2016 Global Temperatures

[chubbs]

I was suggesting if you go back a few years, the current bounce is more pronounced. The conditions in early 1997 weren't comparable to early 2015.

Yes there was warming in 2013+2014 before this super nino began due to recovery from the prolonged nina conditions in 2011/12. That is when the delta vs 18 years ago increased to 0.4C. 2013 was roughly 0.2C warmer than 1995.

[skierinvermont]

Yes there was warming in 2013+2014 before this super nino began due to recovery from the prolonged nina conditions in 2011/12. That is when the delta vs 18 years ago increased to 0.4C. 2013 was roughly 0.2C warmer than 1995.

 

Yeah so what I'm really getting at is that ENSO-correct temperatures have only been running ~.3C warmer vs 18 years ago until the current Nino. One possible explanation (unproven yet) is similar to what Isotherm was suggesting which is that this ENSO is stronger than perhaps indicated by the ONI.

[chubbs]

Yeah so what I'm really getting at is that ENSO-correct temperatures have only been running ~.3C warmer vs 18 years ago until the current Nino. One possible explanation (unproven yet) is similar to what Isotherm was suggesting which is that this ENSO is stronger than perhaps indicated by the ONI.

My pet theory is that the slow progression of this ENSO cycle since the nina ended in 2012 allowed more warming than the shorter cycle in 97/98. Some of the warming is merely a recovery from the cooler hiatus years. In any case will have to wait until the nino is over to evaluate the role of ENSO vs other factors in the warming surge of the past few years.

[nflwxman]

My pet theory is that the slow progression of this ENSO cycle since the nina ended in 2012 allowed more warming than the shorter cycle in 97/98. Some of the warming is merely a recovery from the cooler hiatus years. In any case will have to wait until the nino is over to evaluate the role of ENSO vs other factors in the warming surge of the past few years.

I agree. The heat does not just appear and disappear in a matter of months. I think after nearly a decade of mostly La Niña conditions, we are just playing "catchup" with the current radiative imbalance.

These temperatures are even blowing the door off my expectations with Feburary likely to come in absurdly high on all datasets.

[bluewave]

I agree. The heat does not just appear and disappear in a matter of months. I think after nearly a decade of mostly La Niña conditions, we are just playing "catchup" with the current radiative imbalance.

These temperatures are even blowing the door off my expectations with Feburary likely to come in absurdly high on all datasets.

 

Yeah, the full basin nature of this event across the whole Tropical Pacific has no parallel.

 

[skierinvermont]

Yeah, the full basin nature of this event across the whole Tropical Pacific has no parallel.

 

SST.png

 

That's kind of a different point than nfl or chubbs is making though and more in line with what I'm thinking. He's suggesting the bigger spike for this Nino is due to AGW catchup that was just hidden by Nina conditions. I'm suggesting that the spike is bigger because this Nino is off the charts and even stronger than 1998. 

[chubbs]

That's kind of a different point than nfl or chubbs is making though and more in line with what I'm thinking. He's suggesting the bigger spike for this Nino is due to AGW catchup that was just hidden by Nina conditions. I'm suggesting that the spike is bigger because this Nino is off the charts and even stronger than 1998.

Here are SST for Dec to Feb 2015/16 minus 1997/98. Overall the oceans are warmer in 2015/16 except for cooling in the eastern tropical Pacific and near major ice sheets. The warmth in the western tropical Pacific is in-line with warmth elsewhere and if anything argues for a weaker nino vs. 98.

[Isotherm]

 

 

Are these raw or trendline removed anomalies? Because if they are raw, the .3C warmer reading in 2015 is explained by a baseline increase in global temperatures and ocean temperatures. Thus the two Ninos would be exactly comparable in terms of magnitude. 

 

Moreover, a .3C difference is splitting hairs.

 

Plus that is one month. What was the peak trimonthly? And what was the peak across the whole ENSO region not just 3.4?

