Global Annual Daily Temperatures, 1929-2010

Mi Cro

While standing outside in the middle of a cold night tending my telescope, I started wondering about how much heat is getting radiated into space under clear skys, that increases in CO2 should be reducing that radiation, and whether or not it would be detectable in temperature records.

I dug around on NOAA's data site looking for hourly data thinking I could get some rate of change data from say the 30's and compare that to current data. Unfortunately, that data isn't free. I did find I could download their Global Summary of Days data.

Free is free, so what the heck.

GSOD data consists of year by year records going back to 1929 for weather stations around the globe, some 426,000 station reports, 112 million daily records for 29,903 calendar days.

The methodology I used was to take the daily maximum temperature, and subtracted the morning low temperature. That provides the energy into the planet, I then took the Max temp, and subtracted tomorrow mornings low temp, which gives me the energy lost overnight.

I originally wanted to mine the data for clear nights, but decided that there were to many ways for that to go wrong, and would never be untainted.

So, I process everyday that meets the following criteria:
   The Min and Max temps for today, and tomorrows low are in the range of 199 to -199.
       The record set uses 10000 on days that the temp is missing.
   A record exists in the data for tomorrow.

You can see from the sample count, there are some years that have a lot more data than others, I would expect there to be a range of quality in the data.

According to this:

http://www7.ncdc.noaa.gov/CDO/GSOD_DESC.txt

Global summary of day data for 18 surface meteorological elements
are derived from the synoptic/hourly observations contained in
USAF DATSAV3 Surface data and Federal Climate Complex Integrated
Surface Data (ISD).

Or more simply it's what I have.

Once I have this filtered data(112,426,376 samples), I can generate various graphs.

One of first was an Annual Average of the daily Temperature Increase and Cooling for all stations. And was amazed at how close the 2 values were, all the while changing every year.

I then did one of the difference between them.

And then lastly, The Min and Max temps plotted along with the rising and falling data.

If you calculate the radiation balance over the 80 some years, there is over all a -0.0198 degree F/Day (Edited to add '/Day') imbalance, where it's a little colder tomorrow morning than it was today.

I'm working on some Monthly averages where you can see Rising and Falling values change with the seasons.

*Edit. I updated the graphs of the averages. The new ones I Average the days value from all of the stations into one value, then I average the 365 days per year, this ways Days get the correct weighting.
I decided to just add the new graphs.

The sql code/data can be made available for anyone wanting it, export restriction applicable.

Comments

I realized I averaged the annual values incorrectly, by undervaluing the difference by Day of month. So I redid the various tables and graphs. The visual differences are very minor. But I wanted to try to reduce the imbalance between Rising and Falling temps.

It didn't really work.
I took the daily average imbalance, and multiplied it by the number of days I had records for in each year.
Annual Imbalance.

—

Never is a long time.

Mi Cro | 07/22/11 | 13:22 PM

Link

Interesting article Mi Cro, did you also see this article by News Staff at _http://www.science20.com/news_articles/where_does_earths_44_trillion_watts_heat_come-80917

—

Make love not war

Helen Barratt | 07/22/11 | 13:32 PM

Link

Yes I did.
I still can't quite get how Temps go up when on average it's a little colder tomorrow than it was today.

But it does seem that the shape of the 10 year Rising/Falling temp and 10 yr temp curves are very close to each other.

—

Never is a long time.

Mi Cro | 07/22/11 | 13:58 PM

Link

I am not sure what you are trying to show here, firstly:"The methodology I used was to take the daily maximum temperature, and subtracted the morning low temperature. That provides the energy into the planet,"
If the oceans or a similar large heat sink is absorbing energy then this is simply not true at all. The daily maximum would be less than it would be if the heat wasn't absorbed. Similarly for the overnight temperature.
Are you trying to say that the earth isn't absorbing heat in spite of decadal average temps increasing and the top layer of the ocean heating up?

