret what the numbers mean. They should be prepared for a possible hit regardless of the prediction. Businesses may incorrectly assume an increased or decreased level of risk based upon the numbers.

However, as a meteorologist I love to hear what my colleagues are thinking about upcoming seasonal activity. I don't think that such forecasts can pass between forecast groups without the media wanting to make a story out of it. It's unfortunate, but I don't think we (forecast groups) can just say that we're not issuing any more forecasts. I think the forecasts, and the attention they garnish, force us to find the flaws in our methods. Why was this year so quiet in all basins? (I think I know)

I think this topic becomes an exercise in understanding and point of view which are all different depending on who is looking at the numbers. Those on this board have way more understanding than nearly all in the general public. The media and especially social media is changing the way forecasting is done and in many cases if you are not saying it is raining at location X and it is...then you are wrong...good luck with that during the convective season. Warnings are even issued now based off of Facebook and Twitter posts of tornadoes and flooding...maybe that is a good thing...at least the warning is getting out, but the event is already in progress making the forecaster look somewhat silly. As far as hurricane seasonal forecast, I don't think they have much value and don't think they ever will. As 57 stated, where storms will tend to affect could be more helpful, but suggesting in May that an area could see increased activity (even if it were a TS) could cause its own set of problems. I do countless public events on hurricane education and am amazed at the amount of people who just don't care until it is 3 days away. They surely don't care in May of what MAY happen in September!

Very good points from both of you gentlemen. We ALWAYS appreciate your input! I don't do anything like you or Wxman57 as far as public speaking, but I do talk to/with friends, family, church members, etc. that know of my weather geekieness and I too have found what you have Jeff. It amazes me that they just don't care unless it is close and all they want to know what is going to happen at my house? I do post watches and warning on FB when they are affecting areas where I have contacts either through S2K or FB to further get the warnings out. I also post verified/able weather as it is occurring in the hopes that it will be helpful, but my info is either taken from a verifiable agency or direct observation. That is the only way it should be and hopefully those of us that do this are "helping" the general public to become more weather aware and possibly safer.
I'm not sure how "helpful" beginning or end of season forecasts are, but I do like them since they usually have information in them that helps me and other geeks learn and/or understand more about weather.

[Frank2]

Aside from something forming in the SW Caribbean Sea the season is essentially over, and hopefully the seasonal forecasters have learned something from this season and will follow Dr. Chan's example and not issue press releases for future seasons until their accuracy at least improves...

There are a number of scientific journals where their forecast can be sent as a journal article (probably the Monthly Weather Review would be the most likely) and from that anyone in the public who is interested can review their estimate, but to release a dire and possibly incorrect forecast to the media in the form of a press release is wrong...

We have enough real events to worry about than to worry about something that might or might not happen next summer and fall...

Again, this is why the forecasters and researchers of 30 years ago did not want anything to do with seasonal forecasts - they knew it was difficult enough to forecast a hurricane 5 days in advance, never mind 5 months...

Frank

[gatorcane]

Frank2 yeah I agree 100% with your thoughts on this. The key is " until their accuracy at least improves."

[supercane4867]

Breaking News=CSU verification of 2013 season is up,Will suspend future forecasts

The big news is that they will not issue forecasts unless they get funded. Read it at Read it at first post of thread.

NOTICE OF FORECAST SUSPENSION
The Tropical Meteorology Project has been issuing forecasts for the past thirty years. These predictions have served as a valuable information tool for insurance interests, emergency managers and coastal residents alike. While these forecasts were largely developed utilizing funding from various government agencies, recent attempts at obtaining continued grant funding have been unsuccessful. Funding from several insurance companies enabled the continuation of these forecasts in recent years. However, the forecast team has recently lost some of its financial support from industry. Consequently, new sources of revenue are required to keep the forecast going. Interested parties are invited to contact Phil Klotzbach directly via email at philk@atmos.colostate.edu for additional discussion of potential sponsorship opportunities.
The Tropical Meteorology Project will suspend issuing seasonal forecasts beginning in April 2014, unless additional funding for the forecasts is forthcoming. The CSU forecast team is currently seeking partnerships with the private sector in order to continue these predictions. Please see the sponsorship brochure if you are interested in supporting the forecast team.