Pages

Friday, November 3, 2017

Bering Sea Rule Case Study Moment

Hello everyone!

I noticed a tweet from the BAMwx twitter account which reminded people that even though a -WPO configuration will provide for short term cooler weather, the long term outlook isn't as pretty.



As a reminder, here are the EPO/WPO combinations at the 500mb level.

 

Upon seeing the tweet I went ahead and looked up the Bering Sea Rule aspect 19 days earlier below.
Shemya Island is our Springfield, MO and Kodiak Island is our Bangor, ME correlations. So...let's place the two ESRL maps side by side...


Note how in the CONUS ESRL map has AN (Above Normal) heights just North of the Dakota's diminishing as it works South and the BN (Below Normal) heights to the SE of Maine. Those mimic what happened 19 days earlier in the Bering Sea.

Throwing in my SOI (Southern Oscillation Index) research and this is what I came up with 19 days after a 3 day drop of 20 points! Yes...I know it was 17.15, but it also mimics the one day drop of 10 points in our research with AN over the top and BN off the East Coast.

20pt drop in three days

10pt drop in one day

Thursday, October 19, 2017

Bering Sea Rule and East Asia Rule Explanations

Hello all,

A special request has been made that I explain the Bering Sea and Typhoon Rules to the masses.  Since the Bering Sea Rule is the first "heads up" in a pattern I will begin with it.

The Bering Sea Rule was a pattern that I recognized back in 2011/2012 while performing my moderator duties at Accuweather.com.  Joe Bastardi made a claim that after some monster storms of 1950 and 1974 in the Bering Sea, that within 3 weeks of those storms we saw monster storms for the East and left it at that.  Sadly enough, I was playing firefighter in the thread below and the corresponding OBS thread because his hype didn't come true.

Forecast thread:

OBS thread:


Here is a picture that he tweated showing the 50/74 storms in question.



...and the wiki on both storms...



Over the years, I have amassed multiple post where I have correlated the above to a pattern that follows.  That is where JB got himself into trouble.  He was attempting to get people involved in the hype of some monster storms instead of looking at the pattern in general.  One of my favorite "JDism's" on the forum is "It's all about the pattern, and knowing the right pattern is what it's all about!" or "We sniff out the pattern, specifics come later!"

Is there any research outside of me that has looked deeper into this...not that I can find. In fact, if one performs a google search on this, you will find two good friends of mine who have typed up blogs about the subject based on what I've taught them.

OSNW3 Blog

http://www.scmweather.com/

As for the Typhoon Rule...this rule has been around for decades.
http://www.dtic.mil/dtic/tr/fulltext/u2/a277210.pdf

It became popular by Joe Bastardi who used it while at Accuweather.  The rule is quite simple and applies all year round!  If a typhoon recurves as it approaches Japan, whether it be too late to miss the Korean Peninsula or completely miss Japan OTS, the weather in the Eastern US is teleconnected 6-10 days later.  How does this apply all year you ask?  Easy way to think about it is that a cold front is forcing the typhoon to recurve one way or the other.  Cold fronts aren't seasonally dependant like typhoons are...they happen all year round!  The same applies if the typhoon heads into Mainland China.  That translates to a ridge blocking the typhoon from recurving towards Japan and the Eastern US will have a heat ridge develop in 6-10 days as a result.


A few resources that I use to help me with the pattern recognition techniques...

WPC 3-7 Day Lower 48 Forecast


WPC Alaska Day 4-8 500mb Forecast


Ocean Prediction Center Pacific Tab


Weather Online Expert Charts 500mb
GFS:
Euro:

Accuweather Pro Animator
North Pacific View

@crankywxguy maps

Sunday, October 15, 2017

Where Do We Go From Here?

Hello everyone!

It's been a pretty wild summer for me and I can imagine with the other research projects I have going it will only get worse.


