2013 Global Temperatures

[ORH_wxman]

Will's comparison is valid, but it should be noted that the CFS has a different baseline temperature than GISS, thus with the baseline the CFS uses, the anomalies will naturally be cooler than GISS. The CFS uses a 1981-2010 baseline, whereas GISS generally uses a 1951-1980 baseline. Since the late-20th Century was generally warmer than the mid-20th Century, this would skew the anomalies.

 

For example, the GISS anomaly for March 2010 with a 1981-2010 baseline is 0.41 Degrees C. Still warmer than the CFS' anomaly with a small disagreement, but the CFS actually is not too dissimilar to the Global Temperature Anomaly from GISS when using the same baseline.

 

 

 

 

I don't think anyone argued that the weatherbell maps and GISS had the same baseline.

 

We were trying to figure out if there was a robust correlation with the weatherbell anomalies and GISS anomalies. We hadn't bothered to do it. But it seems (at least based on the last 3+ years of data) that there is quite a strong correlation between the two. We just have to add about 0.54 or so to the weatherbell maps and we'll almost always be within 5 hundreths of the GISS anomaly.

 

There's obviously a few outliers, but I think that happens even amongst the regular datasets like Hadcrut/NCDC as well. Hopefully the weatherbell maps continue to have a strong correlation with the others as it will allow us to track monthly anomalies in progress fairly accurately. Previously the only real-time data we've had was AMSU channel 5 on the satellites but that has failed...and it also told us nothing about the surface datasets.

[Snow_Miser]

I don't think anyone argued that the weatherbell maps and GISS had the same baseline.

 

We were trying to figure out if there was a robust correlation with the weatherbell anomalies and GISS anomalies. We hadn't bothered to do it. But it seems (at least based on the last 3+ years of data) that there is quite a strong correlation between the two. We just have to add about 0.54 or so to the weatherbell maps and we'll almost always be within 5 hundreths of the GISS anomaly.

 

There's obviously a few outliers, but I think that happens even amongst the regular datasets like Hadcrut/NCDC as well. Hopefully the weatherbell maps continue to have a strong correlation with the others as it will allow us to track monthly anomalies in progress fairly accurately. Previously the only real-time data we've had was AMSU channel 5 on the satellites but that has failed...and it also told us nothing about the surface datasets.

 

Yeah, it definitely looks like adding 0.5-0.55 Degrees C to the CFS anomalies will roughly be in the ballpark of the GISS monthly anomalies based off of your analysis.

[skierinvermont]

Will's comparison is valid, but it should be noted that the CFS has a different baseline temperature than GISS, thus with the baseline the CFS uses, the anomalies will naturally be cooler than GISS. The CFS uses a 1981-2010 baseline, whereas GISS generally uses a 1951-1980 baseline. Since the late-20th Century was generally warmer than the mid-20th Century, this would skew the anomalies.

 

For example, the GISS anomaly for March 2010 with a 1981-2010 baseline is 0.41 Degrees C. Still warmer than the CFS' anomaly with a small disagreement, but the CFS actually is not too dissimilar to the Global Temperature Anomaly from GISS when using the same baseline.

 

 

 

 

If the only difference was the use of different baselines, then yeah we could just switch baselines. But, if I understand correctly, the CFS is modelling SAT not actually measuring SAT, unlike GISS which measures it. Thus there was some doubt as to whether the CFS was even correlated to GISS, even after putting on the same base period. It turns out there is a pretty good correlation, although the correction that has to be applied is not just a correction for base period. 

 

If the correction was only for base period, the correction would only be .40C (the 1981-2010 base is .40C warmer than the 1951-1980 base on GISS). However, the correction is a full .54C. I think this is because the CFS doesn't show the correct amount of, if any, long-term warming probably related to the fact that it's a model not a measurement. It's still showing temperatures similar to 1980, when in reality the earth has warmed a lot since then. 

 

Thus, over time (5 or 10 years), the .54C correction would have to be gradually enlarged. 

[The_Global_Warmer]

SST's have gone back up last week. 

 

[The_Global_Warmer]

I wouldn't say set in stone either. More like 70% odds the next 6 months are warmer than the last 3, roughly similar to the 80% odds I would give of 2013 being warmer than 2012. 