—

Thor Russell

Thor Russell | 02/01/12 | 21:07 PM

Link

These are all land based stations, the NOAA data set doesn't include the location of any of the ocean based stations, since there's no location I don't include them.
But consider a single station, in the morning (usually) it's radiated away some amount of heat and obtains the low temp of the day. The Sun comes up, and inputs some amount of energy raising the temp to the daily maximum. The ground and building absorb some amount of that heat, but however much is absorbed the temp stabilized. The sun goes down, all the absorbed heat radiated into the air, and the air(and ground) radiated into space. That heat is lost from the earth's system forever. If there's a net positive the air and ground are a little warmer. All of this shows up in the temperature record.
If you look at the daily data, starting in January there's a positive diff(in the northern Hemisphere), at the end of July diff goes negative as the days shorten and there's more time to radiate into space.

This is Daily Diff * 100
The diff in the yearly charts is the average of these values. I did figure out that the diff on these charts is much larger than it should be, and I'm reworking my sql to correct these errors (I truncated the source data at the decimal point losing the tenths of a degree the record contained, the chart in the WSJ article contains the full data).

But, the minimum temp in the Southern Hemisphere is still going down, and while the max temp is trending up a little, you can see that daily rising temp is the source, and that falling is still almost an exact match to it.

What I think this proves is that there is no loss of cooling, while co2 has increased, and in the SH the reason temps are going up is the sun's daily input is increasing.

Lastly, I got my data from NOAA, the same source that GISS uses, and much of it is the same data used in the CRU Hadcrut3 data.

http://maps.google.com/maps/ms?msid=218403362436074259636.0004aab94d7a95fab9567&msa=0

Here's a map with both the NCDC and CRU stations. I'm working on code to create a map for the specific stations in the same area I average(such as the SH chart on the other page).

—

Never is a long time.

Mi Cro | 02/02/12 | 10:29 AM

Link

Firstly lets go back a few steps so you know where I'm coming from

1. You can tell whether an object is heating up because of increased insulation or increased energy input because of its temperature profile. If the sun's output increased all layers of the atmosphere would heat up. If insulation was the cause, the inner layers would heat up, the outermost layer would cool down. This is a pretty common principle, you see it with thermos flasks, double glazed vs single glazed windows etc.

2. CO2 and other gasses trap around 1.5W/m^2 additional heat since the industrial revolution. The sun's output doesn't vary enough to equal this. The average insolatation is 1366W/m^2 divide by 4 because of geometry. It varies something like 0.1% which is only 0.34 W/m^2. The CO2 effect is much larger and also the suns output is cyclical causing it to mostly cancel. Furthermore the suns output has if anything dropped in the last 30 years with this cycle being especially weak. This appears to completely rule out the sun as a cause of warming recently.

3. Can you please spell out our logic for me? Are you saying that if insulation increased the average temperature, then the day/night temps would go say from 20-9C to 22C-13C with a 11C day/night difference becoming 9C or from 20-9C to 24C-11C with a 11C day/night difference becoming 13C?

Which situation and why, can you give a simple equation to show?

4. I just don't think your energy budget claim is justified. You are forgetting about wind. If a wind blows through the day and takes warm air away from a station that is measured say offshore then that heat will not be counted by your method. To me this appears to completely rule out being able to draw any global conclusion from this data. If in the last 30 years wind patterns changed, because of say PDO or a consequence of global warming itself that would completely change the data. e.g. if it became common for a breeze from land to sea in the middle of the day, this would introduce a systemic error in your calculations. Also if there are now more clear nights with less cloud cover over the land that would also skew the results.
Finally if the sun was doing the warming, there should be an unmistakable 11 year sunspot signal in your data.

—

Thor Russell

Thor Russell | 02/02/12 | 17:14 PM

Link

2. I don't think we can rule the Sun out yet. And the Climate Sensitivity used in all of the GCM's is selected with the intent of making a raising CO2 raise the temp generated so it matches measured values. I know of no experiments that actually attempt to measure CS. So it's an open question if the value used is accurate.

3. What I expect is that since CO2 doesn't block shortwave IR, solar output is a main driver of the daily temperature increase(when averaged across many stations over many days, ie minimizing weathers impact). While CO2 would reduce daily falling temps. Over the long term of increasing CO2 levels, I would expect Falling to decrease when compared to Rising. But even while rising changes many degrees in various years, falling is incredibly close, some years diff is slightly larger, other years it's slightly smaller. In no case that I've found yet does it monotonically increase along with Co2. Also if you look at the daily diff above, it isn't a constant, and yet the annual value is imo remarkably close to zero.

4. I'm using the same data that's being used to prove AGW, in the past when I've complained about the accuracy of the data I was told that errors should cancel out. Beside there are many, many stations no where near large bodies of water. Oh, and during the day, sea breeze is from water to land, and at night it's land to water due to rapid heating and cooling of land compared to water.

—

Never is a long time.

Mi Cro | 02/02/12 | 17:23 PM

Link

2. Climate sensitivity to a doubling in CO2 has been calculated for a simple earth with no feedbacks for a long time, it is about 1.2C for each doubling in CO2. This can be done with atmospheric physics etc. The water vapor feedback is simple to understand (there are many references online) and approximately doubles the sensitivity to 2.4C. The best guess according to climate scientists is about 3C temperature rise for a doubling in CO2, with still significant uncertainty. While some models try to fit the climate sensitivity to CO2, there is strong physical evidence not requiring simulations that it is somewhere around the 2.4-3C. If there were no climate feedbacks it wouldn't be such a problem, but that is clearly not the world we live in.
If you are to claim that the temperature rise is not caused by CO2, then you need to explain TWO things
1. What stops the CO2/water feedback from causing this temperature rise.
2. What amplifies the suns effect (or some other effect to cause the rise) by >5* but does not do the same to CO2. There isn't a strong if at all significant correlation between Galactic Cosmic Rays and temperature to do this.

If the sun is getting colder how can we not rule the sun out? Also where are the 11 year sun cycles in your data. This would provide both an independent sanity check, and is required to appear if the sun is in fact the cause.

3. I'm sorry but I still don't understand fully your methods here.

If the climate is in equilibrium then surely the average daily falling temp equals the daily rising one no matter what is happening? The climate isn't in equilibrium, but the long term temp increase is 0.1C per decade, which is .000027 deg per day, so how is that going to show up anywhere unless you look at the smoothed yearly global temps.

Do you expect the average difference between max daytime temp and min nighttime temp to be decreasing or not?

4. Land data is a piece of the evidence to me, but by itself doesn't prove it. There is consensus even among former sceptics now that the world has warmed, land data, ocean heat data, and satellite data all show that. So if we accept that world is warming, to me the pattern with the outer atmosphere cooling fits with insulation well. If you reject insulation, how do you explain this pattern? I regard such global data more reliable than the land based temperature measurements, and you seem to think there may be issues with them also.

My point about wind etc is to highlight potential ways the data could be skewed.
If the pattern of wind changed over the decades e.g. more/less breeze whenever during the day/night, it would still affect your data in terms of energy balance. If the wind was to/from a place with no/less records it could definitely introduce error. Also as I said before if there was a global change on the land based stations to less cloud cover at night that would also explain the effect. If this was balanced by increased cloud cover in places with no records e.g. ocean the global effect would cancel out, but not do so in the land records, again making them less reliable than global measurements on such details.

My point is that because of potential effects like this, it makes sense to trust the global satellite record over the land based ones in terms of energy balance etc.

—

Thor Russell

Thor Russell | 02/02/12 | 18:28 PM

Link

2 Oh, I understand the co2(heat)/water vapor feedback loop, I'm skeptical. For instance there are many places that are plenty warm to carry more humidity, yet don't. Many places that have high humidity, yet it doesn't go to 100%(where there's plenty of water to do so), nor does the temp spike upwards. Plus such a positive feedback system is unstable, yet the planet hasn't experienced run away warming in it's entire history. I've spend half a life time involved in electronics, and more than a decade as an expert in simulation technology. I'm not convinced.

I don't have a specific answer to this:

If you are to claim that the temperature rise is not caused by CO2, then you need to explain TWO things

1. What stops the CO2/water feedback from causing this temperature rise.
2. What amplifies the suns effect (or some other effect to cause the rise) by >5* but does not do the same to CO2. There isn't a strong if at all significant correlation between Galactic Cosmic Rays and temperature to do this.