 

 

EDIT: Looks like Mallow already hit on almost all these points.

 

 

I'm not ruling out the possibility this Nino cause a greater spike in temperature due to the nature of the Nino. But I don't think the evidence you've offered thus far comes nearly close to supporting that idea. It seems more like wish-casting based on one metric rather than a comprehensive analysis with respect to time, space and dynamics.

 

If I had to guess, I'd actually guess you are right simply because the temperature spike IS bigger than 1998 (I think - somebody correct me if I'm wrong). Whether this is due to the magnitude/nature of the Nino is the question. My guess is you are probably right. But I'm not willing to just assume that.

 

 

 

 

The most significant difference occurred in region 4. The trimonthly peak for absolute temperatures in region 4 was 0.69C warmer in 2015-16 versus 1997-98.

 

2015-16 trimonthly peak in region 3.4:  +2.79C; 29.39C

1997-98 trimonthly peak in region 3.4:  +2.63C; 29.23C

 

2015-16 trimonthly peak in region 4:   +0.78c; 30.07C

1997-98 trimonthly peak in region 4:  +1.47c; 29.38C

 

More importantly though, if we examine absolute SST’s across the entire tropical Pacific basin for both events, it seems that the area of 29-30C SST’s was much more expansive in the western and central tropical Pacific. Additionally, there was a significant zone of 30C+ SST’s in 2015-16 which was not present in 1997-98. Further east, although region 1+2 temperatures were much warmer in 1997-98, note the area immediately to the south of Mexico [equator – 15N]; that region was quite a bit warmer in 2015-16, which is why we saw quite a bit of convection occurring far to the northeast of the core upward motion cell over the past several months.

 

I think the warmer baseline SST’s probably play a role in this, but I tend to think it’s mostly the orientation / behavior of the Nino event which promoted much more widespread, basin-wide warmth compared to 1997-98. If we were to calculate the total area of > 29C SST’s in the tropical Pacific in 2015-16, I would bet that the result would be quite a bit higher than 1997-98. We don’t know for certain without doing the calculations, but the reanalysis maps suggest that the area of > 29C SST’s was larger. And thus, one could infer that the available potential energy for tropical convection was also greater – consequently, yielding an enhanced injection of latent heat into the lower troposphere. It would be an interesting study for sure.

 

We should probably wait until the Nino event is completely over. Maybe another way to address this would be to calculate the total outgoing longwave radiation values across the tropical Pacific for the entire period of the +ENSO event, as often times, OLR can be an invaluable proxy for deep tropical convection, and thus give us a sense of the amount of latent heat release occurring. Edit: A brief check/comparison of OLR for the period June-Feb of both years reveals a more widespread/elongated zone of strong forcing in 2015 versus 1998 (strong forcing this year extended from approximately 160E all the way to near South America; 1997-98 was much more focused in the 160W-120W region).

 

However, going by surface SST’s alone, it seems the amount of energy available for tropical convection was quite a bit greater in the 2015-16 El Nino event. In terms of ascertaining the relative influence of the Nino itself vs. background AGW forcing on those warmer SST’s, that is a topic for more study.

 

 

Oct-Feb SST's 1998:

 

 

Oct-Feb SST's 2016:

 

[The_Global_Warmer]

Looking at the graphics above.

2015-16 is warmer all over except the far Epac.

[Mallow]

Why are you focusing on 29°C when the convective threshold is closer to 27°C? I agree that the area of 29°C is greater in the 2015-2016 El Niño, but I would estimate that the area of 27°C is about equal between the two of them. If anything, because of the addition of the large convective region in the East Pacific in 1997-1998 that isn't present in 2016, it may be that 1997-1998 ekes out the area battle for that particular temperature threshold.

[bluewave]

That's kind of a different point than nfl or chubbs is making though and more in line with what I'm thinking. He's suggesting the bigger spike for this Nino is due to AGW catchup that was just hidden by Nina conditions. I'm suggesting that the spike is bigger because this Nino is off the charts and even stronger than 1998. 

