2015 Global Temperatures

[ORH_wxman]

The shape of the graph looks similar to how it always looked...it's just a bit less egregious than the 5.6 version. RSS still warms the most in the 1990s and then cools in the most recent years.

[StudentOfClimatology]

Disagree. There is clearly some sort of drift there throughout the dataset. Chubbs is aware of the baseline difference.

He obviously hasn't corrected for the baseline difference. The corrected data I have doesn't diverge anywhere near the extent that his comparison depicts.

[AvantHiatus]

I found evidence of a cold bias in the GFS long-range. Apologies for it being a shade unprofessional but watch the full temperature anomaly loop, you will see the el nino magically melt away post-180. This is why I am skeptical of strong PV consolidation over the Arctic in the day 10-15 range.

 

http://www.tropicaltidbits.com/analysis/models/?model=gfs&region=namer&pkg=T2ma&runtime=2015043012&fh=0&xpos=0&ypos=290.9090846038063

[chubbs]

He obviously hasn't corrected for the baseline difference. The corrected data I have doesn't diverge anywhere near the extent that his comparison depicts.

The baseline shift is just an offset and doesn't affect the up and down cycles of UAH vs RSS. There are still  systematic differences in the satellite analyses even though the overall 36-year trends are close..

[StudentOfClimatology]

The baseline shift is just an offset and doesn't affect the up and down cycles of UAH vs RSS. There are still systematic differences in the satellite analyses even though the overall 36-year trends are close..

Of course. I was just pointing out that you kind of presented the baseline offset as if it were an actual deviation between the datasets when it's not. Doing that essentially makes your graph pointless because your differential anomalies won't be real.

[The_Global_Warmer]

I found evidence of a cold bias in the GFS long-range. Apologies for it being a shade unprofessional but watch the full temperature anomaly loop, you will see the el nino magically melt away post-180. This is why I am skeptical of strong PV consolidation over the Arctic in the day 10-15 range.

http://www.tropicaltidbits.com/analysis/models/?model=gfs&region=namer&pkg=T2ma&runtime=2015043012&fh=0&xpos=0&ypos=290.9090846038063

Its a climo bias.

Its leans cold for obvious reasons.

The PV is fully established by day 7-8.

Global temps tho will be roasting.

[SVT450R]

For those interested here is the new UAH yearly ranking compared to the old version.

[The_Global_Warmer]

For those interested here is the new UAH yearly ranking compared to the old version.

UAH Temp ranking old vs new.png

That is comical.

Wouldn't be so questionable if every non temperature piece of evidence didn't say this is wrong.

Even more strange is the raw amsu channel 6 temps have run quite warm. So the discrepancy is somewhere closer to the surface.

It took them 4 years to come up with this.

I look forward to seeing their work critiqued.

I give it a 90% chance there will be a major inconsistentcy.

35% chance it looks unquestionably deliberate.

[chubbs]

Of course. I was just pointing out that you kind of presented the baseline offset as if it were an actual deviation between the datasets when it's not. Doing that essentially makes your graph pointless because your differential anomalies won't be real.

Let me break it down a little more to make it easier to understand. If UAH6 and RSS were completely reconciled the 12-month running average red line would be relatively flat. That would indicate that most of the difference between the two satellite series was due to random fluctuation. Instead the red line looks a little like the surface of the ocean with relatively large peaks and valleys. This shows there are periods when UAH6 is systematically biased vs RSS. There are both short-term and longer-term cycles in the bias between the two datasets. It may be a little fortuitous that they ended up with similar long-term trends or there could be some tuning involved. In any case will be interesting to track going forward.

[SVT450R]

That is comical.

Wouldn't be so questionable if every non temperature piece of evidence didn't say this is wrong.

Even more strange is the raw amsu channel 6 temps have run quite warm. So the discrepancy is somewhere closer to the surface.

It took them 4 years to come up with this.

I look forward to seeing their work critiqued.

I give it a 90% chance there will be a major inconsistentcy.

35% chance it looks unquestionably deliberate.

I will wait to see the data critiqued before bashing the work.  We have to keep in mind in all of this that this is a different data set when trying to compare it to the surface in the short term.  In the long term they both basically agree with each other so we should just look at it as another tool for global temps.

[StudentOfClimatology]

Let me break it down a little more to make it easier to understand. If UAH6 and RSS were completely reconciled the 12-month running average red line would be relatively flat. That would indicate that most of the difference between the two satellite series was due to random fluctuation. Instead the red line looks a little like the surface of the ocean with relatively large peaks and valleys. This shows there are periods when UAH6 is systematically biased vs RSS. There are both short-term and longer-term cycles in the bias between the two datasets. It may be a little fortuitous that they ended up with similar long-term trends or there could be some tuning involved. In any case will be interesting to track going forward.

I understand what you're trying to do, and it's a good idea. That said, none of this addresses what I was saying. You still haven't corrected for the baseline differential between the two datasets which is approximately 0.11C. This will give you a better relative vantage point for comparison.

[StudentOfClimatology]

I will wait to see the data critiqued before bashing the work. We have to keep in mind in all of this that this is a different data set when trying to compare it to the surface in the short term. In the long term they both basically agree with each other so we should just look at it as another tool for global temps.

Not to mention UAH and RSS are now in relative agreement, for the most part. I haven't looked into exactly what they've done in the upgrade, but my guess is the correction was in the right direction.

[SVT450R]

Not to mention UAH and RSS are now in relative agreement, for the most part. I haven't looked into exactly what they've done in the upgrade, but my guess is the correction was in the right direction.

I haven't either that's out of my league which is why i will let the professionals look into it and critique the data.  I think some are looking to hard into the data and not putting their bias to the side.  For example UAH and GISS agree 100% with a step up warming trend since 2011 and that hasn't changed with the update. 

[AvantHiatus]

Global temps recovering to their native state, now starting out May with a 0.152C anomaly on CFSv2. Equivalent to roughly 0.73C on GISS.

 

http://models.weatherbell.com/climate/ncep_cfsr_t2m_anom.png

[skierinvermont]

He obviously hasn't corrected for the baseline difference. The corrected data I have doesn't diverge anywhere near the extent that his comparison depicts.

 

A different baseline would have no effect upon the slope. I think chubbs is perfectly aware that the slope is what matters.

[skierinvermont]

Of course. I was just pointing out that you kind of presented the baseline offset as if it were an actual deviation between the datasets when it's not. Doing that essentially makes your graph pointless because your differential anomalies won't be real.

 

Pointless? Are you really that obtuse? I think everybody here - except you - is capable of visualizing the slope of the graph with the whole line just shifted down .11C. 

 

So either you can't see something that obvious, or you think that the baseline would actually effect the slope. Or, you're just intent on disagreeing with everything everybody else, which is incredibly annoying and puts you in the wrong most of the time.

[Polar Vortex 2014]

UAH6 isn't measuring the same thing as UAH5.6, which in turn isn't measuring the same thing as RSS. Although they all purport to be an analysis of the lower tropospheric temperature anomaly, the definition of lower troposphere is left to the whims of the respective investigators.

 

Dr. Roy Spencer makes perfectly clear that nearly half of the difference in trend between UAH6 and UAH5.6 is due to a redefinition of what the lower troposphere is. This redefinition is accomplished by reweighting the function to reduce sensitivity "to direct surface emission by the land surface, which surface thermometer data suggests is warming more rapidly than the deep troposphere."

 

Only about half of the trend change is actually due to a new method of correcting for diurnal drift. This portion of the trend change, according to Dr. Spencer is "well within our previously stated range of uncertainty for this product’s trend calculation."

 

Tl;dr version: UAH6 is measuring something funadamentally different than UAH5.6, even though they purport to measure the same thing.

[FloridaJohn]

Where is the hue and cry about satellite temperature data being "adjusted?"

[StudentOfClimatology]

Pointless? Are you really that obtuse? I think everybody here - except you - is capable of visualizing the slope of the graph with the whole line just shifted down .11C.

So either you can't see something that obvious, or you think that the baseline would actually effect the slope. Or, you're just intent on disagreeing with everything everybody else, which is incredibly annoying and puts you in the wrong most of the time.

Yeah, let's "visualize" the anomaly w/i the slope instead of actually calculating it. If I tried to present something like that, I'd be laughed out of the room.

It's so incredibly easy to adjust for something like that.

[skierinvermont]

Yeah, let's "visualize" the anomaly w/i the slope instead of actually calculating it. If I tried to present something like that, I'd be laughed out of the room.

It's so incredibly easy to adjust for something like that.

 

And the adjustment will not change the slope. So what's the point?

