In the media one reads that the first quarter (Jan Feb Mar; JFM) of 2015 was ‘the hottest on record’. This was based on NASA (GISS?) surface T records. Having recently downloaded HADCRUT4 on mojo’s recommendation, I decided to look at JFM there. In that compilation, 2015 was second to (slightly lower than) 2002. This is the first lesson, that while these various compilations are highly concordant, they do differ a bit. Record not awarded. HADCRUT4 also comes with 95% confidence intervals. Second lesson is that to prevail in a statistical sense, more than 0.1 oC is required. This is true of short time intervals. Whether the ‘each decade increase’ (that I have argued) holds in a statistical sense could be worked out. Doubt that would merit a journal publication though, so the impetus is just not there (for me). But I observed another interesting thing. As you know, all these compilations show two T decreases since 1850 on approximately 60 year intervals. It turns out these do not occur during JFM. So I worked out the other quarters, and they are seen most strongly during the Jul Aug Sep (JAS) quarter. I don’t recall reading about that before. Did not have this hypothesis, ‘going in’, but thought you might like to know a third lesson. I can offer a facile mechanism for this; the ocean does its excess-energy sinking when air is warmest (T differential greatest). Of course the Southern Hemisphere has an opposite summer, which calls for some sort of explanation. My top three are that it has many fewer T measuring stations, less land surface area, or that the Southern Ocean has more floating ice. In a recurrence sense we are ‘just about due’ for another of these downturns, and it is confidently expected by some (but not by me). The fourth and final lesson is that if it does come, we should expect to see it first in the JAS quarter. Of this or some later year. 2014 JAS was the warmest (by a margin that just might be significant), so it has not happened yet. You’re just not going to find media headline stories providing this level of tiresome detail. Only here, friends and neighbors.
My understanding is you are looking at what are called 'forcing functions.' These are variable length events that lead to the 'noise' seen in the data: Near as I can tell, identification of forcing functions remains an area of investigation to identify the noise. But in the meanwhile, here is a fun tool: The DENIAL101x temperature tool BTW, the second week covered a lot of technical details about dealing with surface temperature data. Bob Wilson
And thinking about: 2015_rel2: Global Mean Sea Level Time Series (seasonal signals retained) | CU Sea Level Research Group It might be interesting to X-Y plot HADCRUT4 temperature vs sea level. My eyes see the 1998 local peak temperature in both sets of data and 2011 is a great ending year to show the smallest global warming temperature. I'll try Berkeley vs sea level. Bob Wilson
Plot of Berkely Temperatures vs. Sea Level Rise (*) Berkeley temperatures consists of monthly records, 12 per year Sea level records consist of more frequent records, ~36 per year We see three peaks and four valleys. Plotting the local peaks and valleys results in two, parallel, straight-line trend lines. Averages can be misleading because it suppress the peaks and valleys. This means 'cherry picking' the end points in the averages can mislead indicating say 'global warming paused.' But peaks and valleys show the trends in the limits and helps identify the true start and stop points for an average. As an experiment, try the temperature record tool I pointed to earlier and add or subtract single years around 1998 to observe how much the rate of change varies. It looks as if there is periodicity in the 3 peaks and 4 valleys. So I took the shortest and longest peak intervals and projected in the "red" segment when the next peak temperature might show up. Three samples are not enough but they do give a minimum and maximum within this recent interval. The Berkeley data goes back to 1850 which means if the 3 peak and 4 valley forcing function existed before 1997, it should be evident in the earlier data. There are research showing La Nino and El Nino events are world temperature forcing functions and the 3:4 peak-valley pattern needs to be compared to the La Nino and El Nino index. I was curious if the temperature record and sea level changes would show some correspondence and the earlier research indicates aerosols (volcanic) and solar irradiance may yet weak forcing functions in the seal level data. But La Nina and El Nino are sea temperature gradients found at the ocean surface. So warm or cold water may appear BUT the water that was there has gone somewhere else in the ocean basins . . . sea level should be net-sum zero for La Nina and El Nino. BACKGROUND In 1991 I was researching something called 'broadcast storms' that would cause a brief but sometimes fatal amount of traffic to terminal servers and PCs. These short, 30-100 ms duration but back-to-back packet storms were invisible when looking at averaged data. So I had to write code to find the local peaks and suddenly we could track and match failures to these deadly but otherwise invisible events. It is what you have to do with dealing with noisy data. Bob Wilson * - A recent paper suggests the satellite, sea level observations may need adjustment based upon tidal record to handle adjustments for rising and descending land masses. The initial paper suggests the 3.2 mm/year is closer to 2.6 to 2.9 mm/year and the rate increasing due to ice melt water. It is too soon to see if this adjustment will be applied to the U. of Colorado data.
Wjtracy, for me the most interesting thing about atmospheric CO2 is how hard Keeling had to look to find an appropriate place to measure it. Started at a California streamside as I recall. CO2 had an enormous daily cycle, so it’s anybody’s guess what was the ‘right number’. Ended up atop a mountain in mid ocean, without very much plant life around. The volcano does emit CO2, but researchers have been on to that since the beginning. Wind blows the wrong way, CO2 shoots up, data gets set aside. The first airborne CO2 measurements in the 19th century where done very accurately in terms of chemistry. But they were made in cities (where the researchers had their labs), and were very high. Just like what you find today in any city; burning very much outweighs photosynthesis on a local scale. Also cities tend to be built in lowland areas, river valleys, where ‘ventilation’ can be limited. So they also get smoggy. Go out to a cornfield, midday, and if the wind is slight, CO2 will go very low. Corn eats it. Go to a forest at night. Lot of decomposition, photosynthesis=zero, and you get very high concentrations. If the forest is in a valley, a CO2 lake can form. The atmosphere tries its best to even such things out, but the daily local opposing processes are so strong that the atmosphere can’t keep up. So patterns develop, even at regional scales. With enough knowledge of airflow, and math mumbo called inversions, you can figure out large CO2 sources. Pretty much. Where each of you is sitting now, I can't guess the CO2, but it is probably not 400.
Yep, not that 2015 couldn't have been the hottest year since the last interglacial, it just is with measurement and reconstruction uncertainty, we don't know whether it is, but we know it is not statistically hotter than other years. 2005-2014 was warmer than 1995-2004 but not by much, both were significantly (statistically speaking) than 1985-1994, and all decades after the mideval warm period.
Hopefully the 'Great Firewall' gives access to the source of these reports: Global Analysis - March 2015 | State of the Climate | National Centers for Environmental Information (NCEI) Temperature anomalies and percentiles are shown on the gridded maps below. The anomaly map on the left is a product of a merged land surface temperature (Global Historical Climatology Network, GHCN) and sea surface temperature (ERSST.v3b) anomaly analysis developed by Smith et al. (2008). Temperature anomalies for land and ocean are analyzed separately and then merged to form the global analysis. For more information, please visit NCDC's Global Surface Temperature Anomalies page. The percentile map on the right provides additional information by placing the temperature anomaly observed for a specific place and time period into historical perspective, showing how the most current month, season or year compares with the past. . . . In the atmosphere, 500-millibar height pressure anomalies correlate well with temperatures at the Earth's surface. The average position of the upper-level ridges of high pressure and troughs of low pressure—depicted by positive and negative 500-millibar height anomalies on the March 2015 and January–March 2015 maps—is generally reflected by areas of positive and negative temperature anomalies at the surface, respectively. . . . It is not unusual to find different 'metric' systems have different results. But the relative numbers and all qualifications should be checked. The temperature tool referenced earlier only goes back to 1900, a shorter interval than the NOAA range. Berkeley goes back to 1850, a super-set, and is organized by yyyy-month. So it should be fairly easy to see what these other temperature records have for these time windows. Bob Wilson
I believe that the berekely earth provides all the data and open sources algorithms if you want to try some of your own. Berkeley Earth Nobody thinks that temperatures in the last 300 years were as high as the last 50, that is not in the margin for error. Going back over 50 years though we have many fewer temperature readings, different heat island effects, so there can be disagreements about how much hotter we are now. Going back more than 400 years past the little ice age to the Medieval Warm Period(MWP or sometimes MCA(climate anomaly) or MCO (climate optimum) ) and roman warm period there is considerable disagreement on how to use proxies. IMHO we are fairly safe statistically saying that no 50 year period was as warm as the current one going through the MWP but individual years may have been, we just don't have the data.