 

 

I was thinking that the summer and fall have the biggest anomalies based on the fact that the NH has been seeings its largest SST anomalies in summer and fall, probably related to the loss of ice. However, I just went and checked global LOTI GISS and over the last 13 years (2000-2012) the months with the biggest +anomalies are actually March, April, September, October, and November. The summer months of JJA actually have the lowest +anomalies. 

 

So for whatever reason, the bigger SSTAs in NH summer/fall are not translating into bigger global GISS anomalies and Friv's argument the past few months is not correct. I can only think of a few reasons this would be the case:

 

1) The anomalies Friv is posting are somehow incorrect or some alteration is made before it is "final" and incorporated into GISS

2) The bigger summer/fall +SSTAs in the NH are being counterbalanced by the SH

3) The bigger summer/fall +SSTAs in the NH are being counterbalanced by land temps

 

I'm guessing some combo of #2 and 3

 

 

You are right about that.

 

Thank You for doing that research and finding that out. 

 

Obviously it's not a full proof guarantee that it will be warmer.  I can readjust for that and I still see a 90% or better shot it will be warmer.

 

1.  SST's the past 5 weeks have averaged slightly above .25+.

 

Which is a 5 week average barely under any 5 week average on record. 

 

And it is likely the record high 5 week average when the MEI was Negative the prior month.

 

The MEI the last 6 months was barely above 0.00.  Yet we have been very warm for that.

 

Methane levels are at record high's concentrated over the Nothern Hemispheres Northern 1/3rd.

 

C02 levels are at global highs.  Averaging almost 400PPM globally.  Solar influence has lessened it's cool grip. 

 

the PDO has been a powerful cooling mechanism.

 

But the PDO has weakened a bit from the big recent low's.  This argues for a possible NINO inside of a -PDO.

 

The AMO is running hot and as you can see April's AMO is going to be robust.  Likely .300 or higher.

 

 

 

 

ENSO has cooled again.  We can see Nino 1-2 just dropped a degree.  NINO 3 cooled.  NINO 3-4, 4 is steady in that period.  Overall it's cool.  But other parts of the Ocean are torching.

 

 

 

 

 

Without a NINA I can't see any month the next 6 being under a .60 monthly anomaly on GISS/NCDC. 

 

We will see.

[ORH_wxman]

 

 

 

 

 

 

 

 

Without a NINA I can't see any month the next 6 being under a .60 monthly anomaly on GISS/NCDC. 

 

We will see.

 

 

 

Its looking an awful lot like GISS will be below +0.60C for April if our weatherbell calculations are correct. We'll need to see a big rise in the final week of the month

[skierinvermont]

Yeah like we said earlier it is looking like .47 for April. There's a lot more variability in short-term temperatures than you're making there out to be. 

[The_Global_Warmer]

I highly doubt it will be that low. 

 

They essentially use the Reynold's SST's data set.  I am not sure how much your calculations adjust for that but it's running near .25 this month on the 1971-2000 anomolies.  They obviously use the other CLIMO set but it has to be even larger on that or close to the same.

 

If the land is that cold then there will have to be regions pretty cold over a large area to off set the warmer ocean.

 

2006 and 2008 were during big negative ENSO events.

 

Comparing there sst's and looking at the charts like the one above.  I would think land would be warm enough even with cold in North America and a small part of Eastern Asia.  The arctic like most months is torching.  Thanks end of the long negative AO.

 

I am going for .57-.58 on GISS and .61 on NCDC for April as it stands today.

 

Models show the Deep Freeze ending over North America but slowly.

 

Warmth coming to Eastern Asia and SST's .25C above normal middle of April.

 

If it comes out to be .47 without much cooling or warming the last half of the month I will concede how I see it as of now.

[ORH_wxman]

I highly doubt it will be that low. 

 

They essentially use the Reynold's SST's data set.  I am not sure how much your calculations adjust for that but it's running near .25 this month on the 1971-2000 anomolies.  They obviously use the other CLIMO set but it has to be even larger on that or close to the same.

 

If the land is that cold then there will have to be regions pretty cold over a large area to off set the warmer ocean.

 

 

2006 and 2008 were during big negative ENSO events.

 

Comparing there sst's and looking at the charts like the one above.  I would think land would be warm enough even with cold in North America and a small part of Eastern Asia.  The arctic like most months is torching.  Thanks end of the long negative AO.

 

I am going for .57-.58 on GISS and .61 on NCDC for April as it stands today.

