Mike's Blog: Understanding models - WLOX.com - The News for South Mississippi

Mike's Blog: Understanding models

Posted: Updated:
  • Most Popular StoriesMost Popular Stories

  • Friday, May 24 2013 7:26 PM EDT2013-05-24 23:26:15 GMT
    Several Jackson County neighbors are extra cautious of their surroundings these days after some unusual sightings in their neighborhoods. A big black bear has been spotted this week roaming around the
    Several Jackson County neighbors are extra cautious of their surroundings these days after some unusual sightings in their neighborhoods. A big black bear has been spotted this week roaming around the St. Martin community. Those close encounters have some neighbors a bit nervous.
  • Monday, May 13 2013 12:24 PM EDT2013-05-13 16:24:04 GMT
    MARION COUNTY, MS (WDAM) - A weekend-long drug raid left 22 individuals behind bars, and more arrests to come. According to Marion County Sheriff Berkley Hall, the drug raid is the result of several
    A weekend-long drug raid left 22 individuals behind bars, and more arrests to come.
  • Friday, May 24 2013 10:24 PM EDT2013-05-25 02:24:32 GMT
    Those who are helping people overcome drug addiction said they are seeing more and more people hooked on Meth. One man shared his story of addition and recovery with the hope of helping others.
    Those who are helping people overcome drug addiction said they are seeing more and more people hooked on Meth. One man shared his story of addition and recovery with WLOX News with the hope of helping others who may be struggling.
BILOXI, MS (WLOX) -

There has been a lot of discussion about the models tracks associated with Tropical Storm Debby. Some models predicted the storm would track west toward Texas, while others tracked the storm east toward Florida. The storm finally moved into and across Florida.

The question is why didn't the models handle the storm better? The simple answer is the old computer model adage 'garbage in garbage out'. Meaning that if you don't have good data or very little data you aren't going to get a good result. You ask why run it then? You have to start somewhere. The data has to start building at some point. The longer you input data, and the more data you put into a model, the better the outcome.

The next question is why so many models? I often ask that question. I just got a list of models used to forecast tropical systems; it numbers 75. Some of the models are different versions of the same models, but 75? Now understand that not all the models are used on a regular basis.

Each model has strengths and weaknesses. Each model has biases based on the scientists that put it together. Models can have large scale and small scale resolutions. These all are part of the strengths and weaknesses of the particular model. Also keep in mind that computer models are for forecasters to use as a tool to decide what the storm is going to do. They were never intended for public use.

The final issue is who sees the model output and their understanding of the model. There are statistical and dynamic models. Statistical models use climatology as part of the solution. That means they look at a storm and its location and compare it to past storms to come up with track and strength information. When I was tracking storms in the Navy, many years ago, this was the only data we had. It wasn't very accurate. Dynamic models use mathematical formulas to track storms. They are the most used models. They take a lot of data and computing time to come up with a solution. They can be very accurate when enough data is used.

Just keep in mind that computer models are getting more accurate every year as long as the storm is well formed and in a well organized atmosphere. Computer models are going to have problems handling new storms, weak storms and storms in a weak atmosphere.

Copyright 2012 WLOX. All rights reserved.