ENSO Updates (2007 thru 2023)

[Kingarabian]

With 3.4 having warmed back up to +0.5 for last week and with signs of further warming being evident, I see almost no way that June as a whole will end up at the +0.2 of the POAMA. It looks more like +0.4 to +0.6 is likely. Currently it's near +0.5 MTD. So, that means POAMA's near La Nina down the road is in even more in question.

June will end up with a -SOI for June. Currently it is at -11 MTD. I'm now roughly estimating June of 2017's SOI comes in at -7 based on these SOI's guesstimates:

6/20: -4; 6/21: -2; 6/22: -5; 6/23 -10; 6/24: +4 (would stop -SOI streak at an impressive 23 days); 6/25; +19; 6/26: +8; 6/27: +1; 6/28: +3; 6/29; 6/30: -3

A -7 in June is of no more than moderate significance as far as El Nino prospects are concerned. A -7 (or lower) in July, especially following a -7 in June, would be of much more significance as far as El Nino's chances are concerned.

Based on these same projections, Darwin's SLP for June will come in near 1013.4. This is above the longterm average of 1012.9 for all ENSO and is actually near the longterm average for strong Nino's. Weaker ones have averaged only near 1013.0. However, I've already determined that a 1014 was needed for strong statistical significance as far as El Nino prospects are concerned.

In summary, whereas there is some El Nino leaning for this year based on June of 2017's SOI and Nino 3.4 SST anomalies, I feel that we really need to see how July is going to be SOIwise to get a real good feel.

Good post, Larry. Although I'm sure your analysis was based on the 12Z Euro, I see a considerably negative SOI compared to the numbers you posted. Here are the pressures I derived from the 00z Euro:
Note: I rounded the Darwin pressure down and the Tahiti pressures up to remove any bias, and to account for the raw pressures coming in higher at Tahiti and lower @ Darwin.

Code: Select all

00z Euro Pressures  June 20 2017

Hr: Darwin   Tahiti
24  1015     1013
48  1015     1010
72  1014     1012
96  1013.50  1013
120 1013.50  1013.50
144 1014.50  1013
168 1015     1013
192 1015     1012.50
216 1016     1012
240 1016     1012.50

After 120hrs confidence is lower in MSLP forecasts no matter what model is used, but we can say with confidence that the SOI should be more negative than positive in at least the next 3 days.

Thanks, King. There is a problem with using just the 24 hour pressure progs you presented because they are fully based on 0Z each day. 0Z is both near the highest SLP of the day at Darwin as well as near the lowest SLP of the day at Tahiti. So, using only 0Z biases the SOI progs far in the -SOI direction for those two reasons (probably on the order of -22 to -25). That's why the 24 hour progs based on the 12Z Euro run are always going to look less bullish for a -SOI overall than that for the adjacent 0Z run. IF you must use just 24 hour progs, the 12Z run will almost always be closer to reality than the 0Z run. (A much more minor point is that I averaged the 12Z GFS and the 12Z Euro as opposed to just the 12Z Euro. By the way, I averaged over 6 hour intervals.)

But if LongPaddock uses 24 hr averages, shouldn't it be fine to do the same with the models? Or they're not reliable past 12hrs?

For 6/20, LongPaddock came in with -7, 1012.53 at Tahiti, 1012.25 at Darwin.

So taking a look at the pressures at Tahiti from the high resolution Euro which has 6hr intervals,

+6hr = 1015
+12hr = 1013.4
+18hr = 1016
+24hr = 1013

Looking at the past three days as well, +12 and +24 line up very closely with +12 occasionally coming slightly higher.

[LarryWx]

Good post, Larry. Although I'm sure your analysis was based on the 12Z Euro, I see a considerably negative SOI compared to the numbers you posted. Here are the pressures I derived from the 00z Euro:
Note: I rounded the Darwin pressure down and the Tahiti pressures up to remove any bias, and to account for the raw pressures coming in higher at Tahiti and lower @ Darwin.