But I also don't think much of Holmesian logic(if you eliminate a,b and c, it must be x) as scientific proof.
The cosmic ray count is based on a proxy measurement of (iirc) neutrons, all of this type of evidence is based on a trail of proxies, and convoluted calculations, I'd really like to see some physical experiments proving this(co2/water vapor) link. Also prior to particle accelerators, scientist carted cloud chambers to the tops of mountain to detect cosmic rays as they nucleated water vapor.

Also where are the 11 year sun cycles in your data.

I haven't looked for it yet, but it's a great idea. I started looking at the daily diff in search of seasonal changes as a sanity check.

3 If they can detect 0.1 degree/ decade from averaging annual temps, I should be able to detect it in annual diff. As for detecting 27 millionths of a degree/day, weathers variability is what 10's of thousands of times larger than that(look at the 60 years of daily diff above)?
What I went in search of is falling getting smaller as compared to rising as co2 increased.

4 I'm inclined to trust satellite temps more than land based temps, it's just that we have such a short time period of measurements. As for insulation, it was just a year or so ago that we figured out how large the impact of solar UV output had on the heights of the atm, there's so much we don't know or understand.

As for ways the measurements might be altered, I don't disagree, but if they can be used as proof, where the logic in saying they're to inaccurate to use as disproof.

I also want to disclose the chart in the WSJ post has issues with station selection, I believe it's not a complete set, I've been modifying my sql so I can give it a set of lat/lon values and build an average from that, as opposed to a hard coded lat/log, and it's still not working properly. The charts in my 2 articles don't have that problem, but they did lose the tenth of a degree resolution in the ncdc data, which I'm also fixing. But if you compare the the new chart to the old charts there's little difference in the averages due to the large sample sizes involved.

But at this point I don't think this invalidates my position that co2 at the foreseeable level of the coming century is not going to cause a climate catastrophe. Which might give us time to find a viable power source to replace fossil fuels without crippling modern society first.

—

Never is a long time.

Mi Cro | 02/03/12 | 09:08 AM

Link

2. I've also studied circuits and positive vs runaway positive feedback, I don't understand your scepticism. The way it works to me is that if the temp raises, (by any process, CO2 or extra sunlight etc) then that means the air is capable of holding a little more air, the water warms and evaporates a little more, the equilibrium amount of water in the air rises a little bit. This then traps heat and raises temp by an amount equal to half the original temp raise. This process continues, the result being 1+1/2+1/4+1/8... = 2 i.e. the original temp rise from whatever cause is doubled. The final equilibrium wont have full humidity or a massive temp spike.
There has been quite a bit of work looking into the 1/2 figure which means we can fortunately rule out it becoming >1 and giving us the runaway positive feedback that the alarmists talk about.

3. I am tempted to download all the data now, however I'm worried it will take up a lot of my time!
I have my own ideas, though they probably aren't easily able to do.
To me the best way to isolate the CO2 signal would be to look at the difference between temp at sunset and minimum temperature on a still clear night and see if that changes over decades. A desert would seem to be an ideal place for this. I wonder if there is data. I'm not sure how much difference there would be however as I expect there is only something like a 1% difference in a the amount of heat trapped now compared to 50 years ago. That would make a 20C-10C drop become 20C-10.1C, which may not be detectable.

Have you thought about cloud feedback as a possible reason for the change? Some kind of global cloud feedback is an expected response to rising temps/CO2, and this could be evidence of it kicking in.
Thats why I wouldn't use the maximum temperature, because differences in cloud cover during the afternoon would effect the night time min temp as less/more sunlight would hit the ground after the max temp. If there is a global change, then these daily weather effects would not average out in the long run. If the daily temperature profile changed that would perhaps be evidence for this. Far too many things to try ...

I don't think there is going to be a catastrophe, just that the evidence justifies global action over inaction, and a 3C-6C degree temp rise is worth avoiding if possible. I am quite happy to have some of my money/taxes go towards non-fossil energy research etc.

—

Thor Russell

Thor Russell | 02/03/12 | 17:32 PM

Link

2. In my experience positive feedback always creates instability (as in runaway feedback), unless there's something that counter acts the feedback (in this case say the extra moisture creates extra clouds blocking solar input,ie negative feedback).
Do you have any links to studies on the 1/2 degree figure? I've not seen reference to this before.
It's nice to see you've concluded that the alarmist are over-exaggerating the issue, if only more of them figured that out.