One thing for sure, the Arctic Ice cover strongly suggests the 2014-2015 winter has been relatively warmer than the previous years: When we were having cold, blizzards, it was the Arctic air that should have been freezing sea ice. Darn those polar vortex. FYI, found new pages about the Arctic: Arctic Change: Ice - Sea Ice U.S. NIC Very timely: The fourth week of Denial 101x addresses the paleo record. Bob Wilson
Remember the earth heats and cools unevenly. The best way to tell the temperature is to measure it, and not use proxies. We only use proxies where we don't have a record, or to calibrate old proxies with the record. Arctic and antarctic ice melt not only have to do with temperature, they have to do with snow fall patterns and ocean circulation. You can use ice melt to talk about how much sun light gets reflected today versus in icier periods, or in context of release of methane calithitrates, but it is beyond worthless as part of the temperature record, if you have thermometers and satellites. That seems as unscientific as pretending warming not levy construction caused katrina damage. Why would a polar vortex increase sea ice? The mechanism for sea ice is more calving than melting, and I don't see why it would cause calving of glaciers. Is the artic warmer than it was 20 years ago, 100 years ago, 300 years ago. Absolutely. We know the first 2 from thermometers, and the third from ice cores. The amount of sea ice, or the thickness of the glaciers don't have much to do with it. Climate Change Impact on Sea Ice Decline | Weather Underground Temperature in the Arctic has increased at twice the rate as the rest of the globe
Calm my friend. The Arctic would certainly qualify as a 'region' and sea ice formation is a pretty good metric about how cold/warm it was. Sure, snow fall contributes to the sea ice once it has formed but cold has got to be a major determinate. This is one trustable proxie. I didn't mention glacier ice loss which is impacted by snow fall. In fact, the Antarctic has seen some increases in the central ice cover because there has been more snow. But central Antarctica snow accumulation has not been enough to over come the coastal and glacier loss of ice mass. This was covered in Denial 101x and they have been pretty good about providing references for everything discussed in the class. If you need them, I'll go back and post here . . . or you could take the 'free' version of the course. I'm fairly calm about Katrina vs any other major storm since at the coast, the sea level continues to rise. When the major storms stop having storm surge flooding, I'll stop singing the song: What do you do with a drunken builder? What do you do with a shore line agent? What do you do with a Dixie Governor? Early in the storm surge. Way, hey, up sea rises, Way, hey, up sea rises, Why, hey, up sea rises, FEMA in the morning. Two different things: Sea ice requires cold, a lot of cold. When I used to surf, I've enjoyed seen snow hit the liquid ocean and quickly disappear while it accumulated on the shore, jetty, roads, and cars. But there is a simple experiment. Put a cup of water in your freezer and come back in 24 hours. Assuming the freezer is working, you'll have ice. Then the snow can accumulate. If the freezer is cold enough, use salt water. Calving of glaciers is a shore phenomina and I'm pretty sure I referenced sea ice which has a very, very small amount of glacier ice because it soon moves to warmer regions and melts. Just a reminder, I was only talking about Arctic sea ice which floats on the ocean, not glaciers which are on the land. These really are two different types of ice with different chemical and physical characteristics. Also, I'm talking about a more recent record, in the 'weather' time frame of years: This chart only covers 1981 to current. Yes, I know there is a claim that weather != climate but there are regions where weather meets climate: polar regions apparently Australia's weather is also driven more by climate change than others due in no small part to being isolated by the oceans. There are credible reports that the early El Nino indicators are going to whack them pretty badly. I look forward to seeing how the Aussis deal with it. Bob Wilson