Code: Select all

00z Euro Pressures  June 20 2017

Hr: Darwin   Tahiti
24  1015     1013
48  1015     1010
72  1014     1012
96  1013.50  1013
120 1013.50  1013.50
144 1014.50  1013
168 1015     1013
192 1015     1012.50
216 1016     1012
240 1016     1012.50

After 120hrs confidence is lower in MSLP forecasts no matter what model is used, but we can say with confidence that the SOI should be more negative than positive in at least the next 3 days.

Thanks, King. There is a problem with using just the 24 hour pressure progs you presented because they are fully based on 0Z each day. 0Z is both near the highest SLP of the day at Darwin as well as near the lowest SLP of the day at Tahiti. So, using only 0Z biases the SOI progs far in the -SOI direction for those two reasons (probably on the order of -22 to -25). That's why the 24 hour progs based on the 12Z Euro run are always going to look less bullish for a -SOI overall than that for the adjacent 0Z run. IF you must use just 24 hour progs, the 12Z run will almost always be closer to reality than the 0Z run. (A much more minor point is that I averaged the 12Z GFS and the 12Z Euro as opposed to just the 12Z Euro. By the way, I averaged over 6 hour intervals.)

But if LongPaddock uses 24 hr averages, shouldn't it be fine to do the same with the models? Or they're not reliable past 12hrs?

Using 24 hour averages is absolutely fine since that's what LongPaddock supposedly does and is what I'm approximating in my estimates by looking at the average of 0Z, 6Z, 12Z, and 18Z maps. The problem is that you're not using 24 hour averages. You're just taking one point in time each day in this case, 0Z. You're not also doing 1Z, 2Z, 3Z......23Z. At Darwin, the normal diurnal pattern's highest SLP just happens to be at/near 0Z. At Tahiti, the normal diurnal pattern's lowest SLP just happens to be at/near 0Z. That means that using just the 0Z SLP of each day causes a double whammy -SOI bias (something like -24 to -25 in June)!

At Darwin, I've found average diurnal SLP range to be near 4 mb. At Tahiti, I've found average diurnal SLP range to be near 3 mb. So, the 0Z SLP for Darwin is on average ~2 mb higher than the 24 hour average, and the 0Z SLP for Tahiti is on average ~1.5 mb lower than the 24 hour average. 2 + 1.5 = 3.5 mb of bias toward a -SOI. Each mb equals ~7 SOI points in June. So, 3.5 x 7 = ~24.5 of -SOI bias if you use just the progs for hour 0Z, which is what the 24 hour intervals of the 0Z Euro run maps show. Does this make sense?

[Kingarabian]

Thanks, King. There is a problem with using just the 24 hour pressure progs you presented because they are fully based on 0Z each day. 0Z is both near the highest SLP of the day at Darwin as well as near the lowest SLP of the day at Tahiti. So, using only 0Z biases the SOI progs far in the -SOI direction for those two reasons (probably on the order of -22 to -25). That's why the 24 hour progs based on the 12Z Euro run are always going to look less bullish for a -SOI overall than that for the adjacent 0Z run. IF you must use just 24 hour progs, the 12Z run will almost always be closer to reality than the 0Z run. (A much more minor point is that I averaged the 12Z GFS and the 12Z Euro as opposed to just the 12Z Euro. By the way, I averaged over 6 hour intervals.)

But if LongPaddock uses 24 hr averages, shouldn't it be fine to do the same with the models? Or they're not reliable past 12hrs?

Using 24 hour averages is absolutely fine since that's what LongPaddock supposedly does and is what I'm approximating in my estimates by looking at the average of 0Z, 6Z, 12Z, and 18Z maps. The problem is that you're not using 24 hour averages. You're just taking one point in time each day in this case, 0Z. You're not also doing 1Z, 2Z, 3Z......23Z. At Darwin, the normal diurnal pattern's highest SLP just happens to be at/near 0Z. At Tahiti, the normal diurnal pattern's lowest SLP just happens to be at/near 0Z. That means that using just the 0Z SLP of each day causes a double whammy -SOI bias (something like -24 to -25 in June)!

