Station Trends Based on the RRWT Average Correlation Wavelength

I added a new test specimen. I want to compare the group velocity wavelength method versus the average wavelength method. In order to do so I needed data and decided to implement the average wavelength method into a Station Trend output like I currently have for the group velocity method. The forecast data for the average wavelength method can be found here. I will be conducting verification on both methods. On a side note, the average wavelength method acts as a numerical genesis for the Lezak Recurring Cycle and the Doug Heady Pattern cycle length. This means anyone can forecast like Gary or Doug as the LRC/HP cycle length is automatically generated on a daily basis, all year round. If there are any questions, comments, or suggestions on the material presented please let me know.

Fundamental Frequency

A changing of the guard? Notice the temporal elongation of frequency, ~30 transitions to ~35 and ~60 transitions to ~70. A phasing of the wave is taking place with a moderate ~40 now showing up.



This paints an anomalously warm ECONUS the next few weeks, with a late July, early Aug reprieve.



My calcs are showing an upper 40s for a Lezak/Heady style cycle length. A forecast value going above 50 as we head through September. I wonder what these guys will come up with through Summer and into Fall. Both of their blogs are listed in the Blog Roll on the right.



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!

Using the Bering Sea Rule to Predict Severe Weather Events and Verifying the Prediction with the Recurring Rossby Wave Train

The images in the Tweet below show the Bering Sea Rule correlation to the CONUS. Based on our research (page 46), Shemya, AK connects to Columbia, MO. The premise is that observed weather in the Bering Sea region shows up in the CONUS 17-21 days later. Click on the Tweet below to see the larger images. (More predictive data related to BSR here)

Potential severe weather events ~May14 & ~May20 based on #BSR? What do you think @TriStatesWx and @jnelsonweather? pic.twitter.com/BWI50VQhy9

— Josh Herman (@OSNW3) May 3, 2015

JD suggests deep south and wants a harmonic relationship based on the Recurring Rossby Wave Train.

The PNA has been cycling ~32 days. (click on PNA) As far as the CONUS 500s, strong correlation exists in these windows as well. (click on REGIONAL+). Seems that fancy LRC length has been showing up in the REGIONAL+ the past 10 days too. Harmonics for everyone.

Getting to the point.

~May 20 going back ~32 days...
150416_rpts
150417_rpts
150418_rpts
150419_rpts

~May 20 going back ~LRC days...
150331_rpts
150401_rpts
150402_rpts
150403_rpts

~May 20 going back ~64 (nothing special)
150318_rpts
150319_rpts

The current frequency and harmonic certainly show a possibility. I know JN has been using late Oct as an analog for his version of the LRC lately. If you go back to Nov 16, it could connect to ~May 20, if you like averaging 185 days to find the wheelhouse prophecy. On a side note, going further into the future with the GFS, #chasecation2015 might get sucked into a vortex.

@islivingston @TriStatesWx @jnelsonweather let the hilarity ensue. See yellow box. #BSR #GFS :) pic.twitter.com/z0dtuAEBlX

— Josh Herman (@OSNW3) May 3, 2015

If there are any questions, comments, or suggestions on the material presented please let me know. Thanks for reading!

High Correlation in the Mid 30 Day Frequency

I am currently recording near 0.8 correlation in the mid 30 day frequency in Region 7. I plotted the 500s for two stations (KILX and KILN) in what I label Region 7 to show the pattern consistency. Region 7 consists of stations in the Ohio Valley (Central). This region has all stations reporting within the daily top 10 frequencies.





I would suspect that if someone analyzed CONUS 500 maps during this time frame there would be regions that do not match up within this frequency. Analysis suggests that a transient cycle frequency with multiple phasing cycles is taking place. Something similar to a standing wave. Not an all encompassing macro cycle that Lezak suggests. See the heat-map table for more.

Lezak tweeted me that his cycle range has been 43-50 days since 10/1. The chart below shows this range correlating well near the end of November and early December. The wheelhouse time when Lezak discovers his cycle length each year, then back dates it to when it "began" in October. The past few months his range of days hasn't been performing well.



Calculating the median and average of the top-10 daily frequencies yield results very close to what Lezak states as his cycle length range. Most recently the range has been 42-49 days. The chart below shows the range of the median and average since 8/1/14.



If there are any questions, comments, or suggestions on the material presented please let me know. Click on the images for a larger view. Thanks for reading!

