Thursday, December 3, 2009

2009 Atlanta rainfall and a look ahead

Let's start at the beginning.  Have you ever wondered whether you can find an average year of rainfall in the historical records?  Does such a thing exist in the wild?  With the assistance of the state climatologist, I was able to accumulate Atlanta rainfall information going back to 1930.  Weighting each month equally, 1997 (blue curve shown above) is the best match to the average rainfall year (purple curve shown above).   Put another way, I did not worry whether the area beneath the curve matched (annual accumulation), but compared each month separately using a least squares method.  What about the curve differences?  Atlanta rainfall is not "dependable" and no two years look exactly the same.  More about this latter.

Shown above in the red curve is the 2009 rainfall (so far).  The big 2009 hump in September-October brought flooding, loss of property, lives and bee lives too.  I speculate that the next largest hump (March) adversely affected nectar gathering during our best in-town nectar flow (tulip poplar flowers).  Why so much rain in 2009?  Has the train left the tracks or has the train always been off its tracks?  Probably a little bit of both--Atlanta rainfall is complicated in part to tropical weather influences and let's not forget the butterfly effect.

Back to data mining--I looked at 80 years of record keeping and found that 1977 (shown in blue) is the best rainfall match to our current year (2009 shown in red).  With me so far?   From a honey harvest point of view, I am only attempting a short range rainfall forecast.  Using a simplistic forecast scheme, what can I expect during February through April 2010?  Shown in the green curve are rainfall values for 1978 (the year following the 1977 best match).  In other words, my February through April fall prediction for 2010 is one of lower than average rainfall.  What if I had used the 2nd best rainfall match to 2009?  So as not to clutter the graph, the 2nd best match is not shown.  If you are still interested, the February through April 2010 forecast is similar using the best and 2nd best match to 2009 rainfall.

Shown above is the "official" February through April 2010 rainfall forecast.  It's hard to see, but Atlanta lies between that brown Indiana/Ohio bulls eye and green Florida shading--in a region of no shading (white map color).  In other words, the "official" forecast for Atlanta is one of "equal chance" of 1) above normal, 2) normal or 3) below normal rainfall.  Is this some kind of super computer joke?   What, unable to forecast? 

I don't have much skin in the bee game, but I'll stick my neck out and say that in 2010, Atlanta will not have too much rain during the tulip poplar nectar flowering.  Next spring, I'll review my rainfall forecast using hind sight.

No comments: