Coldest Winter in the CONUS (so far) since the late 1970s?

[blizzard1024]

Weather balloons were launched well before the satellite era, but sea ice coverage was crap before 1979. Shipping records were the main technique at reconstructing the arctic ice coverage, we know how well shipping records worked out for sea temps. Junk.

 

You can spend hours and hours of time trying to mesh datasets and it is still an educated guess. There are so many problems with comparing older and coarser datasets to current finer ones I don't know even where to begin. The whole CAGW has become a religion and big business. The objectivity I fear could be gone or so tainted that I am quite skeptical of the extraordinary claims that are often made by some...NOT everyone...some.  I do agree with the basic premise that some warming (~1K) would occur with a doubling of CO2 from preindustrial times. That is on pretty solid ground. But how do we know the climate was in stasis in the 1800s? We were heading out of the LIA and much of the warming we have seen could easily be attributed to that plus some from CO2. Climate sensitivity is the 64 million dollar question. It has been warming for 100+ years....yes...how much is up for debate but we know this is basically true. But how much more are we going to see??  my scientific opinion is maybe another .5 to .7C or so assuming we have seen .5C or so from the 1880s from CO2. But this will be spread out over many many decades and natural variability could easily swamp it. Why do you have to label people also? Now I am a contrarian?? What does that make you (speaking to Studentof Climatology NOT Jonger)....a "believer". I would say that fits the bill well for many on this forum ...."believers"....they have faith. good for you folks. 

 

Anyway, I get to sit in on a briefing on AR5 tomorrow...maybe there will be something that will convince me otherwise. who knows.  I do have an open mind....

[blizzard1024]

Weather balloons were launched well before the satellite era, but sea ice coverage was crap before 1979. Shipping records were the main technique at reconstructing the arctic ice coverage, we know how well shipping records worked out for sea temps. Junk.

Yeah but weather balloons are not accurate because they show that water vapor is decreasing in the upper troposphere which is counter to the predetermined model based positive water vapor feedback. So it can't be true.... /sarc 

[Jonger]

Yeah but weather balloons are not accurate because they show that water vapor is decreasing in the upper troposphere which is counter to the predetermined model based positive water vapor feedback. So it can't be true.... /sarc 

 

My point is that we have a decent picture of land based observation, ocean and arctic sea ice data isn't so clear before ARGO and the satellite era. I'm sure the usual suspects will jump all over this, but they see this topic with warmer goggles.

[sokolow]

You can spend hours and hours of time trying to mesh datasets and it is still an educated guess. There are so many problems with comparing older and coarser datasets to current finer ones I don't know even where to begin.

Actually I am interested in this. Would you maybe have some time to provide some common examples from the climate record you think currently exceed our ability to analyze? And an explanation of why the problem is intractable?

[New Englander]

Been cold here but not very cold for the whole month of January for SNE standards..... These kind of winters and cold outbreaks have not happened in a little while and it is easy to forget. 

[The_Global_Warmer]

The cold has been impressive.  But the warmth out West has been equally impressive

 

 

 

 

[The_Global_Warmer]

[blizzard1024]

 

 

Based on this the CONUS is having a mild January. Didn't realize how warm it was out west. Of course, most

people live east of the Rockies so the perception is different than reality. December was 2.39F below the 1981-2010

normal period for the CONUS.  This map would suggest ( just eyeballing it ) that it will be above normal slightly. This winter

probably will end up near normal....unless we get major arctic cold again later in February. 

[blizzard1024]

Actually I am interested in this. Would you maybe have some time to provide some common examples from the climate record you think currently exceed our ability to analyze? And an explanation of why the problem is intractable?

 

The sea ice record before the satellite era, SST and ocean temperatures pre-Argo, even land based stations that have moved or have been encroached by buildings and pavement. Also land use changes like fields changing to forests in the east as an example. Even the surface record is rife with a lot of potential inconsistencies because many stations are measuring a microclimate that could easily be changed by growing forests, pavement, buildings. Many of these stations have been moved so are you really measuring the same climate? I prefer the satellite temperature data the best. it measures a slab of the lower troposphere. It also measures the mid and upper troposphere and stratosphere. The satellite measurements of the sea ice and SSTs are also good. Unfortunately a lot of this data was not present before the 1970s which we know was globally a cool period. Time of observation changes too have introduced a warm bias in the past that has to be adjusted down. Has UHI been properly accounted for? I know that scientists have

