Analogs

[bluewave]

analogs...

[tacoman25]

The thing about pattern analog matches is that the results can really vary based on what you decide to look at. For instance, if I look at -ENSO Septembers that had a similar U.S. pattern to what we're seeing in 2011, here are the years.

If we roll forward to October, a complete pattern reversal is the theme.

And then again to November, yet another reversal.

And finally the following winters.

[bluewave]

The thing about pattern analog matches is that the results can really vary based on what you decide to look at. For instance, if I look at -ENSO Septembers that had a similar U.S. pattern to what we're seeing in 2011, here are the years.

If we roll forward to October, a complete pattern reversal is the theme.

And then again to November, yet another reversal.

And finally the following winters.

I like to look more to the October 500 mb pattern for winter clues since it's difficult to correlate September temperatures to October and the winter.

So far this September is running a bit warmer to the north than the September analogs that you posted.

[Lee]

The thing about pattern analog matches is that the results can really vary based on what you decide to look at. For instance, if I look at -ENSO Septembers that had a similar U.S. pattern to what we're seeing in 2011, here are the years.

If we roll forward to October, a complete pattern reversal is the theme.

And then again to November, yet another reversal.

And finally the following winters.

Tacoman, I see you have 1949-50 as an analogue. Hmmm?? Just curious to whether you intended that one or the more preferred by others, '50-51 ? You have just 2 mild years in there('74-75, '49-50) . Take away '49-50, replace it with '50-51 and you have a significantly different outcome particularly in the SE. The 74-75 one , unfortunately looks pretty legit as a decent analogue.

[tacoman25]

I like to look more to the October 500 mb pattern for winter clues since it's difficult to correlate September temperatures to October and the winter.

So far this September is running warmer to the north than the September analogs that you posted.

The same basic pattern (warm west of the Rockies, cool east) exists, though, and looks to be strengthened through the end of the month per models.

[tacoman25]

Tacoman, I see you have 1949-50 as an analogue. Hmmm?? Just curious to whether you intended that one or the more preferred by others, '50-51 ? You have just 2 mild years in there('74-75, '49-50) . Take away '49-50, replace it with '50-51 and you have a significantly different outcome particularly in the SE. The 74-75 one , unfortunately looks pretty legit as a decent analogue.

I'm not saying 1949-50 is a great analog (it just happens to be among the -ENSO years that has matched 2011's Sep pattern pretty well), my point was just that you can come up with a lot of different years for pattern matches, depending on what criteria you use.

I agree that overall, 1974-75 is a better analog than 1949-50.

[Lee]

I'm not saying 1949-50 is a great analog (it just happens to be among the -ENSO years that has matched 2011's Sep pattern pretty well), my point was just that you can come up with a lot of different years for pattern matches, depending on what criteria you use.

I agree that overall, 1974-75 is a better analog than 1949-50.

I know what you're saying in general. As far as 49-50 versus 50-51 Septembers with the parameters you used, 50-51 is better precip wise with 49-50 better temp. i.m.o. So, those alone being about equal. Now, if you add in second year Nina versus 1st year....lol.

Sept.49-50 Temp.

http://climvis.ncdc....-map-display.pl

Sept. 49-50 precip

http://climvis.ncdc....-map-display.pl

Sept. 50-51 Temp.

http://climvis.ncdc....-map-display.pl

Sept. 50-51 precip.

http://climvis.ncdc....-map-display.pl

Now, just for fun, comparison of winters 49-50 versus 50-51:

dec-feb. 49-50

http://climvis.ncdc.noaa.gov/cgi-bin/cag3/state-map-display.pl

dec-feb. 50-51

http://climvis.ncdc.noaa.gov/cgi-bin/cag3/state-map-display.pl

[bluewave]

The same basic pattern (warm west of the Rockies, cool east) exists, though, and looks to be strengthened through the end of the month per models.

Sure, you can see an overlap between different groups of analogs where they agree on certain elements of the pattern.

I was just trying to focus on whether or not we see a return to blocking after a break this month. It was not meant to

get that much into the specifics of some of the other teleconnections that we have been seeing or may see.

[tacoman25]

Sure, you can see an overlap between different groups of analogs where they agree on certain elements of the pattern.

I was just trying to focus on whether or not we see a return to blocking after a break this month. It was not meant to

get that much into the specifics of some of the other teleconnections that we have been seeing or may see.

I hear you. Just a reminder that depending on what criteria you use, there will be different analog years that pop up. Not that you were doing this at all, but I think it's easy to get too focused on one factor and then ride analogs mainly based on that - whether it be NAO, pattern matches in a certain area, PDO, ENSO, AO, QBO, etc. The hardest part of long range forecasting is determing what emphasis to place on the different factors, imo.