A Random ERSL Map Verification Analysis

I am not sure how I will conduct auto verification for these ESRL map type projections. A random case below.

-Surface Temperature Anomaly
-Observed 20150308-20150312


-Surface Temperature Anomaly Forecast
-Initialized 20150215, 21-25 Day Forecast 20150308-20150312


This might look OK, but it is just one instance. I am archiving all of the past projections here. On a side note, I am currently working on auto verification for the Trends. It's a long way off yet, but it is in the works. All of this is an ongoing work in progress. If there are any questions, comments, or suggestions on the material presented please let me know. Thanks for reading!

A Transient Cycle Length with Multiple Phasing Cycles

So where was I? Ah yes, I dropped the high amplitude and frequency states (ISO short-term component, 10-20 days) from the CONUS output and plotted the 30-90 day correlations exclusively. More

The map below shows the RAOB stations that I collect data from. I don't collect Canada or Mexico and it seems I randomly do not collect in the deep south US. I blame it on laziness. I am super lazy.



My thoughts are not well organized or educated. They are likely difficult to follow. I will attempt to explain them in short detail. The image below shows which region the highest correlations stem from. The range is 0-1. The higher the value, the more high correlations stem from that region. Example; the daily analysis for 2/21 shows region 5 has 9% of it's stations reporting a top 10 value. Region 8 60% and region 9 67% of it's stations in the top 10. This correlation could be considered "east based", where this region shows the most correlation.



These correlations are charted in a heat-map like table form. The image below shows the first 21 days of February. The left most table is the top 10 cycle lengths, listed from left to right, 1 through 10. The table directly to the right is the corresponding correlation values. The 3 tables to the right are mode, median, and average of cycle lengths for the previous 30 days, since December 1st, and since August 1st. The entire table can be found here.



A quick analysis of the heat-maps suggest a transient cycle length with multiple phasing cycles taking place. Similar to a standing wave. If there are any questions, comments, or suggestions on the material presented please let me know. Thanks for reading!

Framework: Use current NOAA/ESRL Radiosonde Database to analyze large-scale upper atmosphere patterns in standing wave notation. Described specifically to harmonics, reflecting the temporal/transient behavior of the frequency wavelengths in correlation and relating Intraseasonal Oscillation to Mid-Latitude recurring weather patterns.

Goal: Forecasting skill of upper-air and surface weather trends.

Doug Heady Joins Gary Lezak at Weather2020

Great to read that Heady and Lezak are back together. Gary puts together a wicked map comparison in a recent blog. Can someone point out 4 similar features in the maps just 9 days prior to this latest map comparison? I have provided the map analogs below.





I find it fascinating what Gary and Doug see. If there are any questions, comments, or suggestions on the material presented please let me know. Thanks for reading!

I dropped the high amplitude and high frequency states and plotted the 30-90 day correlations exclusively

I dropped the high amplitude and frequency states (ISO short-term component, 10-20 days) from the CONUS output and plotted the 30-90 day correlations exclusively.



What is this junk? Back when I followed Lezak's Recurring Cycle and/or the Heady Pattern, comparing 500mb maps with the naked eye was the only way to find/follow the cycle. I quickly learned everyone has a different eye. Instead of looking for blessings from the cyclic masters I decided to seek the cyclic nature in data. Much like how they (LRC/HP) compare maps I decided to compare legitimate peer reviewed oscillations in a similar manner. Rather than look for a recurring cycle at ~50 days every couple days (or what is perceived as the cycle length) in maps, I formulate correlation of a 10 to 90 day range from sounding data on a daily basis. I find the dominant recurring patterns and plot the output.

A correlation example. Say I am finding the correlation for the date 9/1, for day 20 in the 10-90 day range. The data arrays that are used to find the correlation consist of data within the dates 9/1 thru 8/12 and 8/11 thru 7/22. I find the correlation for day 10 through day 90 and plot the highest correlations. This is what is plotted in the tables of the image above. The left table is the region* the top correlations in the middle table originated from, the middle table is the top 10 oscillation frequencies (aka "cycle" length) and the right table is the corresponding correlation value of the middle table. At the very bottom of the table is mode, mean, and average statistics as well.

*I categorize the CONUS RAOB stations in regions of climate. I simply label them region 1 through 9, see image below.