 

The warmer overall Tropical Pacific SST's than 97-98 also resulted in a increase and expansion westward of the convection. Increased tropical rainfall over 97-98. But I am not sure how good the data before 1970 is.

 

Strongest convection was shifted poleward north of the Equator from R1+2 west through 3.4. R4 west to 100E saw a 

large increase in convection over 97-98. 

 

Another interesting feature of this event was the stronger MJO activity which may be the result of the record SST's

from R4 across to the IO.

 

 

 

 

 

 

[Msalgado]

That's kind of a different point than nfl or chubbs is making though and more in line with what I'm thinking. He's suggesting the bigger spike for this Nino is due to AGW catchup that was just hidden by Nina conditions. I'm suggesting that the spike is bigger because this Nino is off the charts and even stronger than 1998. 

 

There's reason to believe that these two lines of thinking aren't mutually exclusive.  A decade long of more of the heat going into the ocean could have provided the backdrop for this type of basin wide SST pattern to manifest.  

[Mallow]

Why are you focusing on 29°C when the convective threshold is closer to 27°C? I agree that the area of 29°C is greater in the 2015-2016 El Niño, but I would estimate that the area of 27°C is about equal between the two of them. If anything, because of the addition of the large convective region in the East Pacific in 1997-1998 that isn't present in 2016, it may be that 1997-1998 ekes out the area battle for that particular temperature threshold.

 

It's more complicated than this, of course. Recent research suggests that the tropical troposphere has warmed pretty evenly at all levels (consistent with moist-adiabatic adjustment, and therefore consistent with climate models). This means that the convective threshold is also warming (to get the same instability, you need a warmer surface if you have a warmer upper troposphere). A true analysis of the 1997-1998 El Niño vs. the 2015-2016 El Niño would take this into account. Let's do a hypothetical scenario to illustrate...

 

If the convective threshold in 1997-1998 was 27.0°C, but the middle/upper tropical troposphere has warmed approximately 0.5°C since then, then the new convective threshold would be approximately 27.5°C (not exactly, since convective instability is not linear with the upper-lower temperature differential, but for small changes, it is a reasonable approximation). So, if we are attempting to address the question of which year's ENSO event had a larger region of convective precipitation, what we really should be comparing is the area of SSTs exceeding 27.0°C in 1997-1998 to the area of SSTs exceeding 27.5°C in 2015-2016. I won't pretend to be able to eyeball that.

[skierinvermont]

OK so I finally got around to creating an ENSO and TSI corrected chart. I made some assumptions about 2016. I assumed it would finish at 1.05 on GISS. And I assumed that the 3 month lagged ONI (October 2015 to September 2016) would be .16C higher (1.0 instead of .84).

 

You can see that, even assuming a slightly stronger ONI, the spike is quite unusual in 2016 if my assumptions are correct. 

 

I think this suggests that either the Nino is stronger than the region 3.4 ONI is indicating or there is some other 3rd variable. I would guess the former.

[Mallow]

The warmer overall Tropical Pacific SST's than 97-98 also resulted in a increase and expansion westward of the convection. Increased tropical rainfall over 97-98. But I am not sure how good the data before 1970 is.

 

Strongest convection was shifted poleward north of the Equator from R1+2 west through 3.4. R4 west to 100E saw a 

large increase in convection over 97-98. 

 

Another interesting feature of this event was the stronger MJO activity which may be the result of the record SST's

from R4 across to the IO.

 

RN.png

 

TRP.png

 

Cool plots, especially the second one! Do you have a link to the numerical values for that chart?

[bluewave]

There's reason to believe that these two lines of thinking aren't mutually exclusive.  A decade long of more of the heat going into the ocean could have provided the backdrop for this type of basin wide SST pattern to manifest.  