  1. The weather.kopn.org website has been dissolved. The radio station is under new management and have developed a new website dedicated to our independent contribution in the community. That being said, a new "weather" section will be up and running soon.
  2. The SOI research is coming along nicely with 33 folders dedicated to ENSO/Season/SOI Fall or Drop. Within those folders I have created ESRL maps spanning 25 days from a 20 point and 10 point rise or drop.
  3. I'm still working on a new Pacific index that could help with ENSO relationships to the lower 48 weather.
  4. Josh's RRWT, Recurring Rossby Wave Train, submissions to the Forecast Rodeo have been doing very well.  There were 118 teams to start with and now they have narrowed it down to the top 5 at the half-way mark. We are optimistic that once the fall and winter weather starts hitting the forecast regions that the RRWT will shine more than the other competitors. He is outperforming the "benchmarks" of 'Damped Persistence' and the 'CFS' in both Week 3-4 & 5-6 Precipitation and beating 'Damped Persistence' in Week 5-6 Temperatures! I've been attempting to contribute via a different sub-seasonal method that had been doing very well utilizing CPC 6-10 & 8-14 day analogs. The issue is that ESRL maps only provide 'Precipitation Rate' or 'Precipitable Water' as options and not the 'Accumulated Precipitation' that we need for the contest. That conversion is giving me fits to say the least. 
  5. Twitter handle @crankywxguy has taken over the creation of our BSR maps and has done an amazing job with it! 
So...as you can see above, we have quite a few items that are coming along and are looking forward to their results as time comes along!

Tuesday, May 2, 2017

Forecast Rodeo Comparisons

Hello everyone!

The team received this email today regarding the Forecast Rodeo

I am looking forward to the 'official' scores mid-month, but in the meantime, let's look at how we fared.

Here are the forecast that we provided the rodeo with on 4/04/17 for the period of 4/18-5/1.

Accumulated Precipitation in millimeters

Average Temperature in Celsius

A few different looks at the results...

Precipitation in inches

Temperature in Fahrenheit

Now...since the Western Region Climate Center doesn't extend into the Plains, I've included the complete US maps below. Otherwise, I'd have to include 2 of each from the High Plains & Southern regions.



Due to the difference in data, mm vs in & (C) vs (F), note the patterns instead...unless you are fine with converting each.

The evaluation criteria information for said rodeo is listed below...
Evaluation criteria
Forecast skill will be evaluated for temperature and precipitation separately since the drivers responsible for prediction of these variables are different and the subsequent skill level is also expected to be different. Moreover, the 15-28 day and 29-42 day periods will be evaluated individually for similar reasons. Winning forecasts must outperform CFSv2 and damped persistence forecasts (see definitions below). Specifically, skill will be evaluated individually for temperature and precipitation for weeks 3-4 and weeks 5-6 as the highest skill over the competition domain, averaged over the entire competition time period. To be prize eligible, Solvers must also demonstrate historical skill of statistical significance that is equal to or greater than that of the CFSv2 through submission of a hind-cast analysis described below.
So, with the above in mind, let's look at the CPC Weeks 3&4 forecast. Yes, I'm familiar that it's not the CFS Week 3&4, but I couldn't grab the archived forecast.

Precipitation Anomaly Forecast

Temperature Anomaly Forecast


The resulting anomalies...



Saturday, April 15, 2017

February Verification

Hello everyone! I know, I know..."Joe, you are quite a bit late on the February verification post!" There were so many different routes that I couldn't decide to take on this that it left me confounded.

One thing that I'd like to touch base on was the 3 day ROC scores for the Northeast during February. Astonishingly good would be the best way of describing it!

 As you can see above, not only did the 3 day ROC of the BSR nail the pattern overall during February, but it literally had a 1:1 correlation on multiple days!

The 7 day pattern correlation of the ROC shows the high correlation even better with score verification.

The first 5 days of the month showed this as it pertained to the BSR 500mb map.
 As you can see above, the Western ridge/Eastern trough was foretold. The trough was focused over Quebec, extending into most of the Eastern US, and the ridge was focused over the Intermountain West.
The above graphic from the ESRL daily composite site, with a focused point of the US, shows the ridge verified SE of the BSR while the trough apex was split between Newfoundland and James Bay. The ridge extending to the SE coast nullified the minimal BSR below normal look South of the Mason-Dixon line.