spent hundreds of hours on trying to mesh datasets with as much accuracy as possible. Still that doesn't mean that they are correct. We just don't know. There is so much uncertainty the farther back you go.  There is also so much uncertainty regarding how well the current class of GCMs actually replicate our climate system. It doesn't mean these scientists are idiots or wasting time and money...it is just an intensely complicated problem. I think there is over confidence in these predictions based on climate models. My background is weather forecasting so I know how things can go wrong real fast even with high res computer models, hours of analysis time etc. The climate forecast problem is no different.   I think this is why many METs tend to be skeptical of the higher end warming scenarios...most of us believe there will be some warming including me. The higher end scenarios that rely on heavily model based positive feedbacks are the most uncertain IMO. 

[e pluribus unum]

Based on this the CONUS is having a mild January. Didn't realize how warm it was out west. Of course, most

people live east of the Rockies so the perception is different than reality. December was 2.39F below the 1981-2010

normal period for the CONUS.  This map would suggest ( just eyeballing it ) that it will be above normal slightly. This winter

probably will end up near normal....unless we get major arctic cold again later in February. 

 

Keep in mind that the West had a very cold December, and February looks to reshuffle things again back towards a cold West. In spite of the ridgy January, 2013-14 has a good chance of ending up cooler than normal across a lot of the West.

[FloridaJohn]

Unfortunately a lot of this data was not present before the 1970s which we know was globally a cool period.

 

If the data record before the 1970s in in error, how do you know this is true?

 

 

Time of observation changes too have introduced a warm bias in the past that has to be adjusted down. Has UHI been properly accounted for? I know that scientists have spent hundreds of hours on trying to mesh datasets with as much accuracy as possible. Still that doesn't mean that they are correct. We just don't know. There is so much uncertainty the farther back you go.

 

Can't this uncertainty be defined through the use of error bars/ranges?

[ORH_wxman]

If the data record before the 1970s in in error, how do you know this is true?

 

 

 

Can't this uncertainty be defined through the use of error bars/ranges?

 

 

UHI adjustments have been pretty subjective and generally unreliable form what I have seen (in fact there's been papers in recent years about their unreliability). But they aren't a huge part of the temperature dataset so their impact isn't that large. Even if the UHI had a gross warm bias in the adjustments, it probably only changes the longterm trend by roughly 10-20%.

[sokolow]

The sea ice record before the satellite era, SST and ocean temperatures pre-Argo, even land based stations that have moved or have been encroached by buildings and pavement. Also land use changes like fields changing to forests in the east as an example. Even the surface record is rife with a lot of potential inconsistencies because many stations are measuring a microclimate that could easily be changed by growing forests, pavement, buildings. Many of these stations have been moved so are you really measuring the same climate? I prefer the satellite temperature data the best. it measures a slab of the lower troposphere. It also measures the mid and upper troposphere and stratosphere. The satellite measurements of the sea ice and SSTs are also good. Unfortunately a lot of this data was not present before the 1970s which we know was globally a cool period. Time of observation changes too have introduced a warm bias in the past that has to be adjusted down. Has UHI been properly accounted for? I know that scientists have

spent hundreds of hours on trying to mesh datasets with as much accuracy as possible. Still that doesn't mean that they are correct. We just don't know. There is so much uncertainty the farther back you go. There is also so much uncertainty regarding how well the current class of GCMs actually replicate our climate system. It doesn't mean these scientists are idiots or wasting time and money...it is just an intensely complicated problem. I think there is over confidence in these predictions based on climate models. My background is weather forecasting so I know how things can go wrong real fast even with high res computer models, hours of analysis time etc. The climate forecast problem is no different. I think this is why many METs tend to be skeptical of the higher end warming scenarios...most of us believe there will be some warming including me. The higher end scenarios that rely on heavily model based positive feedbacks are the most uncertain IMO.

The reason i asked is because well, yes, it is difficult to match datasets of differing resolution and type temporally and spatially. This is notoriously the case for persons trying to give evidence-based guidance to local planning authorities. Its also a prominent issue in paleoclimate. But it is not all shifting sands and hopeless uncertainty.

Thing is, what you're saying is inadequate: the inconsistencies you propose in the instrumental record -- given your expertise and training you should be able to put a range of numbers around that potential error.

You might have guessed from the other thread that I'm interested in glaciers. The worldwide temperature reconstructions derived from glaciers as proxies are informative on century out to millenial timescales. It is getting warmer. The warming is unusual and abrupt. It is worldwide. Temperature reconstructions using glacial data support the instrumental record. Glaciers are approaching minima they havent seen for five or six thousand years. Some places we're looking at the imminent loss of land ice that has an age at base of eleven thousand years.

All sorts of factors -- precipitation, temperature, slope, orientation, size & thickness -- affect glacier length and mass balance and as in any scholarly endeavor demand care. All sorts of datasets require meshing. However at the end of the day there's not a lot that's ambiguous about "these melting glaciers are spitting out artifacts and human bodies and trees and crap that haven't seen the light of day since the Times of Methuselah".

If you are not confident in the models' ability to deal with feedbacks and ae skeptical of so called long tail of extreme warming scenarios I suggest you consider from the fate of land ice here that is no long tail of *mild* warming scenarios and that we have already rapidly reversed a cooling trend and matched the temperatures of the mid-Holocene maximum. "Some warming" on top of that puts us back to the last interglacial.

[blizzard1024]

The reason i asked is because well, yes, it is difficult to match datasets of differing resolution and type temporally and spatially. This is notoriously the case for persons trying to give evidence-based guidance to local planning authorities. Its also a prominent issue in paleoclimate. But it is not all shifting sands and hopeless uncertainty.

Thing is, what you're saying is inadequate: the inconsistencies you propose in the instrumental record -- given your expertise and training you should be able to put a range of numbers around that potential error.

You might have guessed from the other thread that I'm interested in glaciers. The worldwide temperature reconstructions derived from glaciers as proxies are informative on century out to millenial timescales. It is getting warmer. The warming is unusual and abrupt. It is worldwide. Temperature reconstructions using glacial data support the instrumental record. Glaciers are approaching minima they havent seen for five or six thousand years. Some places we're looking at the imminent loss of land ice that has an age at base of eleven thousand years.

All sorts of factors -- precipitation, temperature, slope, orientation, size & thickness -- affect glacier length and mass balance and as in any scholarly endeavor demand care. All sorts of datasets require meshing. However at the end of the day there's not a lot that's ambiguous about "these melting glaciers are spitting out artifacts and human bodies and trees and crap that haven't seen the light of day since the Times of Methuselah".

If you are not confident in the models' ability to deal with feedbacks and ae skeptical of so called long tail of extreme warming scenarios I suggest you consider from the fate of land ice here that is no long tail of *mild* warming scenarios and that we have already rapidly reversed a cooling trend and matched the temperatures of the mid-Holocene maximum. "Some warming" on top of that puts us back to the last interglacial.

 

Many glaciers advanced through the little ice age and reached their maximum extent sometime in the 1800s, in coastal Alaska around 1890. There has been a decline related to the current warm period. Glaciers also retreated during the medieval warm period. How do we know this time is unusual or not just part of an overall reduction in land glaciers related to the present interglacial. Plus depending on the glacier, glacial mass balance can have time scales of hundreds of years or more. Again warming of the climate system is not what is in doubt here. Glaciers, sea ice, snow cover etc all decline in a warming climate. I don't think we have enough confidence to know if the majority of it is from increasing CO2 and GHGs or more from natural variability. As far as error bars on the other datasets, not sure what you are after here. I would say they are very large....especially for coarse data sets meshed with fine high resolution. Or satellite data vs inferred data from before the satellite data. Ship reports of sea ice vs satellite data of ice. You can try to mesh these...but come on, the errors are going to be huge and uncertain. 

[sokolow]

Many glaciers advanced through the little ice age and reached their maximum extent sometime in the 1800s, in coastal Alaska around 1890. There has been a decline related to the current warm period. Glaciers also retreated during the medieval warm period. How do we know this time is unusual or not just part of an overall reduction in land glaciers related to the present interglacial.

Because using careful, comparative reconstructions of past glacier thickness, length, and area via stratigraphy, radiocarbon dating, exposure, morphology &c. to gauge glacial advance and retreat on historical to millenial timescales and according to (say) Johannes Oerlemans and Wilfried Haeberli, worldwide glacial volumes and lengths and rates of loss by the end of the 20th century approached or exceeded natural variability of the last 5kya, and possibly the entire Holocene. That's unusual.