So it wouldn't be enough to average out the the 6, 12, 18, 24 hour intervals?

Say I looked at the Euro run, and at Darwin @ 00, it initiated with a pressure of 1015mb because that' just so happened to be when the highest pressure of the day is. So after that, at the 6th hour, it goes back down to 1013mb, and at the 12th hour it goes up to 1014mb, and at the 24hr it comes in at 1013mb. Is it not enough to average out those intervals? Is a four point sample size from a day sufficient enough to get a ballpark estimate on what the 24hr average would be? For example with what I did for the June 20th based on the 00z Euro for Tahiti:

Code: Select all

Tahiti
+6hr = 1015
+12hr = 1013.4
+18hr = 1016
+24hr = 1013

Note: I know that these numbers were nowhere near what LongPaddock had and what the raw values are.

So when I take into account the lower bias of -1.5 that you calculated:

Code: Select all

Tahiti
+6hr = 1013.5
+12hr = 1012
+18hr = 1014.5
+24hr = 1011.5[

Which would = to 1012.875, almost close to what LongPaddock had at Tahiti.

Code: Select all

Darwin
+6hr = 1012
+12hr = 1013
+18hr = 1012
+24hr = 1013.5


Tacking on the +2mb bias @ Darwin:

Code: Select all

Darwin
+6hr = 1014
+12hr = 1015
+18hr = 1014
+24hr = 1015


Which would = to 1014.5 @ Darwin, a higher than LongPaddocks 1012.25.

At Darwin, I've found average diurnal SLP range to be near 4 mb. At Tahiti, I've found average diurnal SLP range to be near 3 mb. So, the 0Z SLP for Darwin is on average ~2 mb higher than the 24 hour average, and the 0Z SLP for Tahiti is on average ~1.5 mb lower than the 24 hour average. 2 + 1.5 = 3.5 mb of bias toward a -SOI. Each mb equals ~7 SOI points in June. So, 3.5 x 7 = ~24.5 of -SOI bias if you use just the progs for hour 0Z, which is what the 24 hour intervals of the 0Z Euro run maps show. Does this make sense?

Certainly makes sense. But with the bias taken into account, it should've been a -4 on June 20th, as you've mentioned in your previous post. But LongPaddock had a -7.

[LarryWx]

Let's get back to what you originally said that got this discussion going:

"Although I'm sure your analysis was based on the 12Z Euro, I see a considerably negative SOI compared to the numbers you posted. Here are the pressures I derived from the 00z Euro:
Note: I rounded the Darwin pressure down and the Tahiti pressures up to remove any bias, and to account for the raw pressures coming in higher at Tahiti and lower @ Darwin.

Code: Select all
00z Euro Pressures June 20 2017

Hr: Darwin Tahiti
24 1015 1013
48 1015 1010
72 1014 1012
96 1013.50 1013
120 1013.50 1013.50
144 1014.50 1013
168 1015 1013
192 1015 1012.50
216 1016 1012
240 1016 1012.50"

These are essentially 0Z SLP's that you're listing such as the 1015's and 1016's at Darwin and the 1010 at Tahiti. There's no 6Z, 12Z, 18Z or any other hour mentioned here. Just 0Z. I was saying you shouldn't just use 0Z like you implied here. Using the average of 0Z, 6Z, 12Z, and 18Z would be way closer to matching the actual SOI.

[Kingarabian]

:uarrow: Let's get back to what you originally said that got this discussion going:

"Although I'm sure your analysis was based on the 12Z Euro, I see a considerably negative SOI compared to the numbers you posted. Here are the pressures I derived from the 00z Euro:
Note: I rounded the Darwin pressure down and the Tahiti pressures up to remove any bias, and to account for the raw pressures coming in higher at Tahiti and lower @ Darwin.