 

Yes we could calculate the slope if we want to be exact, but everybody else can get the picture (except you) without calculating it. 

[The_Global_Warmer]

So UAH 6.0 is an entirely different dataset

If NOAA or Giss did that and saw such abrupt warming there would be federal inquries and global outcry by denier blogs of cooking the books

[StudentOfClimatology]

And the adjustment will not change the slope. So what's the point?

Yes we could calculate the slope if we want to be exact, but everybody else can get the picture (except you) without calculating it.

So you don't want to know what the actual slope and subsequent anomalies are? Got it.

Personally, I prefer to make sure my data reflects reality. That's just me, though.

[skierinvermont]

So you don't want to know what the actual slope and subsequent anomalies are? Got it.

Personally, I prefer to make sure my data reflects reality. That's just me, though.

 

Matching the baselines has absolutely nothing to do with the slope. Do you really not understand this? Slow down for a minute and think about it.

 

And no, I don't really care what the slope is to the .001C. The overall pattern is quite obvious to me. It's negative with a lot of other ups and downs.

[StudentOfClimatology]

Matching the baselines has absolutely nothing to do with the slope. Do you really not understand this? Slow down for a minute and think about it.

And no, I don't really care what the slope is to the .001C. The overall pattern is quite obvious to me. It's negative with a lot of other ups and downs.

What?

The differential anomalies are semi-homogenous. The slope itself is irrelevant..the graph aimed to depict divergence between the two datasets on a relatively high resolution..the two base periods need to be analogous if you want to see what the differential anomalies are over the stretch of the dataset(s).

[skierinvermont]

What?

The differential anomalies are semi-homogenous. The slope itself is irrelevant..the graph aimed to depict divergence between the two datasets on a relatively high resolution..the two base periods need to be analogous if you want to see what the differential anomalies are over the stretch of the dataset(s).

 

No, the base periods don't need to be the same. 

 

If the slope is negative or positive, the datasets are divergent. If the slope is zero, the datasets are not divergent.

 

The graph shows long periods of divergence, as well as short sharp divergences. This fact would be completely unchanged if you put the two on the same baseline.

[StudentOfClimatology]

No, the base periods don't need to be the same.

If the slope is negative or positive, the datasets are divergent. If the slope is zero, the datasets are not divergent.

The graph shows long periods of divergence, as well as short sharp divergences. This fact would be completely unchanged if you put the two on the same baseline.

That has nothing to do with what I'm saying. Correcting for the offset won't change the slope, but it will give you the actual value of the deviation(s), on every relevant timescale. The UAH dataset is not actually cooler than RSS by ~0.1C.

I know you're smarter than this.

[skierinvermont]

That has nothing to do with what I'm saying. Correcting for the offset won't change the slope, but it will give you the actual value of the deviation(s), on every relevant timescale. The UAH dataset is not actually cooler than RSS by ~0.1C.

I know you're smarter than this.

 

Nobody thinks UAH is cooler by .1C. Do you really think anybody here is this stupid? Do you really think it's necessary to re-do the graph because the rest of us are too stupid to ignore the different baselines and just look at the slope?

[AvantHiatus]

The big heat

http://www.tropicaltidbits.com/analysis/models/?model=gfs&region=nhem&pkg=T2ma&runtime=2015050306&fh=0&xpos=0&ypos=109.09090672642736

[StudentOfClimatology]

Nobody thinks UAH is cooler by .1C. Do you really think anybody here is this stupid? Do you really think it's necessary to re-do the graph because the rest of us are too stupid to ignore the different baselines and just look at the slope?

It has nothing to do with being stupid, it's about getting the correct specific deviation value(s). That was the point of the graph in the first place. You'd rather "visualize" these anomalies, I'd rather know them. Whatever floats your boat.

[ORH_wxman]

So UAH 6.0 is an entirely different dataset

If NOAA or Giss did that and saw such abrupt warming there would be federal inquries and global outcry by denier blogs of cooking the books

GISS and NOAA have already changed their data sets significantly. Both completely changed their SST data source without undergoing peer review or even a mere explanation. Nobody has made much of a stink about it though.

UAHv6 will eventually get published and when it does, it can be scrutinized a bit more accurately. RSS version 3.3 was published in 2009 and nobody has published anything striking to discredit it. So that is one reason I'm not overly skeptical of the UAH changes. They are probably not perfect but they don't seem ridiculous.