 

Models show the Deep Freeze ending over North America but slowly.

 

Warmth coming to Eastern Asia and SST's .25C above normal middle of April.

 

If it comes out to be .47 without much cooling or warming the last half of the month I will concede how I see it as of now.

 

 

 

GISS doesn't use Reynolds SST data anymore. They switched over to ERSST.v3b recently.

 

 

Regardless of what data they use, a fairly strong correlation between the weatherbell maps and the GISS anomalies exists. So if the weatherbell maps don't warm at least 0.05C for the whole month between now and a week from now, then a GISS value of +0.60C would be extremely unlikely. Even a value near your revised 0.57C figure would be unlikely.

 

But we'll see what the map says in a week.

[skierinvermont]

I'll go with .52 for GISS (.05 higher than weatherbell would suggest) on the assumptions that 1) the use of SST data which is running warm leads to a higher anomaly than weatherbell maps would suggest and/or  2) weatherbell maps may warm a bit before the end of the month

 

I think .57 is pretty unlikely given the correlation between the two sources, but can't rule it out as a possibility especially if April finishes warm

 

One also has to remember that a .25C SST anomaly on a 1971-2000 baseline is probably only a .40-.45C anomaly on a 1951-1980 baseline. The land has to be warm (.7C+) to achieve big anomalies of .5C+ globally. Most of the variance in global temperatures is driven by the land anomalies, not ocean anomalies. 

[PottercountyWXobserver]

I'm thinking it comes in near .5

[Snow_Miser]

Is there a link to the CFS global temperatures?

[tacoman25]

I'll go with .52 for GISS (.05 higher than weatherbell would suggest) on the assumptions that 1) the use of SST data which is running warm leads to a higher anomaly than weatherbell maps would suggest and/or  2) weatherbell maps may warm a bit before the end of the month

 

I think .57 is pretty unlikely given the correlation between the two sources, but can't rule it out as a possibility especially if April finishes warm

 

One also has to remember that a .25C SST anomaly on a 1971-2000 baseline is probably only a .40-.45C anomaly on a 1951-1980 baseline. The land has to be warm (.7C+) to achieve big anomalies of .5C+ globally. Most of the variance in global temperatures is driven by the land anomalies, not ocean anomalies. 

 

We also have to remember that the GISS/Weatherbell comparisons are based off temps that have been adjusted after they initially came out. So the initial number we see for April may change later in the year to be closer to Weatherbell.

[ben4vols]

Is there any way to see the GISS anomaly for the area of 70N to 70S?

[The_Global_Warmer]

Global sst's come in again around .25C for last week.  That is Thursday to the most recent Wednesday.

 

This is a break down of them in 3 regions.  30N-90N, -30-30N, -90-(-30N).

 

 

 

30-90N:

 

We can see the large upswing is about to take place.

 

 

 

-30 to 30N:  Despite no nino the tropics are running warm.

 

-90 to -30N: Running well above the most recent year.

 

 

The NH warming is guaranteed.  Something must give or global ssts will warm into the .3 to .4 range. The range of the two warmest years on record.  2010 and1998 both caused by Strong to Super Nino's.  Another similar period was always caused by persistent Nino's in the early to mid 2000s.

 

 

 

During such a strong negative PDO/negative enso periods/strongest solar min/max in a century things wouldn't be running so warm.  But they are.

 

I believe we are at the end of one of the short lack of warming periods during a century of warming.

[ORH_wxman]

Weatherbell maps have warmed slightly to about -0.05 since last week, so it might warm another hundreth or so with 2 days to go. If the past 3+ years holds true, we are very likely to see GISS come in between +0.46-+0.56 this month.

[skierinvermont]

We also have to remember that the GISS/Weatherbell comparisons are based off temps that have been adjusted after they initially came out. So the initial number we see for April may change later in the year to be closer to Weatherbell.

 

Aren't any adjustments always tiny (if there is one)?

[LakeEffectKing]

HadCrut 4 hasn't updated for March, and I believe this will be the first time that the figure didn't come out within the following month.

[ORH_wxman]

Weatherbell maps have warmed slightly to about -0.05 since last week, so it might warm another hundreth or so with 2 days to go. If the past 3+ years holds true, we are very likely to see GISS come in between +0.46-+0.56 this month.

 

 

The final weatherbell April anomaly was -0.04 which would place the best guess around 0.50 for GISS...but with a likely range of 0.46-0.56.

