NOAA Reinstates July 1936 As The Hottest Month On Record

[tacoman25]

That's where I don't get your logic.  A few have overestimated warming (per the reasons we just mentioned), but they are generally in the bounds of uncertainty.  I'm not sure how you think positive feedbacks have been overestimated based on a few years of model/observation discrepancy.  It's way too early to tell that.

 

Apparently you haven't been paying much attention to many of Will's posts.

 

Water vapor feedback has almost certainly been overestimated, as well as the positive feedback of methane release. Then's there's the matter of heat release/mixing with the oceans. It's also very possible that forcings from aerosols/clouds have been underestimated, or at other times overestimated, as the links I shared above talked about.

[ORH_wxman]

That's where I don't get your logic.  A few have overestimated warming (per the reasons we just mentioned), but they are generally in the bounds of uncertainty.  I'm not sure how you think positive feedbacks have been overestimated based on a few years of model/observation discrepancy.  It's way too early to tell that.

 

 

The current generation of models are almost all outside the range of uncertainty going on a hindcast until 1993 and then using forecast from 1993 to 2013.

 

I actually posted a paper earlier in this thread on how the CMIP3 (the previous generation of models) was better at forecasting global temperatures than the current CMIP5 models. The paper concluded that there might be a selection bias emerging in the models in trying to reconcile the warmer arctic region at the expense of the rest of the globe.

 

 

The statistical performance of the CMIP5 models is bad. The fact that they are already at 2 sigma too warm after 20 years is not an indictment of too short a time period, but an indictment at the lack of ability by the models to accurately simulate the forcing and temperature response of our climate. The CMIP5 models take into account random ENSO and volcanic events (i.e natural variability)...thus, even taking this into account and they still have <5% of the simulations within the observed trend is poor. That's why we use the sigma values and not the mean...if the mean was off by a massive 1C in 20 years as an example, but we were only 1 sigma off on the model ensembles...then we could say "still well within the model spread".

 

I'm sure there's some legit reason for this discrepency. The models are not 100% useless. Maybe its too many Chinese aerosols. Maybe they didn't accurately model the amount of deep ocean storage of heat. Maybe there was more volcanic activity than they showed. Maybe they are just too sensitive to increase in GHGs. The truth is probably some combination.

 

This is an area of research that is really intense right now in the literature. A lot of new stuff has come out in the past 3-4 years, and I'm sure it will only get more intense as the CMIP5 suite performs worse (the problem with "catching up" to the CMIP5 models is they keep accelerating the warming in BAU scenario).

[tacoman25]

The current generation of models are almost all outside the range of uncertainty going on a hindcast until 1993 and then using forecast from 1993 to 2013.

 

I actually posted a paper earlier in this thread on how the CMIP3 (the previous generation of models) was better at forecasting global temperatures than the current CMIP5 models. The paper concluded that there might be a selection bias emerging in the models in trying to reconcile the warmer arctic region at the expense of the rest of the globe.

 

 

The statistical performance of the CMIP5 models is bad. The fact that they are already at 2 sigma too warm after 20 years is not an indictment of too short a time period, but an indictment at the lack of ability by the models to accurately simulate the forcing and temperature response of our climate. The CMIP5 models take into account random ENSO and volcanic events (i.e natural variability)...thus, even taking this into account and they still have <5% of the simulations within the observed trend is poor. That's why we use the sigma values and not the mean...if the mean was off by a massive 1C in 20 years as an example, but we were only 1 sigma off on the model ensembles...then we could say "still well within the model spread".

 

I'm sure there's some legit reason for this discrepency. The models are not 100% useless. Maybe its too many Chinese aerosols. Maybe they didn't accurately model the amount of deep ocean storage of heat. Maybe there was more volcanic activity than they showed. Maybe they are just too sensitive to increase in GHGs. The truth is probably some combination.

 

This is an area of research that is really intense right now in the literature. A lot of new stuff has come out in the past 3-4 years, and I'm sure it will only get more intense as the CMIP5 suite performs worse (the problem with "catching up" to the CMIP5 models is they keep accelerating the warming in BAU scenario).

 

I don't understand the denial by some on this forum. It's not like AGW is discredited just because the climate models haven't performed very well. Things aren't black and white like that.