[blizzard1024]

Here is the temperatures for December and January in the CONUS...

computed from mostly airport sites. There are adjustments made for

UHI...that's it. The data started in 1973 and it shows this winter so

far is not quite to the level that would make it coldest since the late 70s, but

it is pretty close so far. This data shows that 5 out of the last 7 DEC/JANs

in the CONUS will be below the long term mean. 

 

[Msalgado]

Do you know what the main problem is with using mainly airport sites?  There's far more airports in the blue section of this map than there is in the warm section yet for the last 30 days the country as a whole has been warmer than normal.  

 

http://www.cpc.ncep.noaa.gov/products/tanal/30day/mean/20140201.30day.mean.F.gif

 

We'll see.  Still got a huge part of the winter to go but I have a hard time thinking that this will be a cool one when its all said and done.

[skierinvermont]

Great. Un-peer reviewed work by Spencer instead of using NCDC data. Don't like the result? Make up one you do.

 

http://www.ncdc.noaa.gov/cag/time-series/us/110/00/tmp/2/01/1973-2013?base_prd=true&firstbaseyear=1901&lastbaseyear=2000

 

 

[Jonger]

Here is the temperatures for December and January in the CONUS...

computed from mostly airport sites. There are adjustments made for

UHI...that's it. The data started in 1973 and it shows this winter so

far is not quite to the level that would make it coldest since the late 70s, but

it is pretty close so far. This data shows that 5 out of the last 7 DEC/JANs

in the CONUS will be below the long term mean. 

 

DecJan-USA48-temps-1973-2014.png

 

Good find.

[Jonger]

Great. Un-peer reviewed work by Spencer instead of using NCDC data. Don't like the result? Make up one you do.

 

http://www.ncdc.noaa.gov/cag/time-series/us/110/00/tmp/2/01/1973-2013?base_prd=true&firstbaseyear=1901&lastbaseyear=2000

 

Virtually no trend over the last 80 years.

 

http://www.ncdc.noaa.gov/cag/time-series/us/110/00/tmp/2/01/1973-2013?base_prd=true&firstbaseyear=1901&lastbaseyear=2000

[skierinvermont]

Good find.

 

There are appear to be a number of obvious flaws in this analysis which I will elaborate on later. 

 

Let's just start out with the fact that UAH shows .22C/decade of warming over the U.S. while his proprietary ISH adjusted data shows just .01C/decade (essentially zero) over the same period. How does one explain that?

 

Oh and let's not forget that Roy pretends to hate adjustments but really what he wants to do is pick and choose and create with crazy statistics that don't represent reality, his own very strongly negative adjustments that literally take a very large positive trend and turn it into zero.

 

The same could be said for blizzard1024... I thought you didn't like adjustments?

 

Is UAH just making up the very large significant positive trend?

 

Oh and let's just forget about all the obvious (and scientific) physical evidence of warming in the U.S. since the 70s. 

[Msalgado]

What?  There's a pretty clear upward trend over the past 80 years.  

[e pluribus unum]

Do you know what the main problem is with using mainly airport sites?  There's far more airports in the blue section of this map than there is in the warm section yet for the last 30 days the country as a whole has been warmer than normal.  

 

http://www.cpc.ncep.noaa.gov/products/tanal/30day/mean/20140201.30day.mean.F.gif

 

We'll see.  Still got a huge part of the winter to go but I have a hard time thinking that this will be a cool one when its all said and done.

 

Why?

 

December was the 21st coldest on record, and more than two degrees below the 20th century average.

 

January data hasn't been finalized yet, but if there was a national warm anomaly it would've been very slight. A solid half of the lower 48 was cold.

 

 

This month is starting out very cold across the country. We'll see where it goes with a potential pattern shakeup midmonth, but the first half of the month is going to be significantly below average for the country overall.

 

If anything, it would be surprising at this point if the winter didn't wind up colder than average, and possibly by a significant amount.

[blizzard1024]

There are appear to be a number of obvious flaws in this analysis which I will elaborate on later. 

 

Let's just start out with the fact that UAH shows .22C/decade of warming over the U.S. while his proprietary ISH adjusted data shows just .01C/decade (essentially zero) over the same period. How does one explain that?

 

Oh and let's not forget that Roy pretends to hate adjustments but really what he wants to do is pick and choose and create with crazy statistics that don't represent reality, his own very strongly negative adjustments that literally take a very large positive trend and turn it into zero.