Framework: Use current NOAA/ESRL Radiosonde Database to analyze large-scale upper atmosphere patterns in standing wave notation. Described specifically to harmonics, reflecting the temporal/transient behavior of the frequency wavelengths in correlation and relating Intraseasonal Oscillation to Mid-Latitude recurring weather patterns.

Goal: Forecasting skill of upper-air and surface weather trends.

This is just quick summary that stemmed from this twitter thread. I will likely add to the entry in the future. If there are any questions, comments, or suggestions on the material presented please let me know. Thanks for reading!

LRC'ers using a different cycle duration? How can this be?

GL sees 41.



DH sees 62.

@OSNW3 Hey Josh, You can plot all day long. You can't beat my eye. We are 100% 62 days no doubt. The cycle started July 31st.

— Doug Heady (@DougHeady) December 10, 2014

JN sees 60-63.



I think the LRC/HP was a good starting point a handful of years ago. It helped me realize that things were cycling. Kinda like learning how to dribble a basketball and run at the same time. "Seeing" the "cycle/pattern" just isn't in my vocabulary anymore. I have done my share of research/experimenting based on the core principles of the LRC/HP. I need to keep moving forward.

GL sees 41. DH sees 62. JN sees 60-63. Personally I think it's great. Are they in harmony? Below is a chart showing correlations for each duration mentioned above.

CONUS Correlations


Some Regions are noisy. Coastal usually. Continental a different story. Region 3 is locked in. Closer to the recurring long-term longwave perhaps? (analysis)

Region 3 Correlations


Originally posted on the AccuWx ISO/Rossby/BSR/TR 2014-2015 forum. If there are any questions, comments, or suggestions on the material presented please let me know. Thanks for reading!

OSNW3 | WxClimate 2014-15 Ensemble

This blog post will act as a "help" page for the 2014-15 ensemble member descrptions.

Framework: Use current NOAA/ESRL Radiosonde Database to analyze large-scale upper atmosphere patterns and relate the tropical Intraseasonal oscillation to Mid-Latitude recurring weather patterns.

Goal: Long Range forecasting skill of surface weather trends.

Last year I integrated a three member ensemble seeking better resolution to the cycling weather pattern. The 2013-14 method was based on specific ideas of the LRC/Heady Pattern. Since that time I have discovered that the LRC/Heady Pattern are just names for a specific harmonic in the overall Rossby standing wave. Utilizing this new information I have revised the ensemble members for this year.

Member 1 (M1) scales the 10-30 and 60-90 day daily 500mb correlation to the 30-60 day Madden–Julian oscillation (MJO).

Member 2 (M2) scales the 10-30 day daily 500mb correlation to the 30-90 day tropical Intraseasonal oscillation (ISO).

Member 3 (M3) does not utilize MJO/ISO scaling. This is a 1 to 1 connection of the mid-latitude Rossby standing wave 10-90 daily 500mb correlation.

Doing this has disconnected my data from the LRC/Heady Pattern thought process. The 2014-15 data is ready for testing/review and can be found here.

Often times this blog is neglected. Many of my ramblings can be found in the AccuWeather ISO/Rossby/BSR/TR threads. Please check it out, go here for the 2014-15 version. If there are any questions, comments, or suggestions on the material presented please let me know. Thanks for reading!

Current Recurring Rossby Components and a 2014-15 LRC Duration Hint

Let's begin with last years dominant components. (01AUG13-31JUL14)



75-79
49-51
40-42
26-26
14-15

Requirements are such that the correlation equals 0.4 or greater for any day throughout the year. The chart below shows how using the dominant harmonic may prove to be solid guidance as the short-term, mid-term, and long-term will all have their days in the sun so to speak. With a broad brush, when one component is out of frequency the other components are not. Difficult to imagine that ~57 didn't show up at all.



Below is the average of each component in 2013-14. 10-20, 30-60, and 60-90 days.



We caught onto the LRC duration last year around this time. What about this coming season? 17-18 days are tops this morning (31AUG14). 74-76 days are right behind. However, through August, 58-60 days has been the most dominant. Likely to swerve in the coming weeks.



With the long-term likely to cycle negative soon, the short/mid-term should recover and produce new patterns for the upcoming season. Discussing this with some others, I agree we can use the 74-76 day to find the LRC duration and confirm with the short-term using standing wave harmonics. Some suggest taking the current short-term from the current long-term for this years LRC duration. Give or take a day of course.

