bdgwx

This paper is not relevant to the error for the annual OHC anomaly or annual global temperature. Nor is the quoted 0.6C figure the measurement error that can be expected from an ARGO float which is said to be approximately ±0.002C for individual measurements. The 0.6C figure is the RMS error of ARGO derived hydrographic section temperature fields. These sections are computed even in lieu of being occupied by an ARGO float at each grid cell. Using the WOA (World Ocean Atlas) dimensions we can estimated 75x30 = 2250 grid cells along the cruise line of the hydrographic section used in the paper. If you were to then answer the question...what is the error of computed mean temperature of this hydrographic section then you might expect it to be 0.6/sqrt(2250) = ±0.01C using the standard error of the mean formula. In reality I suspect the actual error to be a bit different for a variety of reasons. I'm just giving you an order of magnitude estimate based on trivial statistical principals using the RMS error of the temperature field on that single hydrographic section mentioned in the publication. Note that this hydrographic section represents but an infinitesimally small part of a much larger 3D volume containing vastly many more grid cells by which to significantly reduce the error in the global mean temperature estimate if such a method were used. I do not see anything in this publication that is inconsistent with Cheng's OHC 2σ envelope. I do see that ARGO reduces the error by a factor 2 relative to the non-ARGO era. Perhaps this is why the 2σ envelope appears to be significantly reduced in later years on Cheng's graph?

Fun with math. The ocean has a mass of about 1.4e21 kg. The specific heat capacity is about 4 kj/kg. This means it would require 5.6e24 joules of energy to increase the mean temperature of the ocean by 1.0C. This would require an EEI of +1.2 W/m^2 to persist for 300 years. A smaller +0.7 W/m^2 imbalance is causing the GMST to increase by about +0.2C/decade. I doubt the relationship would be linear but you can certainly do an order of magnitude estimate on what might happen if we added 5.6 yottajoules of energy to geosphere over a 300 year period. Hint...hothouse Earth might be something worth researching. It's a good thing then that no reputable scientific work claims that climate forcing agents are known with perfect certainty nor is it claimed that anthroprogenically modulated forcing agents are the only contributors to changes in temperature. I highly recommend reading IPCC AR5 WGI chapter 8 on radiative forcings for a brief summary of the agents in play and estimates of their magnitudes and uncertainty. ENSO, clouds, and likely a bunch of stuff that you haven't even thought of are all actively researched. I think what you'll find when looking at the academic literature is that the exact opposite of arrogance is happening in climate science.

You can see in the graph that @chubbs posted that the 2σ (95%) confidence envelope is delineated by the green error bars.

Heat is the transfer of thermal energy. For the body being warmed/cooled it requires a net positive/negative flow of energy. If I am to interpret your question in precise terms then it is equivalent to asking...of the Earth Energy Imbalance (EEI) how much does the Sun contribute? The answer is effectively nothing. The reason is because total solar irradiance is not increasing. In fact, if anything it has actually been decreasing, albeit by a small amount, over the last few decades. I'm going to estimate the RF of the Sun over this period at about -0.01 W/m^2 as compared to the EEI of +0.70 W/m^2. In order of magnitude terms you might say it is 1/100th and in the opposite direction. Now if the question were...how much energy is being added by the Sun then the answer is about 240 W/m^2. This is the effective solar ingress flux near the surface. It is about 300x the EEI. Don't forget about the near surface egress fluxes though!

There are many potential sources of error including instrument bias, station moves, station sighting changes, time-of-day of observation, urban heat island effect, human element, etc. Ocean temperature datasets like ERSST and HadSST go back well into the 1800's as well. They are used as inputs into the global mean surface temperature (GMST) datasets as well. We have reconstructions of the ENSO cycle going back hundreds of years. The ENSO phase was positive (El Nino) in 1885 and into 1886. Speaking of cities and the potential for the urban heat island effect...Berkeley Earth concluded that the UHI bias was flat to even negative since 1950 during the period in which the anthroprogenic influence is most acute. Berkeley Earth shows that the GMST increased about 1.0C since 1880. Temperature proxies like those derived from ice cores, tree ring analysis, etc. can provide insights into the GMST going back millions of years.