 

This is going to provide a ton of research material. The other interesting feature was the exceptional warm blob over the

NE PAC which emerged during the summer of 2013 followed by the strong +PDO shift shortly after. It's very impressive

to see the SST departures actually increase at depth for this feature over the winter.

 

https://alaskapacificblob.wordpress.com/2016/02/09/subsurface-warmth-persists/

 

Looking at the 100-200m layer, it’s remarkable to see that the region of warm water in the northeast Pacific actually increasedslightly in size and intensity from October to January.  The anomaly is slightly east of the NPM center-of-action of 45°N/150°W, but nevertheless it’s clear that “the Blob” lives on under the surface.

[bluewave]

Cool plots, especially the second one! Do you have a link to the numerical values for that chart?

 

This is the ESRL link.

 

http://www.esrl.noaa.gov/psd/data/timeseries/

 

Just plug in the coordinates and choose the value on the menu that you want to plot.

[skierinvermont]

There's reason to believe that these two lines of thinking aren't mutually exclusive.  A decade long of more of the heat going into the ocean could have provided the backdrop for this type of basin wide SST pattern to manifest.  

 

Maybe a slight contribution, but the upper ocean has only warmed like what .1C? Maybe less?

 

I think it's mostly just that this Nino is a different beast, and I think we're starting to get into some of those reasons above.

 

Could also be some significant contribution from some variable besides ENSO or AGW.

[skierinvermont]

Here's a 2nd chart using a 4 month lag. You can see it smooths 1998 out even more, but 2016 would remain a big jump if this year finishes at 1.05. For this chart I assumed than the ONI 4 month lag for 2016 would only be .1C warmer than 1998 (1.25 instead of 1.15). That's a pretty favorable assumption given it is currently average a hair lower than 1998. It basically assumes that it dies much slower from now until August (the last month that counts) .

[chubbs]

Here's a 2nd chart using a 4 month lag. You can see it smooths 1998 out even more, but 2016 would remain a big jump if this year finishes at 1.05.

Could you put a regression line through the points? Part of this year's jump looks like a recovery from a period of slow increase, similar to 2005.

[skierinvermont]

You can also see from my chart that ENSO corrected temps were only running about .2-.25C warmer than 18 years ago until 2016 but are now running nearly .4C warmer.

[skierinvermont]

Could you put a regression line through the points? Part of this year's jump looks like a recovery from a period of slow increase, similar to 2005.

Or the underlying trend has slowed for some other reason and it just a bounce above the trend like 1988 was, due to some abnormality of this El Nino or temporary 3rd variable. If that's the case, temperatures will return to below trend within a couple years.

 

Anyways here it is with trendline. This is with the 4 month lagged ONI, assuming 1.05C on GISS and 1.25C 4 month lagged ONI.

[Mallow]

This is the ESRL link.

 

http://www.esrl.noaa.gov/psd/data/timeseries/

 

Just plug in the coordinates and choose the value on the menu that you want to plot.

 

Thanks!

[nflwxman]

Or the underlying trend has slowed for some other reason and it just a bounce above the trend like 1988 was, due to some abnormality of this El Nino or temporary 3rd variable. If that's the case, temperatures will return to below trend within a couple years.

 

Anyways here it is with trendline. This is with the 4 month lagged ONI, assuming 1.05C on GISS and 1.25C 4 month lagged ONI.

 

 

Nice chart.  I doubt temperatures will revert much below the trendline as long as a) we stay in a +PDO regime and there is no major volcanic eruption.  As the OHC data has shown, global warming has accelerated and thus we should expect temperatures to respond provided the ocean is not in a deep uptake mode (La Nina).  Every ENSO event is not in isolation and can feed or delete off a previous ENSO event (i.e the double dip Nina of 2011-2013).

 

I guess the CIMP5 decadal models are not looking as bad these days, huh?