If there are any questions, comments, or suggestions on the material presented please let me know. New framework, new website, new content will be live sometime around 01OCT14. Thanks for reading!

500mb Rossby Correlation

Finally had a breakthrough in my correlation programming method. I often wonder why these breakthroughs take so much time to come about. Why didn't I think of this before? A long story short, I can now correlate one days worth of soundings in four seconds. Not bad. Still room for improvement. Below are this years sounding correlations in chart format and table format.

500mb Rossby Correlation Chart 01SEP13-20JUL14 (larger chart)

(X = Duration, Y = Correlation Coefficient)

500mb Rossby Correlation Table 01SEP13-20JUL14 (larger table)

(Daily Top 10 Duration)

It is easy to depict that the dominant recurring oscillation is constantly evolving. All of the above jabber stems from THIS and THIS.

Sounding Domain (larger view)

(Keeping it local - OSNW3|WxClimate)

If there are any questions, comments, or suggestions on the material presented please let me know. Thanks for reading! More to come.

500mb Intraseasonal Oscillation Correlation - April through June

The correlation of the 500mb flow over the OSNW3 domain. Y axis is correlation coefficient, X axis is days. Analyzing the oscillations we can apply standing wave harmonics to find the recurring weather patterns based on the dominant ISO component.

500mb ISO Correlation Apr-Jun (larger chart)

ISO and Rossby; Harmony 101

An auto-discovery "cycle" duration. Obtainable by finding the dominant ISO component. On a daily basis I am doing just that. A while back I began following a specific feature and tracking it back through the long-term recurring Rossby.

Observations


Short-term ISO on Apr-13


Midwest 500mb Correlations Apr 1-30


What are these charts showing? The correlation of the 500mb flow over the Midwest. Y axis is correlation coefficient, X axis is days. I correlate each station on a daily basis looped 10-80 days backward in time to find the most positive correlation. The change is in days of the ISO translated to the mid latitude Rossby wave. Analyzing the oscillations we can then apply standing wave harmonics to find the recurring weather patterns.

Below is a KMCI MaxT reanalysis based on ISO (r=0.716) and LRC (r=0.017). Out with the old and in with the new?



If there are any questions, comments, or suggestions on the material presented please let me know. Thanks for reading! More to come.

Current Pattern and Prior Cycle Engagements

Give or take a day, this is where the current pattern engaged within each "cycle" aka the long-term ISO component, ~57 days.

20131008


20131203


20140130


20140329


Please attempt a stare & compare method analysis by moving the maps back and forth (click image). Currently my favorite way to get GFS guidance is via Tropical Tidbits. If there are any questions, comments, or suggestions on the material presented please let me know. Thanks for reading!

Seeking an Auto Discovery Cycle Duration - Part 3

I am putting the 500mb sounding correlations to use and seeking the recurring weather pattern. There is talk about severe weather approaching. The SPC probabilistic outlook for 2/20-21 is below. This feature has occurred before.



I want to track this feature back in time through the ISO. I start by seeking the dominant components in the correlations. The listed correlations are the greatest positive correl from the short and long term ISO components. Short-term is on the left, long-term is on the right. Give or take a day or two on each side, I see a consistent short-term component of 20-21 days in the FEB correls. I utilize standing wave harmonics and pin point a long-term component of 60-63 days. I see a consistent long-term component of ~50 days in the DEC correls.

FEB Correlations


DEC Correlations


Observations and Forecasts


Starting with the forecast date of 2/20/14 I subtract ~60 days and find 12/21/13, I then subtract ~50 days from 12/21/13 and find 10/31/13. Simply put, from within the correlations we are seeing the movement of ISO components in relation to the recurring Rossby Wave Train and cyclic weather patterns. If there is interest in the data it can be found here. Another step closer.

Click for FEB observations
500-Millibar Height Contour Map 20140220
Storm Prediction Center Storm Reports 20140220

If there are any questions, comments, or suggestions on the material presented please let me know. Thanks for reading!

Seeking an Auto Discovery Cycle Duration - Part 2

Stemming from seeking an auto discovery cycle duration I started running the sounding correlations on a daily basis. They now begin from the current day and loop 10-80 days backward in time.

When I first began correlating the soundings I started the sample period on 10/1/13. I recall ~15 days being the short-term component from those runs - a consensus on positive correlation from all stations. Beginning from the current day the short-term 10-20 day ISO still holds the top positive correlation, but now ~20 days has the consistent positive correlation. It seems to me that the short-term cycle of the ISO stands out as the dominant component this season.