Certainly not as accurate as today. But here's the cool thing about math. Trivially if a thermometer has an error of say 2C and there are 1000 thermometers then the error of the mean is only 2/sqrt(1000) = 0.06C. In reality it's far more complicated than that. After it is all said done surface datasets like GISTEMP publishes an error of about 0.10C before WWII and 0.05C after for the global mean temperature. The error envelope expands the further back in the 1800's you go. Berkeley Earth actually has some in depth papers on how it works and how the uncertainty is determined. Fair warning...it's thick stuff. http://berkeleyearth.org/static/papers/Methods-GIGS-1-103.pdf http://berkeleyearth.org/static/pdf/methods-paper-supplement.pdf

Using my +0.7 W/m^2 figure above and dividing by 240 W/m^2 yields = 0.7 / 240 = 0.3% of the surface budget. However, keep in mind that +0.7 W/m^2 is just the imbalance that still needs to equilibriate. The energy imbalance from the past that has caused 1.1C of warming has already equilibriated so is not included in my +0.7 W/m^2 value. This additional energy is probably in the 1.5-2.0 W/m^2 'ish range (just guessing right now). If you include that then we're probably close to 1% of the surface budget.

So to put that in perspective the change from 1985 to 2019 is about 350e21 joules (eyeballing for now). The average uptake is thus... 350e21 (OHC-joules) / (31.56e6 (seconds in year) * 34 (years)) / 510e12 (Earth area-m^2) = 0.64 W/m^2. The ocean takes up about 90% of the imbalance so we can probably estimate the average imbalance as 0.64 / 0.9 = +0.7 W/m^2. That is a pretty large imbalance and is inline with expectations of the net radiative force from all agents subtracted off from what has already equilibriated to raise the global mean temperature. Over the last 10 years this imbalance actually works out to about +1.0 W/m^2. The implication...even if GHG emissions were to cease instantly there is still a lot of warming waiting in the pipeline that needs to equilibriate.

Here's a pretty good (and lengthy) article talking about Arctic sea ice and the odds of seeing an ice-free summer. https://interactive.carbonbrief.org/when-will-the-arctic-see-its-first-ice-free-summer/ Bottom line...around 2050 is the 50/50 point especially if carbon emissions take a more middle-of-the-road trajectory being neither abated aggressively or allowed to grow unmitigated...basecially RCP 4.5.

I'm just wondering if there isn't going to be more warm sector activity out ahead of this line than is being advertised at the moment. The area I'm most interested in is eastern TX and LA as the LLJ ramps up in the overnight hours. Like @CheeselandSkies I'm certainly not sold on that, but I'm not going to eliminate that possibility just yet either.

There are some heavy hitters showing up on CIPS including 2/5/2008.

It's interesting what is happening in the SH as well. According to the IPCC the expectation was for mostly flat trends and possibly even an increase through the 2020's. So to see the SH decline and even drop to record lows tells us that at least some sea ice predictions have underestimated the decline down there too. It seems as though there is a long history of sea ice predictions being too conservative; sometimes shockingly so. For example, back in 2001 and inferring from a graphic in AR3 the IPCC predicted that NH sea ice extent wouldn't drop below an annual mean of 10.5e6 until about 2040. It first happened in 2007 and then 6 times after that including the last 4 years in a row. So by taking a more conservative stance and hinting that the declines may moderate in the 2020's I'm doing so fully aware that I could end up getting burned. But I also understand that trendline reversion is a powerful concept and I'm also trying to stay pragmatic and not come across as overly alarmist either.

So if I remember correctly some of the recent computer modeling studies showed that the 2020's might be characterized as period of stalling out on the declines before picking back up again in the 2030's. What do you guys think? Are we going to see the same dramatic declines or will there be a hiatus? I think I'm more in favor of a moderation in the decline rates. But, it's not lost on me that those who have made similar conservative predictions in the past have gotten burned. So I'm prepared to be wrong.

