Video follows. I had at first thought that this press release from NASA Goddard was telling me that they had taken the data from their new Orbiting Carbon Observatory and put it into a model that used wind data so that distribution and mixing could be tracked. Seems sensible, right? But no, they’ve created a model that is projecting such things years before the OCO even made it in to orbit, while touting that they have it. The model “simulates May 2005 to June 2007” They write: (bold mine):
But the simulation – the product of a new computer model that is among the highest-resolution ever created – is the first to show in such fine detail how carbon dioxide actually moves through the atmosphere.
Uh, sorry, no. Model simulations aren’t actual movements, you need hard tracking data for that. One wonders what sort of science mindset exists where they can substitute modeled output for actual data and publish a press release like this with a straight face.Hopefully, somebody at NASA Goddard will actually use the OCO data instead of model data to make claims. The high resolution model itself has merit, but without hard atmospheric CO2 data put into it, like we have from the new OCO, it really is just little more than a model with guesswork data.
NASA Computer Model Provides a New Portrait of Carbon Dioxide
An ultra-high-resolution NASA computer model has given scientists a stunning new look at how carbon dioxide in the atmosphere travels around the globe.
Plumes of carbon dioxide in the simulation swirl and shift as winds disperse the greenhouse gas away from its sources. The simulation also illustrates differences in carbon dioxide levels in the northern and southern hemispheres and distinct swings in global carbon dioxide concentrations as the growth cycle of plants and trees changes with the seasons.
Scientists have made ground-based measurements of carbon dioxide for decades and in July NASA launched the Orbiting Carbon Observatory-2 (OCO-2) satellite to make global, space-based carbon observations. But the simulation – the product of a new computer model that is among the highest-resolution ever created – is the first to show in such fine detail how carbon dioxide actually moves through the atmosphere.
“While the presence of carbon dioxide has dramatic global consequences, it’s fascinating to see how local emission sources and weather systems produce gradients of its concentration on a very regional scale,” said Bill Putman, lead scientist on the project from NASA’s Goddard Space Flight Center in Greenbelt, Maryland. “Simulations like this, combined with data from observations, will help improve our understanding of both human emissions of carbon dioxide and natural fluxes across the globe.”
The carbon dioxide visualization was produced by a computer model called GEOS-5, created by scientists at NASA Goddard’s Global Modeling and Assimilation Office. In particular, the visualization is part of a simulation called a “Nature Run.” The Nature Run ingests real data on atmospheric conditions and the emission of greenhouse gases and both natural and man-made particulates. The model is then is left to run on its own and simulate the natural behavior of the Earth’s atmosphere. This Nature Run simulates May 2005 to June 2007.
While Goddard scientists have been tweaking a “beta” version of the Nature Run internally for several years, they are now releasing this updated, improved version to the scientific community for the first time. Scientists are presenting a first look at the Nature Run and the carbon dioxide visualization at the SC14 supercomputing conference this week in New Orleans.
“We’re very excited to share this revolutionary dataset with the modeling and data assimilation community,” Putman said, “and we hope the comprehensiveness of this product and its ground-breaking resolution will provide a platform for research and discovery throughout the Earth science community.”
In the spring of 2014, for the first time in modern history, atmospheric carbon dioxide – the key driver of global warming – exceeded 400 parts per million across most of the northern hemisphere. Prior to the Industrial Revolution, carbon dioxide concentrations were about 270 parts per million. Concentrations of the greenhouse gas in the atmosphere continue to increase, driven primarily by the burning of fossil fuels.
Despite carbon dioxide’s significance, much remains unknown about the pathways it takes from emission source to the atmosphere or carbon reservoirs such as oceans and forests. Combined with satellite observations such as those from NASA’s recently launched OCO-2, computer models will help scientists better understand the processes that drive carbon dioxide concentrations.
The Nature Run also simulates winds, clouds, water vapor and airborne particles such as dust, black carbon, sea salt and emissions from industry and volcanoes.
The resolution of the model is approximately 64 times greater than that of typical global climate models. Most other models used for long-term, high-resolution climate simulations resolve climate variables such as temperatures, pressures, and winds on a horizontal grid consisting of boxes about 50 kilometers (31 miles) wide. The Nature Run resolves these features on a horizontal grid consisting of boxes only 7 kilometers (4.3 miles) wide.
The Nature Run simulation was run on the NASA Center for Climate Simulation’s Discover supercomputer cluster at Goddard Space Flight Center. The simulation produced nearly four petabytes (million billion bytes) of data and required 75 days of dedicated computation to complete.
In addition to providing a striking visual description of the movements of an invisible gas like carbon dioxide, as it is blown by the winds, this kind of high-resolution simulation will help scientists better project future climate. Engineers can also use this model to test new satellite instrument concepts to gauge their usefulness. The model allows engineers to build and operate a “virtual” instrument inside a computer.
Using GEOS-5 in tests known as Observing System Simulation Experiments (OSSE) allows scientists to see how new satellite instruments might aid weather and climate forecasts.
“While researchers working on OSSEs have had to rely on regional models to provide such high-resolution Nature Run simulations in the past, this global simulation now provides a new source of experimentation in a comprehensive global context,” Putman said. “This will provide critical value for the design of Earth-orbiting satellite instruments.”
For detailed views of various parts of the world, visit:
www.nasa.gov/content/goddard/a-closer-look-at-carbon-dioxide
For more information about NASA’s Orbiting Carbon Observatory-2, visit:
Discover more from Watts Up With That?
Subscribe to get the latest posts sent to your email.