I am very interested to see what happens to this correlation as the days move along. Should I expect the wave to continue on or will it roll back towards ~15? Am I seeing a seasonal ISO in the 70-80 day range? I should loop to 90 or beyond to get a better glimpse? It's a work in progress, but I think we are very close to finding an index to aid in calculating a daily cycle duration based on a primary harmonic. More to come.

If there are any questions, comments, or suggestions on the material presented please let me know. Thanks for reading!

OSNW3 Trend Comparison to the Jeremy Nelson LRC Based Long Range Forecast

I wanted to analyze the OSNW3 long range numbers for February in conjunction to Jeremy Nelson's forecast for southeastern WI. Jeremy is a meteorologist at WISN12 in Milwaukee who uses LRC techniques to long range forecast. His "February forecast and beyond" can be found at this link.

I will highlight some pieces of the forecast.

1. "The pattern looks to stay busy with a chance of snow around February 4-5 and 7-8 [...] some of the storm systems will have more moisture to work with" - February 1-7 Section (Black)

2. "Following each snow chance a quick hit of cold air will sweep across southeastern Wisconsin. More below zero lows are likely this month. Temperatures should moderate and possibly move above average late in week two of February [...] The average/above average temperatures should last 1-3 days" - February 8-15 Section (Orange)

3. "The middle of the month should see the return of one of the biggest precipitation producers for southeast Wisconsin [...] As this feature returns around February 14-17 another storm system is possible impacting our area" - February 16-28 Section (Blue)

4. "Closing out the month a moderation in temperatures is possible around February 20-22 before another surge of cold air around Feb 23-25. A couple chances for snow showers also possible the final week of February" - February 16-28 Section (Red)

The image below is a snap shot of the Milwaukee, WI trend from the OSNW3|WxClimate website for February. I highlighted the points mentioned above in the image. For Jeremy's entire forecast be sure to click the blog link above. Click here for more information on the image below, which will provide an interactive experience for the Brewer's home opener Jeremy mentions in his forecast as well.



If there are any questions, comments, or suggestions on the material presented please let me know. Thanks for reading!

OSNW3|WxClimate Trends 2013-14

Consider the Rossby Wave, a wave train carrying with it repeating intervals of atmospheric conditions.

Rossby Wave


2013-14 marks the fourth year that I have been following the recurring long term long wave troughs and ridges. Since my first days of learning of them from Jeremy Nelson, with the likes of Gary Lezak, Doug Heady, and the AccuWx Forum crew, my curiosity piqued. To make this story short, I am still here, but now I am looking at this LRC stuff with hopes of forming a reliable point of view.

Reconsider the basics of LRC. Two of the main points can be described well with the opening statement of this blog entry.

* Long term long-wave troughs and ridges become established and also repeat at regular times within the cycle. These dominant repeating features are a clue to where storm systems will reach peak strength, and where they will be their weakest.
* The LRC isn’t just one long-wave trough, storm system, or ridge. It is a sequence of troughs and ridges that are cycling across the Northern Hemisphere.

But, it is these last two that cause confusion and are a hurdle.

* This unique weather pattern sets up every year between October and November
* The weather pattern cycles, repeats, and continues through winter, spring and into summer.

It was Scott Metsker, an advocate of critical thinking, who turned me onto looking at the cyclic patterns with a more focused, data driven, scientific approach. I was introduced to Intraseasonal Oscillations, Boreal Summer Monsoons, Northward Propagation Mechanisms, Rossby Waves, Standing Wave Harmonics, etc. These items soon became embedded into my daily operational thoughts of the cyclic patterns.

How do I get over this hurdle of when the weather pattern sets up? How does this annual unique cycling weather pattern come to life? A simple answer can be obtained by connecting the dots mentioned above. Choosing the correct source material will open a mind blowing path to "organic forecasting" based on the planets cyclic atmospheric behavior.

But, before I ever got to the hurdle, I attempted to prove that the first couple points could be utilized to project weather and climate conditions. The last two years I have developed a 'model', which is the main point to this entry, that attempts to forecast temperature and precipitation weeks and months into the future. My equations are simple and the 2013-14 version is ready to be tested and reviewed. I am looking forward to the results, no matter the outcome.

If there are any questions, comments, or suggestions please let me know. Thanks for reading!