According to NSIDC... For 2019 the NH (Arctic) ended with an annual mean of 10.186e6 km^2 of extent. This is the 2nd lowest after 2016 which ended with 10.163e6. For 2019 the SH (Antarctic) ended with an annual mean of 10.826e6 km^2 of extent. This is the 2nd lowest after 2017 which ended with 10.749e6.

Annual global mean temperatures from most datasets are accompanied with a margin of error. This makes annual rankings probabilistic. Berkeley Earth has a good visualization of this in their 2018 report. You can see how the temperature distribution curve for each year peaks at the reported value and how the tails can overlap with other years. http://berkeleyearth.org/2018-temperatures/ They have an excellent paper describing how the averaging process works and how uncertainties are dealt with and reported. http://berkeleyearth.org/static/papers/Methods-GIGS-1-103.pdf Most other datasets post their uncertainties and make annual rankings in a similar manner.

According to the NSIDC the 5 day average is well outside the interdecile range and just barely inside the 2σ envelope. It is 870,000 sq km below the 1981-2010 mean. Also, I counted 6 other times in which sea ice extent increase from 12/1 to 12/23 was higher.

And with reanalysis running warmer through the first half of December I think it's a near certainty that 2019 is going to clinch 2nd place in most datasets.

It's somewhat of an interesting twist isn't it?

One model that gets a lot of attention is that done by Hansen and published in 1988. See H88 here. He models three different scenarios: A, B, and C. A is a high emissions scenario, B is medium emission scenario, and C is a low emission scenario. Scenario A assumes exponential growth of GHG emissions plus hypothetical new emissions. B assumes modest GHG emissions most similar to what has been experienced so far. And C assumes drastic curtailment of all GHG emissions by 2000. Scenarios B and C inject volcanic eruptions in the years 1995 and 2015 similar in magnitude to El Chichon and Agung. Scenario A has no such volcanic eruptions injected. Scenario A is claimed to produce an equivalent forcing of 2xCO2 by 2030, B by 2060, and never for C. Scenario B was said to be most realistic and likely future trajectory. The actual trajectory of forcing lies between B and C. The HDAS2019 publication referenced in this thread estimates B's forcing as 27% higher than what actually transpired. Today we know this is partially due to the Montreal Protocol and probably the somewhat higher volcanic aerosol forcing that occurred due to Pinatubo 1991 and several VEI 4+ eruptions in the early 2000s as well. Had H88 used correct inputs in a hypothetical scenario that closely matched reality then that scenario would have exhibited very high skill in predicting the global mean surface temperature. This tells us the problem was more with the assumed inputs than with the model physics.

A new study appeared yesterday evaluating the skill of past climate model simulations. https://www.sciencemag.org/news/2019/12/even-50-year-old-climate-models-correctly-predicted-global-warming The news article contains a link for free access to the publication. Here is the paywall link though. https://agupubs.onlinelibrary.wiley.com/doi/abs/10.1029/2019GL085378 And here is the materials available on github. https://github.com/hausfath/OldModels Abstract Retrospectively comparing future model projections to observations provides a robust and independent test of model skill. Here we analyse the performance of climate models published between 1970 and 2007 in projecting future global mean surface temperature (GMST) changes. Models are compared to observations based on both the change in GMST over time and the change in GMST over the change in external forcing. The latter approach accounts for mismatches in model forcings, a potential source of error in model projections independent of the accuracy of model physics. We find that climate models published over the past five decades were skillful in predicting subsequent GMST changes, with most models examined showing warming consistent with observations, particularly when mismatches between model‐projected and observationally‐estimated forcings were taken into account. Plain Language Summary Climate models provide an important way to understand future changes in the Earth's climate. In this paper we undertake a thorough evaluation of the performance of various climate models published between the early 1970s and the late 2000s. Specifically, we look at how well models project global warming in the years after they were published by comparing them to observed temperature changes. Model projections rely on two things to accurately match observations: accurate modeling of climate physics, and accurate assumptions around future emissions of CO2 and other factors affecting the climate. The best physics‐based model will still be inaccurate if it is driven by future changes in emissions that differ from reality. To account for this, we look at how the relationship between temperature and atmospheric CO2 (and other climate drivers) differs between models and observations. We find that climate models published over the past five decades were generally quite accurate in predicting global warming in the years after publication, particularly when accounting for differences between modeled and actual changes in atmospheric CO2 and other climate drivers. This research should help resolve public confusion around the performance of past climate modeling efforts, and increases our confidence that models are accurately projecting global warming.