No, No, No! Bob you are better than this. We have a temperature record and we have proxies. A regional proxies is a poor indicator of global temperature. There is definitely some correlation but even the regional temperature record of the arctic should not used in place of other global temperatures. There are other factors. It is a magicians trick to say don't believe your eyes, look over there. Using arctic temperatures alone to approximate global temperature would give us approximately twice the warming that we think has occurred (source weather underground). Using sea ice we would have vastly overestimated arctic temperatures. Using antarctic sea ice we would have vastly underestimated global temperatures. Don't repeat the magicians tricks. The lady is not sawed in half. If you do this on some proxies you will overestimate, on others underestimate. Why not use the temperature record (from thermometers, and satelites) if you are trying to determine the temperature record. Yes I am overly excited about this, because once you start down the path, it is hard to debunk those that claim antarctic ice extent is larger we aren't warming (extent is large, thickness is smaller, both are proxies, but temperature record trumps all).
Actually Arctic sea ice is a pretty good indicator but if you insist, here is BEST: And: The BEST data file is 400 MB so it may take a while to get it reduced to something we can share. But I'm happy with the satellite records of Arctic sea ice coverage and other inventories. If you're not, OK, we'll get the BEST data soon enough. Bob Wilson
Bob what is a better indicator of changes to global temperature. A) arctic sea ice B) antarctic sea ice C) combination of arctic and antarctic sea ice D) Watching the movie water world E) Watching the movie its a wonderful life F) A set algorithm that uses thermometers from all over the globe. adjusting for heat island effect and checking these algorithms with satellites, or including satellite in the temperature reconstruction. If you answer anything but F, you are ignoring the data in favor of a proxie that can not meaure it as well. D or E might be entertaining for some and a waste of time for others, but like A, B, and C they can't tell you any more about global temperatures than F. A seems to be used most by people that want to predict temperatures are rising faster than they are, which leads to B being used by those claiming that they are rising slower Here are examples of this bad thinking by the most popular climate blog. Good news from #AGU14 ‘Arctic sea ice is holding up to global warming better than expected’ | Watts Up With That? On the Pause in Global Sea Ice Anomalies | Watts Up With That? Now that is convincing that we have paused if you don't actually look at the temperature record. Here is what NOAA says about the possible but not statistically significant record last year. Five things to know about 2014 global temperatures | NOAA Climate.gov I think points 4 and 5 are most pertinant
This technique belongs what are called 'Logical Fallacies', suborder, 'False Dichotomy.' The question and fixed list of choices is nonsense, a poor attempt to frame the issue. So let's start over with a question that can be answered: What is a better indicator of changes to global temperature? In priority: Sea level - this changes as a result of ice melting and ocean warming: It has to be corrected for changes in the tidal observation sites, subsidence and uplift. For example, glacier loss often results in rebounds and the land sinks often from resource extraction: Global ICE inventory - the amount of both land and surface ice which GRACE did and Cryosat measures today. Now you may have noticed I'm more interested in the Arctic regional ice. Living in the Northern hemisphere, there are direct interactions between Arctic and North American, European, Russian, and Asian weather and societies. It is called a 'Polar Vortex' because it comes from the Arctic regions, the North polar axis. The Arctic also has a growing economic activity from increasing summer navigation and mineral exploitation as the sea ice abates. But the climate models generally indicate the polar areas are 'early predictors.' There are other temperature records and over time, they are converging to a warmer climate. But many of them, especially air temperature, appear to be subject to 'forcing functions' that cause short term variability: El Nino and La Nina - ocean exchanges of stored heat and sinks. Aerosols - volcanic but also man-made smoke and soot Solar radiance - the Sun is not a constant source Orbital variance - very long period changes These short-term, variations are used by unethical people to 'cherry pick' start and stop dates to make false claims about global warming: 1998 to 2012 average - Holly C*RP, we're entering a new ice age. 1996 to 2010 average - Uh OH! Here we come Venus, our hair is on fire! That is why with noisy data I plot the local maximums as one slope and the local minimums as the other slope. To the extent they show similar slopes, we have a credible rate of global warming. Bob Wilson