[Mallow]

Or the underlying trend has slowed for some other reason and it just a bounce above the trend like 1988 was, due to some abnormality of this El Nino or temporary 3rd variable. If that's the case, temperatures will return to below trend within a couple years.

 

Anyways here it is with trendline. This is with the 4 month lagged ONI, assuming 1.05C on GISS and 1.25C 4 month lagged ONI.

 

Looks a lot like standard variations around a steady trend to me.

[chubbs]

Or the underlying trend has slowed for some other reason and it just a bounce above the trend like 1988 was, due to some abnormality of this El Nino or temporary 3rd variable. If that's the case, temperatures will return to below trend within a couple years.

 

Anyways here it is with trendline. This is with the 4 month lagged ONI, assuming 1.05C on GISS and 1.25C 4 month lagged ONI.

Thanks. Perhaps this nino will be unusually effective at raising 2016 global temperatures to a small extent (assuming 2016 end up as plotted). The big picture though is that global temperatures are close to the long-term trend line after adjusting for ONI.

[chubbs]

Nice chart.  I doubt temperatures will revert much below the trendline as long as a) we stay in a +PDO regime and there is no major volcanic eruption.  As the OHC data has shown, global warming has accelerated and thus we should expect temperatures to respond provided the ocean is not in a deep uptake mode (La Nina).  Every ENSO event is not in isolation and can feed or delete off a previous ENSO event (i.e the double dip Nina of 2011-2013).

 

I guess the CIMP5 decadal models are not looking as bad these days, huh?

From Gareth Jones' UK Met Office twitter. Looks like obs are through 2015.

[skierinvermont]

Looks a lot like standard variations around a steady trend to me.

 

 

Thanks. Perhaps this nino will be unusually effective at raising 2016 global temperatures to a small extent (assuming 2016 end up as plotted). The big picture though is that global temperatures are close to the long-term trend line after adjusting for ONI.

 

I agree with the statement about the big picture. Any time we talk about annual or even a decade of temperature we are really micro-analyzing because we don't have a good understanding of what affects temperature on an annual or even decadal scale. However, by micro-analyzing it like this perhaps we can learn something. We can't simply dismiss variation from the trend as meaningless. Either there is a physical cause, or the data is wrong.

 

If you exclude 2016, the 10 year trend is much lower than it was before. Either the data is wrong, or there is a 3rd variable other than AGW and my simplistic ENSO-correction. Or more likely, many 3rd variables. 

 

So the question is, is the spike in my chart in 2016 temporary or permanent? My guess is that it is somewhat temporary and is due to this ENSO being stronger than the ONI would indicate. That's an effect that would disappear when we revert  to Nina conditions.

[Isotherm]

The ONI is based upon the ERSST dataset, which was quite a bit different/cooler than the OISST dataset for this El Nino event. I think the latter is more representative of the peak magnitude (as explained with the numbers on the previous page; this Nino peaked at nearly +3 in region 3.4). The 29C was utilized as a benchmark simply to make it easier to eyeball the difference. However, I agree with the likely upward trend in convective threshold and have seen some of those studies as well. I'm not sure that we can effectively quantify the total latent release with each el nino, though I think a less indirect way of quantifying it would be total OLR (as opposed to SST convective threshold analysis).

[Mallow]

The ONI is based upon the ERSST dataset, which was quite a bit different/cooler than the OISST dataset for this El Nino event. I think the latter is more representative of the peak magnitude (as explained with the numbers on the previous page; this Nino peaked at nearly +3 in region 3.4). The 29C was utilized as a benchmark simply to make it easier to eyeball the difference. However, I agree with the likely upward trend in convective threshold and have seen some of those studies as well. I'm not sure that we can effectively quantify the total latent release with each el nino, though I think a less indirect way of quantifying it would be total OLR (as opposed to SST convective threshold analysis).

 

Agreed with your last sentence, OLR over the equatorial regions should vary with convection pretty regularly.