From the early arrivals...CFSR and UAH for November came in 0.10 and 0.09 higher than October respectively. I would imagine the conventional surface station datasets like GISTEMP and Berkeley Earth and the like will be at least as warm as October if not warmer. This would position 2019 with very high odds of being the 2nd warmest on record.

That's pretty typical. Regional trends often do not align with global trends. Another example...Siberia cooled even more than the upper great plains in the CONUS during this same period. Yet...the planet is warmer overall now than in the previous decade. I asked the question above...is this a result of a systematic shift via WACKy, quasi resonant amplification of the polar jet, or something else? Or is it just another chaotic artifact that will disappear in the following decade? Lots of questions...

NSIDC isn't the source though. Their "NSIDC in the News" section is just links to various articles and blogs that mention NSIDC. They have hundreds of links in this section every year. It's not even clear if these articles (which are dead links now) were in reference to a bona-fide peer reviewed study or some random blogger's opinion. Note the disclaimer in the section. The following items link to media coverage of NSIDC in various news outlets, online magazines, editorial pieces, and blogs. The content of these articles and blog posts does not necessarily reflect the views of NSIDC, our collaborators, or our funding agencies. Like I said, I can't see the articles anymore so I have no idea what the details of these "predictions" are. But based on the timing of when the articles appear I can speculate a bit. There were two fellows during this period that made some very aggressive predictions that got widespread media attention. The first was Maslowski and the second was Wadhams. Neither was characterized by broad acceptance in the academic community. In Wadham's case he was pretty much entirely ignored. Maslowski was a legit researcher but in his defense his work was frequently taken out of context. His 2016±3 date (which was often erroneously cited as 2013) was statistical and appeared in a publication that I believe used many methods to arrive at many different estimates with 2016 being the lowest therefore making a cherry pick and really bad at that. Masklowski even warned against taking his work out of context and specifically chided Al Gore for doing just that. The point...be careful about linking media popularity with the mainstream views of bona-fide scientists. They are often at odds with each other.

Broadly speaking the first "ice-free" year has been getting pushed up. You'll find select studies here and there that have really aggressive predictions, but those are either few in number or not well received enough to influence the consensus much. In the 1990's the prevailing prediction was around 2100 or thereafter. And in the IPCC's AR3 report from 2001 it was stated (via a chart) that the first annual mean extent of 10.5e6 km^2 would not occur until about 2040. In reality it actually occurred in 2007. Even today many sea-ice models continue to struggle with the rapid pace of sea ice declines in both the NH and SH. Today it seems as though the consensus lands somewhere in the 2040-2060 range. So we still have a good wait ahead of us before we see < 1e6 km^2 of extent at the minimum. It's certainly possible that it could occur prior to 2040. Some on this forum and the ASIF believe we'll be lucky to make it to 2040. I'm in the more conservative camp and believe it will be after 2040. I'm prepared to be proven wrong though.

Here is the average ONI for each of the last 6 years. Note that because the global mean temperature response tends to lag ENSO by a few months I have computed each year's average ONI using an offset. It seems as though this lag is typically in the range of 3 to 6 months so I've included two values. The first is the 3-month lag and the second is the 6-month lag. For example the 6-month lag value of +1.7 in 2016 is computed from 2015-07 to 2016-06. 2019 = +0.6, +0.6 2018 = -0.4, -0.5 2017 = -0.1, -0.2 2016 = +1.2, +1.7 2015 = +1.0, +0.6 2014 = -0.1, -0.2