[The_Global_Warmer]

April broken into 3 periods:

 

There has definitely been some very consistent deeper cold over the NH lands.  Obviously North America has been well into the deep freezer for most of April over large portions of it.

 

 

 

 

It appears I may have under estimated the large vast region of deeper cold over North America.

 

[SVT450R]

 

[ORH_wxman]

That seems to line up with what we are expecting from GISS based on the weatherbell maps...April coming in a bit lower than March. Though satellites can often be off from the surface on a month to month basis.

[tacoman25]

UAH_LT_1979_thru_Apr_2013_v5.5.png

 

03503-03b9186e-435b-4161-8634-02c45318374d.png

 

.4C drop from January, that's pretty huge. Coldest month for the NH since March 2012 as well.

[The_Global_Warmer]

That seems to line up with what we are expecting from GISS based on the weatherbell maps...April coming in a bit lower than March. Though satellites can often be off from the surface on a month to month basis.

 

 

Yeah.

 

It looks like I am probably to warm with my prediction for April. 

 

It's clear the blocking all along the Northern Hemisphere really was the difference maker.  I will try to incorporate land variability a bit more. 

 

Last year April: G+0.232 NH+0.351 SH+0.114 T-0.242

 

The big kicker last year was the .65 NH land anomaly for April. 

 

Giss had a .59 last April and NCDC had a .68.

[LakeEffectKing]

And still no MARCH #'s for HadCrut 4.....broken??

[LakeEffectKing]

Now May 6th, No March #'s from HadCrut.....anyone have knowledge for the delay??

 

BTW. RSS is up a hundredth from last month to .20

[ORH_wxman]

Now May 6th, No March #'s from HadCrut.....anyone have knowledge for the delay??

 

BTW. RSS is up a hundredth from last month to .20

 

 

I haven't found anything. HadSST2 isn't updated past February either.

[The_Global_Warmer]

Now May 6th, No March #'s from HadCrut.....anyone have knowledge for the delay??

 

BTW. RSS is up a hundredth from last month to .20

 

http://www.ssmi.com/msu/msu_time_series.html

 

From March to April or Feb to March?

 

I figured it would have dropped a bit from the previous month.

[ORH_wxman]

http://www.ssmi.com/msu/msu_time_series.html

 

From March to April or Feb to March?

 

I figured it would have dropped a bit from the previous month.

 

 

It prob didnt drop because Anatarctica was cold...and RSS has a blackout down south of 70S versus UAH....usually when RSS is warmer than UAH, its because of the Antarctic temps.

[LakeEffectKing]

So, the HadCrut4 link for the data (at least the one I was using) went from not updating, to now a link that doesn't work.  But the old HadCrut3 (for March) finally updated:

 

2012/03  0.291  0.311  0.271  0.438  0.143  0.291  0.284  0.440  0.142  0.440  0.141
2012/04  0.499  0.517  0.481  0.643  0.355  0.499  0.492  0.645  0.354  0.645  0.354
2012/05  0.484  0.503  0.465  0.649  0.319  0.484  0.476  0.650  0.318  0.650  0.317
2012/06  0.483  0.500  0.465  0.626  0.340  0.483  0.475  0.627  0.339  0.627  0.339
2012/07  0.447  0.465  0.430  0.632  0.263  0.447  0.440  0.633  0.262  0.633  0.262
2012/08  0.512  0.529  0.495  0.737  0.288  0.512  0.505  0.737  0.287  0.737  0.287
2012/09  0.516  0.533  0.499  0.703  0.329  0.516  0.509  0.704  0.328  0.704  0.328
2012/10  0.491  0.509  0.474  0.646  0.337  0.491  0.485  0.647  0.336  0.647  0.336
2012/11  0.486  0.506  0.466  0.609  0.364  0.486  0.479  0.610  0.362  0.610  0.362
2012/12  0.257  0.279  0.236  0.393  0.121  0.257  0.250  0.395  0.120  0.395  0.119
2013/01  0.390  0.412  0.369  0.542  0.238  0.390  0.383  0.544  0.236  0.544  0.236
2013/02  0.424  0.445  0.403  0.570  0.278  0.424  0.417  0.572  0.276  0.572  0.276
2013/03  0.387  0.408  0.367  0.541  0.233  0.387  0.380  0.543  0.232  0.543  0.232