Another AccuWx 25 Day Forecast Comparison

This is my third installment of following the AccuWeather 25 day temperature forecast and comparing the output to my cycle temperature trend output. One and Two can be found by clicking on the links. In the second comparison last month I eliminated a portion of the cycle in hopes of quantifying the dynamics of the trend. Turns out one "anomaly" did the trick and the LRC came out on top with less error.

I create the comparisons by saving a snapshot of the AccuWx forecast and uploading a 'cached' version of the webpage onto my server where it sits unmolested. (I took the June snapshot on June 4th) I then extract the AccuWx data into the graph shown below. As the month progresses the graph will dynamically update after I enter in the observed daily temperatures. Check back every now and then to see the ongoing results.


(Click on the graph for the full set of data.)

If there is anything substantial to the 'every other cycle' thought process the second half of June should be a scorcher here in Oshkosh. The cycle comparisons used to generate the model trend output are my interpretations of the cycle. My interpretations of the cycle are lined out in my cycle calendar. If there are any questions or thoughts on my research and analysis of the LRC or how I presented the material just let me know in the comments section of the blog. Thanks for reading!

Every Other Cycle?!

Update: June 26, 2012

"If you look at this play what we're trying to get is a seal here, and a seal here, and we're gonna run this play in the alley" - Vince Lombardi

Two cycles ago. Mar 19, 2012.

And two cycles before that. Dec 14, 2011.

Did you see it coming?

Next week. Jun 18, 2012 (Jun 11 - 18Z GFS Forecast)

Turns out Jun 20, 2012 was the day. 
Good job GFS, just 2-ish days early...

Below are duration calculations to put the time frames into perspective.

Dec 14 to Mar 19 = 97 days / 2 cycles = 48.5 days
Dec 14 to Jun 18 = 188 days / 4 cycles = 47 days
*Mar-19 to Jun-18 = 92 days / 2 cycles = 46 days **
**As jdrenken has proven, the computer models are usually a day or two quick this far out. Click on his forum handle to learn more.

Mar-19 to Jun-20 = 94 days / 2 cycles = 47 days

As always, all these comparisons boil down to the beholders interpretation of the cycle and I can't emphasize this feeling enough. My interpretation of the cycle is lined out in my cycle calendar. If there are any questions or thoughts on my research and analysis of the LRC or how I presented the material just let me know in the comments section of the blog. Thanks for reading!

The Battle Continues

Seriously, how fun is this, comparing the AccuWx 25 day forecast to the LRC? At this point, who really cares right? I mean, it is what it is. The theory is going to have critics, the theory is going to have followers. Personally, I am past trying to convince people (myself) that our atmosphere cycles. This stuff has become a passion and I just simply enjoy following along.

Anyone who follows along, anyone who takes the time to understand the theories (yes, it is not cut and dry) knows it is a fantastic magical joy ride crystal ball and is crucial to have in a forecasting bag of tricks. You know exactly what I am talking about, the smoke and mirror gag and the delusional eye strain gag where a woman gets chopped in half.

It all boils down to the beholders interpretation and I can't emphasize this feeling enough. My interpretation is in my calendar. Other interpretations are located on the right hand side of this blog under the "LRC Blog Roll". Once this baby breaks out it will come down to who legitimately follows along and it will be noticeably evident in the accuracy and confidence of the forecast.

Within my model analytic I have these magical buttons that suppress certain portions of a cycle. I decided to run the model twice for this comparison. I enjoying playing with numbers.

AccuWx 25 day forecast May 6-30, 2012 versus the LRC.
(cached version of AccuWx forecast)

Click on the graphs for the entire data set. If you choose not to, I can just tell you that AccuWx won versus the blend of all cycles of the LRC. Not by much, but still. I couldn't let the LRC lose, so I got rid of the March data. Wallah! LRC kicking butt once again.

If there are any questions or thoughts on my research and analysis of the LRC or how I presented the material just let me know in the comments section of the blog. Thanks for reading!