Code: Select all
00z Euro Pressures June 20 2017

Hr: Darwin Tahiti
24 1015 1013
48 1015 1010
72 1014 1012
96 1013.50 1013
120 1013.50 1013.50
144 1014.50 1013
168 1015 1013
192 1015 1012.50
216 1016 1012
240 1016 1012.50"

These are essentially 0Z SLP's that you're listing such as the 1015's and 1016's at Darwin and the 1010 at Tahiti. There's no 6Z, 12Z, 18Z or any other hour mentioned here. Just 0Z. I was saying you shouldn't just use 0Z like you implied here. Using the average of 0Z, 6Z, 12Z, and 18Z would be way closer to matching the actual SOI.

Oh I'm sorry If I wasn't clear. I understand completely that we can't use 24 hour points.

The following posts, I'm just trying to split the numbers up further to get a better estimate on the daily's so they would match up closer to yours.

[LarryWx]

OK, so we're on the same page then. Also, keep in mind that I used 0Z/6Z/12z/18Z of yesterday's 12Z GFS and 12Z Euro and pretty much averaged the two models out.
Yes, today came in -7 vs my estimate of -4. It is hard to get it just right. After the 10 days are up, I plan to quote my post and make an evaluation of how I did overall for the 10 days, whether good or bad.Let's see if the general idea ends up close.

[Kingarabian]

OK, so we're on the same page then. Also, keep in mind that I used 0Z/6Z/12z/18Z of yesterday's 12Z GFS and 12Z Euro and pretty much averaged the two models out.
Yes, today came in -7 vs my estimate of -4. It is hard to get it just right. After the 10 days are up, I plan to quote my post and make an evaluation of how I did overall for the 10 days, whether good or bad.Let's see if the general idea ends up close.

Sounds good. I wish there was a real time model to predict the SOI. It probably would make ENSO forecasts a lot easier.

[Kingarabian]

Looks like raw Tahiti's average today will end up @ 1010.86mb and Darwin will be @ 1012.86, so June 22 will be -20 to -25 ish, depending on if LongPaddock uses the same averages.

Per the 00z GFS Thursday analysis (w/o adjusting for bias),
June 23: the average pressures between the 00z, 06z, 12z, and 18z intervals for Tahiti will be around 1011.75 and Darwin will be around 1013.25. So the daily average will also come in around -20 to -25 if the pressures are to verify.
June 24: the average pressures between the 00z, 06z, 12z, and 18z intervals for Tahiti will be around 1012.40 and Darwin will be around 1012.75. So the daily average will be lower and it would be around -9 to -11, if the pressures are to verify.

Increasingly likely that the SOI will be positive from June 25-28, followed by a negative drop thereafter.

[Digital-TC-Chaser]

El Niño WATCH cancelled; ENSO neutral likely for 2017
The Bureau's ENSO Outlook has been reset to INACTIVE after an easing of climate model outlooks, and a reversal of the early autumn warming in the eastern tropical Pacific Ocean.
In the atmosphere, the trade winds and Southern Oscillation Index are well within the neutral range. Equatorial sea surface temperatures in the central Pacific are slightly warmer than average. However, far eastern Pacific Ocean sea surface temperatures, which were several degrees above normal near the Peruvian coast during March and April, cooled during May and June. This warmth had the potential to spread and develop into an El Niño event with global effects, but eased as trade winds failed to reinforce the ocean warmth. Other ENSO indicators also remain neutral.
All eight international models surveyed by the Bureau of Meteorology now suggest tropical Pacific Ocean temperatures are likely to remain ENSO-neutral for the second half of 2017. This compares to seven of eight models that suggested a possible El Niño in April.
While models have steadily eased back the likelihood of El Niño, most still indicate an increased chance of warmer and drier than average conditions for Australia over winter.
The Indian Ocean Dipole (IOD) is neutral. Three out of six climate models suggest a positive IOD will develop by the end of winter, and three are neutral. A positive IOD is typically associated with a drier than average winter and spring for southern and central Australia.