3. That's exactly what i started out to do, based on temperature readings I took while out doing astrophotography, I was interested in seeing if there was a loss of cooling. What I found was that NCDC hourly data wasn't free (and I think it was quite expensive for a complete set). I decided that the summary of days with a min/max would work. But as I was trying to figure out how to filter out just clear days and nights, it wasn't clear on how to do that with the data in the SOD set, and that any results would be labeled cherry picking and would not be accepted. Even these charts are ignored, though I still think they offer insight and proof that co2 isn't the source of of modern temperature increases.

Whatever the various factors are in climate response, they're all integrated into the temperature record (accurate or not), and it's what's being used by alarmists, so IMO it's the best weapon to prove that their interpretation of the data is wrong.

Oh, I don't mind tax incentives for new energy technologies, I do mind that many want to tax oil out of use, and that will be (IMO)a disaster.

—

Never is a long time.

Mi Cro | 02/06/12 | 13:22 PM

Link

Thats a shame about not being able to get the detailed data. I can't find the detailed explanation of the feedback unfortunately, it was somewhere on RC,
however there is this article
http://www.realclimate.org/index.php/archives/2005/04/water-vapour-feedb...

There is a very strong thing to counteract the feedback and stop it going runaway and that is a warmer earth radiates more heat into space just because it a hotter Black Body. I think this is order T^4 or something so it would really take something to overcome this. Also maybe the higher frequencies of IR emitted by a warmer earth are not blocked by CO2/H20 as well. When people say that a doubling of CO2 will cause a 1.2C temp rise, it is already taking this effect into account. So it is for this reason that I think a further positive feedback will definitely increase this 1.2C rise, but not be capable of causing a runaway effect because of the BB radiation of the earth into space. It is possible to still get a runaway effect, the feedback needs to go strongly exponential as presumably happened on Venus, but the research I saw showed that is very unlikely if not impossible on earth.

—

Thor Russell

Thor Russell | 02/06/12 | 16:54 PM

Link

Maybe you can help me with this, I'd like to quantify the overlap in to 10u-20u wave lengths for co2 with the same chart for water
http://webbook.nist.gov/cgi/cbook.cgi?ID=C124389&Units=SI&-SPEC&Index=1#IR-SPEC

http://webbook.nist.gov/cgi/cbook.cgi?ID=C7732185&Units=SI&-SPEC&Index=1#IR-SPEC

I also wonder if how close the planet is to a blackbody, much of the land has vegetation which I can't help but think this isn't close to BB, and wouldn't the oceans emit in the same wave lengths as water does?

—

Never is a long time.

Mi Cro | 02/06/12 | 17:56 PM

Link

Sorry, I havn't gone into it that deeply.I don't think the earth would be a very good BB for incoming sunlight, but for IR at night its probably better. the order T^4 law probably still holds however. I think its BB in the sense that where heat is actually lost to space which is the outer atmosphere, which won't so matter so much whether its ocean or land. Cloud cover probably makes the most difference, hence it being a significant and still not completely understood feedback.

—

Thor Russell

Thor Russell | 02/06/12 | 23:30 PM

Link

The sun's output doesn't vary enough to equal this. The average insolatation is 1366W/m^2 divide by 4 because of geometry. It varies something like 0.1% which is only 0.34 W/m^2.

http://www.sciencecodex.com/solar_cycle_driven_by_more_than_sunspots_sun_also_bombards_earth_with_highspeed_streams_of_wind_0

I'm not trying to say this is the source of the excess heat, only that we still don't have a complete understand of pretty much any of this complex system, and it's far to early to claim we know it's all due to Co2.

Personally, I think there's a lot to consider with the CG of the solar system, and it's influence on the sun as it moves with the planets, which at times(the CG) is actually external to the sun itself.

—

Never is a long time.

Mi Cro | 02/06/12 | 13:19 PM

Link

Comments

Know Science And Want To Write?

What's Happening

Books By Writers Here

Related Articles on Science 2.0

Comments

Know Science And Want To Write?

What's Happening

Books By Writers Here