 

But there is a definite reluctance by many to acknowledge the models and certain scientists' shortcomings. I guess because they think if they do, it will look like they're giving in to the "other side".

[nflwxman]

The current generation of models are almost all outside the range of uncertainty going on a hindcast until 1993 and then using forecast from 1993 to 2013.

 

I actually posted a paper earlier in this thread on how the CMIP3 (the previous generation of models) was better at forecasting global temperatures than the current CMIP5 models. The paper concluded that there might be a selection bias emerging in the models in trying to reconcile the warmer arctic region at the expense of the rest of the globe.

 

 

The statistical performance of the CMIP5 models is bad. The fact that they are already at 2 sigma too warm after 20 years is not an indictment of too short a time period, but an indictment at the lack of ability by the models to accurately simulate the forcing and temperature response of our climate. The CMIP5 models take into account random ENSO and volcanic events (i.e natural variability)...thus, even taking this into account and they still have <5% of the simulations within the observed trend is poor. That's why we use the sigma values and not the mean...if the mean was off by a massive 1C in 20 years as an example, but we were only 1 sigma off on the model ensembles...then we could say "still well within the model spread".

 

I'm sure there's some legit reason for this discrepency. The models are not 100% useless. Maybe its too many Chinese aerosols. Maybe they didn't accurately model the amount of deep ocean storage of heat. Maybe there was more volcanic activity than they showed. Maybe they are just too sensitive to increase in GHGs. The truth is probably some combination.

 

This is an area of research that is really intense right now in the literature. A lot of new stuff has come out in the past 3-4 years, and I'm sure it will only get more intense as the CMIP5 suite performs worse (the problem with "catching up" to the CMIP5 models is they keep accelerating the warming in BAU scenario).

I actually don't disagree with what you said here.  As a model designer myself, I more emphasize with a critic of a model that was designed for long term projections that is getting quite literally dissected on a yearly basis by skeptics that have no agenda but to discredit the science as a whole.  If there were a reasonable way to accurately predict chaotic natural systems like ENSO and TSI, i'd be all for it. If our CIMP5 temperature mean remains below the P75 level for an extended period of time (20+ years), I'd fully agree we should re-examine short term climate sensitivity.

 

The individual members have not performed statistically well in simulating ENSO.  But, is that really what they are designed for- decadal projections?  Let's truly think about that.  I'll just say, I remain skeptical of the skepticism.  I don't believe the models have performed BAD because I don't believe they should be judged by some arbitrary end point in time when there is 100 years of hindcast data that clearly shows the physics (and even climate sensitivity) are relatively sound.  We should continue to learn and improve upon our models, but let's not miss the forest for the trees.

[michsnowfreak]

Speaking of heat....

 

90F days down 50% in MI since the 1930s

http://stevengoddard.wordpress.com/2014/07/11/ninety-degree-days-in-michigan-down-50-over-the-last-80-years/

 

Midwest and US as well

http://stevengoddard.wordpress.com/2014/04/05/frequency-of-90-degree-days-in-the-midwest-down-almost-50-over-the-past-century/

http://stevengoddard.wordpress.com/2014/06/18/frequency-of-hot-days-in-the-us-have-declined-sharply-over-the-last-century/

[Snow_Miser]

Speaking of heat....

 

90F days down 50% in MI since the 1930s

http://stevengoddard.wordpress.com/2014/07/11/ninety-degree-days-in-michigan-down-50-over-the-last-80-years/

 

Midwest and US as well

http://stevengoddard.wordpress.com/2014/04/05/frequency-of-90-degree-days-in-the-midwest-down-almost-50-over-the-past-century/

http://stevengoddard.wordpress.com/2014/06/18/frequency-of-hot-days-in-the-us-have-declined-sharply-over-the-last-century/

 

Please don't post links to Steve Goddard's blog. He is a fringe extremist who doesn't deserve as much attention as he gets. Globally, heat extremes have increased, and will continue to increase with a warmer world. 

 

 

http://link.springer.com/article/10.1007%2Fs10584-012-0668-1

[tacoman25]

Please don't post links to Steve Goddard's blog. He is a fringe extremist who doesn't deserve as much attention as he gets. 

 

Regardless of Goddard, the stats are what they are.

[Jonger]

Please don't post links to Steve Goddard's blog. He is a fringe extremist who doesn't deserve as much attention as he gets. Globally, heat extremes have increased, and will continue to increase with a warmer world.

likelihood of heat records.png

http://link.springer.com/article/10.1007%2Fs10584-012-0668-1

That's rubbish. This is the result of a high resolution of weather networks. The more stations in place, the more heat outbreaks get recorded. Sounds like more "extreme weather" nonsense.

[Snow_Miser]

That's rubbish. This is the result of a high resolution of weather networks. The more stations in place, the more heat outbreaks get recorded. Sounds like more "extreme weather" nonsense.

 

Do you have any sort of evidence that the paper I posted is "rubbish?" 

[Snow_Miser]

Regardless of Goddard, the stats are what they are.

 

This is true. However the stats globally are far different than they are for the Midwest was my point. 

[skierinvermont]

That's rubbish. This is the result of a high resolution of weather networks. The more stations in place, the more heat outbreaks get recorded. Sounds like more "extreme weather" nonsense.

 

Perhaps you need to look up the definition of 'ratio' in the dictionary. Higher resolution has no effect on the 'ratio' of extremes.

 

Please, try to pay attention. The stupid mistakes and constant mis-information are exasperating. 

[ORH_wxman]

Please don't post links to Steve Goddard's blog. He is a fringe extremist who doesn't deserve as much attention as he gets. Globally, heat extremes have increased, and will continue to increase with a warmer world. 

 

likelihood of heat records.png

 

http://link.springer.com/article/10.1007%2Fs10584-012-0668-1

 

 

Goddard posts a ton of garbage, but his stats are pretty accurate WRT the midwest/Lakes....that area has definitely seen a lack of heat extremes since the mid 20th century as I posted earlier in this thread based on the Peterson et al (2013) study....but its just one region and other regions like the Western US have seen increases.

[michsnowfreak]

Goddard posts a ton of garbage, but his stats are pretty accurate WRT the midwest/Lakes....that area has definitely seen a lack of heat extremes since the mid 20th century as I posted earlier in this thread based on the Peterson et al (2013) study....but its just one region and other regions like the Western US have seen increases.

Yes...and I cannot say it enough...I dont think I ever ONCE said or implied that the earth is not warming to SOME extent. I certainly know that what is happening GLOBALLY can be far different than in my tiny little speck on the globe. HOWEVER....in every single "study" I have seen posted on what the not-so-distant future will hold for THIS REGION (MI, the midwest, lakes, etc) it is almost unanimous saying that heatwaves and extreme heat episodes will dramatically increase. In fact, while some are divided on the winters effects (some say winters becoming much milder and less snowy, others saying snowfall increasing, though that is probably due to what has happened since 2000)....the summer has been more "clear cut". That is why it is SO funny simply knowing how much LESS heat we see now than we did in  the 30-year summer inferno of 1930-1959. I can give you any Detroit stat, but I posted goddards graph because this generalizes MI and the midwest, something I could not do in simply posting Detroits data. But that, along with the graph you posted in the other thread, shows that the decreasing number of hot days is not a localized anomaly here in Detroit, but in the region as a whole.

[michsnowfreak]

Goddard posts a ton of garbage, but his stats are pretty accurate WRT the midwest/Lakes....that area has definitely seen a lack of heat extremes since the mid 20th century as I posted earlier in this thread based on the Peterson et al (2013) study....but its just one region and other regions like the Western US have seen increases.

Obviously this is locally and NOT global, but wanted to add a bit more detroit data on how midwest heat is NOT increasing despite continuous predictions for the last 20 years or so that it will be quickly increasing dramatically (again...NOT saying that heat isnt increasing in parts of the globe, obviously it has). In fact, if anything, the average low has slightly risen while the average high has slightly fallen in the last few decades of summers.

 

Detroits 90F data can be separated into 3 periods. The cool early days, the hot mid-century, and the middle ground since the mid-20th century:

 

1874-1929: 7.6 days per year

1930-1959: 16.8 days per year

1960-2013: 12.2 days per year

 

An unusually high number of 90F days from 2010-12 brought that up a bit, but looking at trends, despite what was said at the time of those few hot summers (which included 2 of the hottest Julys on record, 2011 and 2012) it appears to be an anomaly. Take away that trio of oppressive Julys from 2010-12, and then just 3 of the remaining 12 years in the 2000-2014 period will see above average 90F days in July. If Detroit does not hit 90F on Tuesday (a decent chance, but perhaps our only chance), it would be the 4th July since 2000 (and the 3rd since 2008) to not hit 90F. Prior to 2000, the previous 4 Julys without a 90F spanned 85 years.

 

I guess what Im getting at is...while we certainly still get heatwaves (and every single solitary one of them you will find some sort of article with some scientist quoting said heatwave is magnified because of gw)...and while we certainly have years with above average number of hot days (again, see 2010, 2011, and 2012)...the number of years with below average number of hot days is easily outpacing the number of above average years so far this century. And yet, all we ever see is stories on how heat will be increasing dramatically here.

 

Looking at the average number of annual 90F days broken down by decade at various midwest sites...while each city tended to have its own trends, there is one common theme everywhere. The 1930s-1950s were the hottest, and the 2000s saw a decline in heat. I didnt include the 2010s because of the small sample size, but its pretty cut and dried just about everywhere. 2010, 2011, and 2012 above normal...2013 and 2014 below normal.

 

The xmacis age makes stats for anything you want a quick, easy thing to look up!

 

DETROIT, MI

1870s- 4

1880s- 5

1890s- 9

1900s- 6

1910s- 11

1920s- 9

1930s- 19

1940s- 16

1950s- 15

1960s- 11

1970s- 12

1980s- 13

1990s- 12

2000s- 10

 

MARQUETTE, MI

1870s- 5

1880s- 3

1890s- 5

1900s- 3

1910s- 6

1920s- 4

1930s- 7

1940s- 7

1950s- 4

1960s- 5

1970s- 5

1980s- 4

1990s- 3

2000s- 4

 

MINNEAPOLIS, MN

1870s- 10

1880s- 8

1890s- 14

1900s- 9

1910s- 12

1920s- 13

1930s- 27

1940s- 19

1950s- 16

1960s- 13

1970s- 17

1980s- 17

1990s- 8

2000s- 15

 

MILWAUKEE, WI

1870s- 6

1880s- 4

1890s- 10

1900s- 6

1910s- 8

1920s- 7

1930s- 15

1940s- 11

1950s- 11

1960s- 8

1970s- 8

1980s- 11

1990s- 10

2000s- 7

 

INDIANAPOLIS, IN

1870s- 17

1880s- 16

1890s- 31

1900s- 16

1910s- 25

1920s- 17

1930s- 32

1940s- 27

1950s- 23

1960s- 17

1970s- 13

1980s- 22

1990s- 18

2000s- 14

 

CHICAGO, IL

1870s- 8

1880s- 5

1890s- 11

1900s- 9

1910s- 14

1920s- 13

1930s- 19

1940s- 23

1950s- 28

1960s- 19

1970s- 21

1980s- 22

1990s- 16

2000s- 13

 

CLEVELAND, OH

1870s- 4

1880s- 4

1890s- 6

1900s- 2

1910s- 5

1920s- 5

1930s- 9

1940s- 18

1950s- 20

1960s- 8

1970s- 7

1980s- 10

1990s- 10

2000s- 9

 

COLUMBUS, OH

1880s- 14

1890s- 27

1900s- 18

1910s- 22

1920s- 13

1930s- 29

1940s- 28

1950s- 28

1960s- 17

1970s- 11

1980s- 16

1990s- 21

2000s- 16

[Mallow]

Definitely the opposite in the Pac NW... the same measures would almost certainly show a gradual upward trend with maybe some smaller-magnitude multi-decadal oscillations embedded within.

[ORH_wxman]

Definitely the opposite in the Pac NW... the same measures would almost certainly show a gradual upward trend with maybe some smaller-magnitude multi-decadal oscillations embedded within.

 

 

It shows up quite well on the Peterson et al graph for anomalous heat/cold waves.

 

 

 

 

 

 

 

The southwest has also seen much more heat recently than the rest of the country...though they have the 1930s spike which is definitely more muted in the PAC NW.

[The_Global_Warmer]

Seasonally matters as well.

 

I would presume Russia has had more heat waves in the warm period of the year.

 

But recently outside of 2013-14 has had some very cold winters.