[BethesdaWX]

The thing about pattern analog matches is that the results can really vary based on what you decide to look at. For instance, if I look at -ENSO Septembers that had a similar U.S. pattern to what we're seeing in 2011, here are the years.

If we roll forward to October, a complete pattern reversal is the theme.

And then again to November, yet another reversal.

And finally the following winters.

Comparing patterns in the United States during September isn't the same thing as a higher lattitude connection at 500mb, the NAO isn't really "driven" like the patterns in the United States alone, (By the same factors, that is) how many of those years had warm Septembers in Europe?

ENSO is a global phenonemon, with many intricate processes involved that will alter it's effect/forcing on the climate system, atmosphere + oceans, so it is a silly cmparison, because it is those factors that alter ENSO that are being discussed by Bluewave. it's not just an ENSO base and fake "mirror" patterns that obviously can result from different forcings, the eastern us troughing in september in your years did not result from the same forcings, and many of those yrs did not have the warm 500mb anomaly over Europe which is likely connected to a higher lattitude cycle.

[tacoman25]

Comparing patterns in the United States during September isn't the same thing as a higher lattitude connection at 500mb, the NAO isn't really "driven" like the patterns in the United States alone, (By the same factors, that is) how many of those years had warm Septembers in Europe?

ENSO is a global phenonemon, with many intricate processes involved that will alter it's effect/forcing on the climate system, atmosphere + oceans, so it is a silly cmparison, because it is those factors that alter ENSO that are being discussed by Bluewave. it's not just an ENSO base and fake "mirror" patterns that obviously can result from different forcings, the eastern us troughing in september in your years did not result from the same forcings, and many of those yrs did not have the warm 500mb anomaly over Europe which is likely connected to a higher lattitude cycle.

Please explain how the patterns observed over Europe (the anomalies this month in the U.S. are certainly also connected to 500mb patterns, in part due to what is going on at the higher latitudes) are more important than the patterns in North America.

It sounds like you are basically saying to ignore ENSO state when analyzing pattern analogs, which seems pretty questionable to me.

[BethesdaWX]

Please explain how the patterns observed over Europe (the anomalies this month in the U.S. are certainly also connected to 500mb patterns, in part due to what is going on at the higher latitudes) are more important than the patterns in North America.

It sounds like you are basically saying to ignore ENSO state when analyzing pattern analogs, which seems pretty questionable to me.

???

Thats not what I said. First, you're essentially saying that La Nina winters that feature a trough over the Eastern USA in September (all after 1949 for whatever reason) is a result of the same forcing in every case, thus you blend them together, along with the following Octobers, when in reality if you were to un-blend the mean, you'd see a wild signal for this very reason. Come on.

No one is Ignoring ENSO state, but the NAO state does not depend on the ENSO state, you can Have La Nina with a +NAO or a -NAO, in a low sun activity period a -NAO is more likely than a +NAO.

You are completely off topic, the point here is that a warm pattern over Europe in September tends to be followed by a -NAO/blocking pattern a month later, it doesn't have to turn out that way, but it tends to happen, and this winter has a higher likelyhood of a -NAO than the average winter would.

[tacoman25]

???

Thats not what I said. First, you're essentially saying that La Nina winters that feature a trough over the Eastern USA in September (all after 1949 for whatever reason) is a result of the same forcing in every case, thus you blend them together, along with the following Octobers, when in reality if you were to un-blend the mean, you'd see a wild signal for this very reason. Come on.

No one is Ignoring ENSO state, but the NAO state does not depend on the ENSO state, you can Have La Nina with a +NAO or a -NAO, in a low sun activity period a -NAO is more likely than a +NAO.

You are completely off topic, the point here is that a warm pattern over Europe in September tends to be followed by a -NAO/blocking pattern a month later, it doesn't have to turn out that way, but it tends to happen, and this winter has a higher likelyhood of a -NAO than the average winter would.

No, I'm not. I'm just showing how looking at one factor and then rolling that over to the following months can result in different outcomes, based on the criteria chosen. If you want to look just at European blocking patterns in late Sep, you will get one result. Look just as summer NAO you'll get another result. Look just as ENSO another. PDO another. Pattern matches to the U.S. during any given time frame, another. See what I'm saying?

[CoastalWx]

FWIW the European weeklies try to develop blocking over the Davis Straits area by mid October. Tough to saw where it originates, but it looks like anomalous ridging in Canada moves into the NAO region during that time.