Southern Oscillation Index
The 30-day Southern Oscillation Index (SOI) to 18 June was −4.5 (90-day value −3.6), within neutral territory.
Sustained positive values of the SOI above +7 typically indicate La Niña while sustained negative values below −7 typically indicate El Niño. Values between about +7 and −7 generally indicate neutral conditions.

[Kingarabian]

It would be extraordinary if we do end up getting an El Nino despite the Neutral forecasts.

[Ntxw]

Per the buoys warming continued this week. May see 0.6 or 0.7C this coming Monday. While shallow about 100m down, the area stretching from near the dateline to 120W below is at or greater than 1C warmth. The cold anomalies are at 150-200m and while persistent have remained there the past 3 months+

[Kingarabian]

Ntxw, we've seen a fair of trade bursts over the past 3 months, why haven't the cooler anomalies been able to make it past 150m?

[Hurricaneman]

The PDO wants an El Niño hence why the trades haven't done a lot to the surface and subsurface but the atmosphere doesn't want an El Niño and that's why El Niño hasn't materialized

[Ntxw]

Ntxw, we've seen a fair of trade bursts over the past 3 months, why haven't the cooler anomalies been able to make it past 150m?

While it's true the warmth (anomaly wise) is close to the surface, the pool of warmth overall is a lot more than last year. 28C line was only 50m or so but this year is 100m west of the dateline. There is more than usual volume of warmth out west and the trades while prevalent near the dateline hasn't been all that strong to the east where 3.4 sits. If you go back in the thread to what a Nina subsurface and trades couple do it is more evenly spread out unlike this year where it is pretty much centered too far west.

I'm not sure if this is correct, but from history past all that piled warmth near and west of the dateline is probably waiting for a surge of successive WWB's to relieve it. It may not happen this year (it still could) but at some point it has to.

[LarryWx]

Per the buoys warming continued this week. May see 0.6 or 0.7C this coming Monday. While shallow about 100m down, the area stretching from near the dateline to 120W below is at or greater than 1C warmth. The cold anomalies are at 150-200m and while persistent have remained there the past 3 months+

It looks to me like some of the models that earlier this month significantly cooled fall/winter to neutral from Niño were largely basing that on June cooling off to near +0.2. We'll, that isn't happening at all. Expect to see model consensus warmer in 3.4 in their July progs.
Also, note the recent warming being 2-3 weeks after the re establishment in early June of a consistent solid -SOI. With the SOI at least temporarily going more neutral in late June, don't be surprised if 3.4's warming were to temporarily cease in early to mid July before any possible resumption of warming later in July.

[Kingarabian]

PDO will likely come in warmer for June.

[Kingarabian]

Per the buoys warming continued this week. May see 0.6 or 0.7C this coming Monday. While shallow about 100m down, the area stretching from near the dateline to 120W below is at or greater than 1C warmth. The cold anomalies are at 150-200m and while persistent have remained there the past 3 months+

It looks to me like some of the models that earlier this month significantly cooled fall/winter to neutral from Niño were largely basing that on June cooling off to near +0.2. We'll, that isn't happening at all. Expect to see model consensus warmer in 3.4 in their July progs.
Also, note the recent warming being 2-3 weeks after the re establishment in early June of a consistent solid -SOI. With the SOI at least temporarily going more neutral in late June, don't be surprised if 3.4's warming were to temporarily cease in early to mid July before any possible resumption of warming later in July.

Agreed. Also check the latest CFS runs. Pointing towards an El Nino for the heart of the season again though I'm sure it'll show a moderate La Niña in a couple of days considering its track record.

[cycloneye]

@MJVentrice
You would think atmospheric coupling associated with La Nina would have ended when La Nina was declared over. Nope. Standing wave alive.





 https://twitter.com/MJVentrice/status/878596558815715328

[LarryWx]

There have been no daily SOI updates since 6/21 for whatever reason.

[cycloneye]

@Met_mdclark
Interesting to note the continued warning of the Equatorial pacific water temps. #Nino SST but #Nina background. Fun stuff. #AGwx

 https://twitter.com/Met_mdclark/status/878970092863852544