[tacoman25]

One thing that may be a factor in comparing the Western trends to the Midwest/East is that much of the west has seen more urbanized development the past 40-50 years, while the Midwest/East was already more urbanized and has seen substantially less development and growth.

[sokolow]

One thing that may be a factor in comparing the Western trends to the Midwest/East is that much of the west has seen more urbanized development the past 40-50 years, while the Midwest/East was already more urbanized and has seen substantially less development and growth.

How much do you think is not accounted for through present homogenization and correction methods aimed at mitigating UHI biases?

[ORH_wxman]

How much do you think is not accounted for through present homogenization and correction methods aimed at mitigating UHI biases?

 

 

If referring to the peterson et al study graph I posted above...it would not be taken into account. I believe they only filtered out TOBS in that study. (I'll have to double check when home later though..but this is just on memory)

 

Sites like Phoenix and Las Vegas have had some pretty severe UHI contamination...which shows up in the nightime minimum temps.

[Mallow]

If referring to the peterson et al study graph I posted above...it would not be taken into account. I believe they only filtered out TOBS in that study. (I'll have to double check when home later though..but this is just on memory)

 

Sites like Phoenix and Las Vegas have had some pretty severe UHI contamination...which shows up in the nightime minimum temps.

 

But there are far more rural than urban sites, especially in the northwest. Urbanization would only affect a small handful of the weather stations, so I don't think that could account for much if any of the strong trend you see there.

[ORH_wxman]

But there are far more rural than urban sites, especially in the northwest. Urbanization would only affect a small handful of the weather stations, so I don't think that could account for much if any of the strong trend you see there.

 

 

I don't think it does either. They probably wouldn't get the study published if it were only urban areas supporting their trends. Also, Tmax has not been affected much by UHI...if anything, in the arid regions, it actually causes a cool bias.

 

Its been the Tmins mostly affected. Though their deifnition of anomalous heat and cold events is based on anomalies for both Tmin or Tmax. Unlike some studies which only focus on Tmax for heat waves and Tmin for cold waves.

[tacoman25]

But there are far more rural than urban sites, especially in the northwest. Urbanization would only affect a small handful of the weather stations, so I don't think that could account for much if any of the strong trend you see there.

 

Take a look at the standards they used for the study, though. They only used long term stations with less than 10% missing data. Only 711 stations nationwide qualified, and I can guarantee the vast majority of those were east of the Rockies.

 

I also know from personal research that most rural stations in the West would not be able to qualify.

[michsnowfreak]

Here is another lack-of-heat stat for a midwestern site. Barring a major overpeforming temperature, July 2014 will finish with just 13 days seeing a high at or above 80F at Detroit. This ties for the 4th lowest on record. Only 9 times since 1874, or 141 years, has Detroit seen 13 or less days with maxes of 80F+ in July. And 3 of those 9 are since 2000.

 

1884 - 13 days

1891 - 9 days

1895 - 13 days

1920 - 12 days

1924 - 13 days

1992 - 13 days

2000 - 10 days

2009 - 13 days

2014 - 13 days

 

So if you want to rank in the order of least amount of 80F days....

1.) 1891 - 9 days

2.) 2000 - 10 days

3.) 1920 - 12 days

4.) 1884, 1895, 1924, 1992, 2009, 2014 - 13 days

[Jonger]

Here is another lack-of-heat stat for a midwestern site. Barring a major overpeforming temperature, July 2014 will finish with just 13 days seeing a high at or above 80F at Detroit. This ties for the 4th lowest on record. Only 9 times since 1874, or 141 years, has Detroit seen 13 or less days with maxes of 80F+ in July. And 3 of those 9 are since 2000.

1884 - 13 days

1891 - 9 days

1895 - 13 days

1920 - 12 days

1924 - 13 days

1992 - 13 days

2000 - 10 days

2009 - 13 days

2014 - 13 days

So if you want to rank in the order of least amount of 80F days....

1.) 1891 - 9 days

2.) 2000 - 10 days

3.) 1920 - 12 days

4.) 1884, 1895, 1924, 1992, 2009, 2014 - 13 days

I work outside about 3 days a week, from Sun up till Sun down. It's been amazing. I can only recall being sweaty just a handful of days all summer.