 

The same could be said for blizzard1024... I thought you didn't like adjustments?

 

Is UAH just making up the very large significant positive trend?

 

Oh and let's just forget about all the obvious (and scientific) physical evidence of warming in the U.S. since the 70s. 

 

the trend is based on two things in the UAH...two major volcanic eruptions and an epic el nino....plus the 70s when the good satellite data began was a cool period based on the PDO phase and the AMO phase. So it shud be a little warmer now...anyway you can believe in your climate alarmist religion...that's fine. you are a real climate denier. I and others who agree with me are climate realists.... 

[skierinvermont]

the trend is based on two things in the UAH...two major volcanic eruptions and an epic el nino....plus the 70s when the good satellite data began was a cool period based on the PDO phase and the AMO phase. So it shud be a little warmer now...anyway you can believe in your climate alarmist religion...that's fine. you are a real climate denier. I and others who agree with me are climate realists.... 

 

Wait now I'm confused.. are you saying UAH is right and the U.S. has warmed a lot since the 70s, or are you saying Spencer's proprietary unpublished surface temperature series is right and the U.S. hasn't warmed at all since the 70s?

 

Please make up your mind I'm getting really confused. All I got out of your last reply was you are right and I am wrong, but I don't really know about what exactly.

[blizzard1024]

Wait now I'm confused.. are you saying UAH is right and the U.S. has warmed a lot since the 70s, or are you saying Spencer's proprietary unpublished surface temperature series is right and the U.S. hasn't warmed at all since the 70s?

 

Please make up your mind I'm getting really confused. All I got out of your last reply was you are right and I am wrong, but I don't really know about what exactly.

 

Actually, I think the UAH data is pretty accurate. I have never stated anything to the contrary. You made that assumption. You mention that Spencer's dataset does not accurately sample the west, well there isn't that many stations out west plus it is well known that UHI has severely contaminated data around populated areas in the west. Hence there is a warm bias that has to be adjusted down. How far down do they adjust it?? What is enough? hard to tell...

[blizzard1024]

Here is the latest CFS for December through Feb 5, 2014. Most of the CONUS is below average, anomaly of -1.550C.  Pretty impressive. Globally the CFS average has just dropped to   -.2C   as well...

 

 

[skierinvermont]

Actually, I think the UAH data is pretty accurate. I have never stated anything to the contrary. You made that assumption. You mention that Spencer's dataset does not accurately sample the west, well there isn't that many stations out west plus it is well known that UHI has severely contaminated data around populated areas in the west. Hence there is a warm bias that has to be adjusted down. How far down do they adjust it?? What is enough? hard to tell...

 

Both data sets cannot be right. If you are saying UAH is right, then Spencer's surface methodology is wrong. Pick one.

[Jonger]

Actually, I think the UAH data is pretty accurate. I have never stated anything to the contrary. You made that assumption. You mention that Spencer's dataset does not accurately sample the west, well there isn't that many stations out west plus it is well known that UHI has severely contaminated data around populated areas in the west. Hence there is a warm bias that has to be adjusted down. How far down do they adjust it?? What is enough? hard to tell...

 

There are always going to be problems with anything involving human invention and science, UAH measurements included. The best hope we have for an accurate dataset is lower atmospheric, the system is less tainted by landform changes and measuring discrepancies. The biggest bummer is that the dataset starts in a cold period, so the dataset obviously doesn't include the warm 1930-1950's period, so we can't compare it to the other land based datasets. It would have been wonderful to get a snapshot of a pre-UHI tainted earth and see the impact over time.

[skierinvermont]

There are always going to be problems with anything involving human invention and science, UAH measurements included. The best hope we have for an accurate dataset is lower atmospheric, the system is less tainted by landform changes and measuring discrepancies. The biggest bummer is that the dataset starts in a cold period, so the dataset obviously doesn't include the warm 1930-1950's period, so we can't compare it to the other land based datasets. It would have been wonderful to get a snapshot of a pre-UHI tainted earth and see the impact over time.

 

We have that. There are hundreds of well sited rural thermometers all over the world and the mean trend from them is identical to GISS/Had4/BEST/NCDC. Moreover, all the aforementioned data sets perform well documented UHI adjustments when necessary and/or avoid UHI contaminated data.