Jun 8, 2015

Lezak's Recurring Cycle

The weather pattern for the city of Kansas City is often discussed on the Weather 2020 blog. Maps of similar weather patterns from past LRC cycles that seem to match the forecast pattern will be shown in great content. It is stated that the Weather 2020 LRC forecasts are 75% perfect. When I followed the LRC cycling patterns, map comparisons were memorizing and fascinating. Soon I was turned on to data to gain another perspective of this cycling phenomenon. I will attempt to provide an example.

Using a full blend of cycles (M1), the previous 2 cycles (M2), and the most recent cycle (M3), the calculations I've generated show promise to be 75% perfect (someone would have to perform the verification results) with daily temps and precipitation chances. The charts in the link below pertain to KC using a 47 day cycle. Click on the image for more charts.



If we play the game of "which cycle matches best", see image below, we may find better than 75% perfection (someone would have to perform the verification results). But then how do we forecast from an unknown “which cycle matches best” analog framework? Seasonal jet medians? I will go with Cycle 2 and 3 blend to tell the story for this month. Click image for a larger view.



The cycle length for the chart shown above is derived from a cycle length auto-discovery formula I created. More info and the auto-discovery chart can be found at this link. Also, similar to the first chart in this blog using a 47 day cycle, the image below provides a look at the auto-discovery cycle length daily temps and precipitation chance forecast. Difficult to get better results.



It is obvious my work regarding the LRC is far from perfect. It is likely I will never be able to reproduce the accuracy of the real LRC. If there are any questions, comments, or suggestions on the material presented please let me know. Click on images for larger views. Thanks for reading!