The next event I will bring up is the February 7th, 2017 severe weather that hit the Gulf Coast. The idea which the BSR started to paint was one which had a system develop over the lee of the Rockies and move East affecting the gulf region.


The first "official" hint on the organicforecasting.com BSR section was January 16th, 2017 with the 96hr OPC forecast overlay.


 While these overlays were produced on January 21st, 2017 based on 500mb analysis.




Almost immediately afterwards, we saw the potential for a strong storm to hit the Northeast via the next system based on the 96hr OPC forecast overlays starting on the 18th of January, 2017.




These maps are based on 500mb analysis once again.




Compare the maps for February 10th to what the Weather Prediction Center had for Days 5 & 7 forecast of February 9th on the left side of the next two pictures!

 Day 7 Forecast vs Verification
Day 5 Forecast vs Verification

On the Accuweather.com forums we have "adopted" a naming strategy to storms which the BSR nails..."Dark Knight Blizzard" as my nickname is the "Dark Knight" dating back to the January 20th, 2016 storm. This is the forecast thread, while here is the OBS thread for the DKBIII storm.

I will reference something that was brought up in the weatherboy.com article dated February 12th, 2017.

This is a warm pattern upcoming for much of the U.S., especially east of the Rockies. In the 10 day period ending around the 22nd of this month, many areas of the country will be 10 degrees above normal with some as high as 20 degrees above normal. Though New England will be the slowest to warm, after the current blizzard departs, temperatures will soar even there to well above normal readings about a week from now. While they warm slowly in New England, temperatures in the Midwest, Great Lakes and South will be well above normal for much of this upcoming week.
“A strong area of high pressure over the North Pacific near the Gulf of Alaska retrograded, meaning it headed west, instead of the normal east that most weather systems take (in late January) and this teleconnected to a ridge of high pressure off the Southeast coast heading west and dominating the eastern US,” explains Renken.  This type of weather pattern encourages a strong west to east flow across the Lower Forty Eight  and will combine with the much stronger than normal Pacific jetstream to allow warmth to dominate.

Below you can see the Des Moines, IA data for the period of February 1st-22nd. I placed the 10 day period from the 12th-22nd in red squares.

Even though I didn't see the departures get larger than 20° above normal, the pattern was foreseen and warned about. I suspect that the raging Pacific jet that I spoke about added extra moxie to the pattern.



While the pattern indicates a warm stretch of weather, a change is also likely around the 20th of the month. “This very well could be another severe weather outbreak with the Southeast with the Gulf Coast states being the area seeing the greatest threat for severe weather,” Mr. Renken stressed.

Even though the outbreak didn't come true due to mesoscale features not lining up, the potential was there and foretold.

Here are the reports from the 19th.

The SPC outlooks for the remaining days.



BSR map created on February 3rd, 2017. Note the ULL over West Texas, the corresponding ridge over the Deep South, another ULL off the SE coast and the Mid-Atl.

The West Texas ULL hasn't closed off yet in the OBS, but the ridge is created ahead, ULL between Bermuda and the Bahama's, and the last ULL over Nova Scotia.

A look at the BSR ESRL generated map for the 18th-22nd.

This is the corresponding OBS at 500mb.
This storm will be followed by a change to more normal temperatures.  Mr. Renken explained, “The stronger than normal jetstream slamming into California has been the overwhelming meteorological factor in our mild winter so far. And there is no sign of this factor going away. Plus Canada has been running way above normal temperature-wise. Combining these and other factors, I don’t think a pattern of sustained cold in the East will exist.”  After the stormy period that’ll set-up around February 20-22  and into March, the BSR indicates a volatile back and forth temperature pattern with some cold snaps but the cold is generally muted.

The BSR had this for the last 4 days of February.






Here are the last 4 days of February...


 Boston, MA data showing the cold was "muted" as predicted. That being said, going from 25° above normal to 7° above normal is a nice cold front that I'm sure those in the region were happy to see.

These are the 3 day ROC charts for the Mountain and Central regions.

The Mountain 3 day ROC had the pattern, but not the magnitude during the same period.

While the Central 3 day ROC was off by 3 days in the recovery.

So, overall, I'd say that the BSR did quite well during the month of February!