TwinPlex forecast for the St. Paul / Minneapolis Area for September to December 2006.

 

Long Range weather forecasting can be a tough job, no wait; it is a tough job.  I have my believers and my naysayers.  While next day forecasting is roughly about 90% accurate by day 2 it decreases to 80% and keeps dropping from there.  A good long-range forecaster is roughly at 50%.

 

I'm often kidded about my job as the only job you can get paid at for being wrong.  I occasionally take exception to this considering how many other jobs are out there that tries to predict the future, i.e. sports, stockmarket (much more costly when your broker is wrong). I'm sure you get the idea.

 

Panic is the biggest mistake a long-range forecaster can make.  For example the summer of 2005 saw a record number named storms and Hurricanes in the Atlantic.  There were many stories in the press trying to link global warming to the increase in the number of storms. 2005 was a warm year and 2006 started off as one of the warmest years since the 1930's.  Forecasts called for another above normal year of hurricanes in the Atlantic. 

 

By late summer forecasters down graded their hurricane forecasts.  What happened to make them change their mind?  I'm not sure exactly what made them change their minds but I can tell you when I first heard their predication earlier this year I was skeptical.

 

When I make my long range forecast I follow some does and don'ts. 

 

First, I make my long range forecast, generally 1-year in advance.  This helps eliminate current media hype that can suck you into making a forecast that people want to hear.

 

Don't let one bad forecast period drastically change your forecast.  The weather pattern can change on a dime.  Wet weather can become dry, hot can become cold. Remember, your forecasting for a time period, generally months.

 

My father, the late meteorologist Bruce Watson, often told you can take 30-day periods and often finesse the data so that it supports your forecast.  Lets look at this past summer.  My July forecast called for hot and above normal rainfall.  The hot worked out well but as you recall you might remember wasn't July dry?  Was it?  Looking at the rainfall record from the period July 3 to August 2, I totaled 4.55 inches.  That is normal rainfall for that period.

 

Of coarse looking at my lawn it doesn't reflect "normal rainfall".

 

I try to stick to my original forecast.  I may make adjustments when asked what does the next week look like.  I really don't consider that long range forecasting.  A long-range forecast is a forecast made months and months in advance for a period months and months in advance.

 

I've already published a long-range forecast in my 2006 Watsons' Weather Watch Calendar.  Overall I feel it has worked out pretty good.  Here is my forecast for the remaining months of the year plus a few extras.

 

September, the overall trend for this month is warmer and wetter. The last below average September was back in 2001, otherwise we're coming off September's that have been 4-5 degrees warmer than normal. Precipitation is always a bit more tricky to forecast than temperature.  Storm tracks can be very finicky with a track 5-10 miles north or south can make the difference of receiving rain or not.  Still I'm looking for above normal rainfall for the month.

 

October is always a big transitional month.  The past three October's have been warmer than normal.  I feel this streak will come to an end.  Look for a cooler than normal October.  Precipitation since 2000 has either been feast of famine.  Look for above normal totals.

 

November is the month when temperatures come crashing down.  From an average high on the 1^st near 50 but by the end of the month the average high tumbles to 31. After a couple of mild Novembers look for this one to be closer to normal.  Precipitation and snowfall for the month may set the tone for the winter, below.

 

December, a month that has not been the Decembers of our youth. All signs point to a warmer than normal December.  Don't look for much in the way of snowfall this month, below normal snow.

 

Weather Values:

 

Month:                   Avg. High             Avg. Low              Avg. Precip.          Avg. Snow            Forecast

                                                                                                                                                                Temp.     Precip.

September             70.1                         50.8                         3.01                         Trace                      Mild        Wet

October                  57.7                         38.8                         2.37                         1.0"                         Cool        Wet

November              39.3                         24.6                         2.06                         10.4                         Near        Dry

December              25.7                         10.7                         1.16                         11.1                         Warm     Dry

 

Seasonal Forecast:

 

First Frost: October 8^th

First Hard Freeze: Oct: 15^th

First 1 inch snow: Nov. 15^th

Snow 2006-2007: 37 inches