‘I had at first thought that this press release from NASA Goddard was telling me that they had taken the data from their new Orbiting Carbon Observatory ‘ yes but that may have not given them the results they ‘need’ were as the models could certainly do that if used in the ‘right way ‘
Gosh, another computer gamE
And then they wonder why nobody takes them seriously…
Looking at the MODEL one must wonder why the Aussies have bothered to spend billions on reducing their emissions, only a muppet would do that at the expense of jobs,or perhaps there is another “fix” in pay oops sorry play.
In this video the color yellow represents 383 PPM and Red 385 PPM it seems like a lot of worrying over 2 PPM which is probably smaller than the margin of error. Plus its all still way under 400 PPM.
Plus, this animation always show the south pacific in green which is 381 PPM which is wrong Mauna Loa is currently at 395 PPM and should be pink, I don’s see any pink in the animation. Most of it is in the 381 PPM to 386 PPM range which is what levels were 10 years ago.
The simulation is for the year 2006
Elmer, seems reasonable for results from 2006.
As Rosana Dana used to say “nevermind” but my point about there only being 2 parts per million between yellow and red still stands. Plus this shows it all collecting at the north pole which I’m pretty sure isn’t happening.
Does anyone know how many actual atmospheric CO2 measurements are in the database? The only measurement I have seen quoted is the one from the Keck Observatory in Hawaii.
What struck me most was how much more red was the last frame (Dec 28) at the end of the video than was Jan 3 (at the beginning). I don’t think they have a handle on the physics and chemistry sufficiently to wrap the end of year to the beginning of year
Why do you expect that the pCO2 would be the same in December 2006 as it was in January 2006 when we know that during that year globally it increased by about 2ppm?
Their CO2 model is total BS. The Japanese showed a couple of years ago that Europe and N America are net carbon dioxide sinks and that S. America and Africa are the main CO2 emitters. Clearly, they have a total bias against industry and totally ignore all the dung and wood being burned in the undeveloped regions.
” all the dung and wood being burned in the undeveloped regions.” => “Their CO2 model is total BS”
So you agree (ignoring the burned wood) that the model does have some skill after all. 🙂
If the model is to be believed, the fundamental assumption that CO2 is “well-mixed” in the atmosphere is false. Who knew?
My understanding of this is that CO2 is indeed “well mixed” in the atmosphere, but not perfectly mixed. So there are weak gradients created even by 1 or 2 percentage point differences. The sources and sinks are reconstructed by advection over these weak gradients.
The proof of all this will be in the OCO-2 pudding. When and if they decide to make it public.
I have to agree that NASA gives the impression that they are showing measured OCO-2 satellite data while they display stuff from elsewhere: https://co2.jpl.nasa.gov/#mission=OCO-2
A 172 MB global OCO-2 file for the period 8/9/2010 – 8/24/2010 while the launch was in July 2014!
Back to the Future is finally reality: today’s science has no limits.
I browsed the “co2.jpl.nasa.gov” link, trying to find some OCO-2 data to download. I was able to get some sample “prelaunch” data from 2010, in hdf5 format. Using HDFView it appears to be the 194 raw data variables in a matrix format. Probably, as the name implies, some dummy data generated pre-launch to allow researchers to write their analysis codes while waiting for the real thing. The samples were less than a 1 megabyte.
https://co2.jpl.nasa.gov//build/#dataset=OCO2PreLaunch&product=FULL
I also tried to download this ‘FULL’ product, also labeled ‘prelaunch’, but apparently a larger, 133 megabyte set. The web site generates a Python script for you which contains a list of 5 hdf5 files to download.
But when I ran the script, it got a code:404 (file not found) on each file, but the script ignored the errors and tried to open the error report, thinking it was an hdf5, generating floating point divide-by-zero errors etc.
So we can assume that the real OCO-2 data is not yet available. Why don’t they just say that, and stop pretending the prelaunch samples are the real thing.
Most disturbing, as Anthony Watts points out, is this paragraph by NASA about its model: “But the simulation–the product of a new computer model that is among the highest-resolution ever created–is the first to show in such fine detail how carbon dioxide actually moves through the atmosphere.” “Actually”??? Not quite. Watts rightly says, “Uh, sorry, no. Model simulations aren’t actual movements, you need hard tracking data for that. One wonders what sort of science mindset exists where they can substitute modeled output for actual data and publish a press release like this with a straight face.”
But what we’re seeing here–the failure (inability created by habit?) to distinguish model from real world–is common to much of the AGW (and especially CAGW) scientific community, as Myanna Lahssen documented through years of observation and interviews with modelers at NCAR. See her article “Seductive Simulations? Uncertainty Distribution around Climate Models” (http://sciencepolicy.colorado.edu/admin/publication_files/resource-1891-2005.49.pdf). She found that NCAR scientists consistently and persistently spoke of their models as if they were the real world; of their models’ output as if they were real-world observations; and genuinely couldn’t keep the distinction clear in their minds.
(Quote) “One wonders what sort of science mindset exists where they can substitute modeled output for actual data and publish a press release like this with a straight face.”
Anthony, maybe this is an extension of the simulations they have used in spaceflight coverage (in lieu of actual footage). The general public will have come to perceive this as reality over time.
Reblogged this on Centinel2012 and commented:
I bet DreamWorks could do better then these guys!
If anyone is still interested in this video, I have made a re-rendering of it (http://youtu.be/D_EjLowB-7M) with the flat map re-mapped onto a globe. (actually a pair of globes with North and South poles visible).
If nothing else it is pretty computer art. Also it was a exercise in using Blender, so I got at least that much out of it.
{Thank you, good job. .mod]