Bob. You put number 1) For measure global temperatures sea level. But you pictured a thermometer. Please although I see you don't understand my point .... Think of what you pictured. Then remember that ice can lag or lead temperature by hundreds of years. What do you want to trust? Why look at the ice for global temperature today when we have thermometers all over the globe, and ice in only a small fraction of the places? When you are trying to dicern ice, measure ice, when its temperature look at temperature. Now we can use ice as a proxy for old temperatures, but here we look at concentrations and icatopes, not simply volume. We have ice left from the last 2 interglacials when carbon dioxide had a lower concentration and sea levels were much higher. No we can not just look at the ice and ignore the satelites and thermometers. The ice tells different answers than temperature.
I see no reason at all to ask somebody to choose among these 'data streams'. Each one is fit for some purpose and not others. No 'trump suit'. I started with HADCRUT since 1850. The others are ...others
NCDC, HADCRUT, GISS, Berekely Earth All seem to mostly agree, but have some margins for measurement and adjustment error. I would not use arctic sea ice, antartic sea ice, artic land ice, antartic land ice, sea level or any combination to devine global temperatures post 1950. Again these proxies are only partially correlated and may lead or lag tempearture for hundreds of years. Trying to use them when thermometers and satelites are available can only lead to worse measurements. Its like having a bunch of government accountants and a bunch of champanzees trying to calculate last years gdp. Sure you may not trust the government accountants, but adding chimps into the mix in no way makes the number more accurate. We get those liking to over estimate how bad things are looking at arctic sea ice. We know from theory and thermometers that arctic temperature should and have increased more than average global temperatures. What does the melting ice tell us that the thermometers don't. It certainly does not tell that polar bears will drown (nice image but no they can swim), we can tell that the polar bears aren't drowning by the stable or increasing population of .... yes measuring the polar bear population. They are getting thinner, and we can postulate that lack of sea ice may make them worse hunters. Yes beautiful, when we measure what we are trying to measure. Now those that want to underestimate how bad things are use the top method but instead look at antartic ice extent. Again not shrinking like predicted but who cares. Those were bad predictions, it doesn't mean the temperatures air above the land and water aren't rising. We have thermometers to say that the sea surface temperature lagged land tempertures but now is catching up. Isn't it great that we don't have to rely on these proxies?
Correct! And here is why: http://onlinelibrary.wiley.com/store/10.1002/2013EF000188/asset/eft215.pdf;jsessionid=A80854CD06FF34C68501282B3BC9911F.f01t01?v=1&t=i9orxewe&s=04ca81128fd9b88d631167a9e3ffa2e0b70fea90 http://www.ipcc.ch/pdf/assessment-report/ar5/wg1/WG1AR5_Chapter13_FINAL.pdf Water - Thermal Properties So you have some source for this wonderful ice that does not melt,". . . ice can lag or lead temperature by hundreds of years. . ." This is Nobel prize material. Where can I get this wonderful water? Your words, stand by them! With a standard microwave I can make super-heated water any day of the week. But super-cooled water, that is a trick that requires special equipment and water not found in nature. Still you assert we can find ice that does not melt above 0C nor freeze below 0C, we need to bottle it and submit our paper . Bob Wilson
