Debunked: Carbon Monoxide "Explosion" on US West Coast Feb 26th? [Terra Satellite Glitch]

Mick West

Administrator
Staff member
20160301-112912-ffh25.jpg

The Carbon Monoxide overlay on earth.nullschoool.net appears to show a sudden spike in CO levels over most of California on 2/26/2016. There's a large area shown at off-the-scale levels.

The previous day looked like:
20160301-113203-vebgd.jpg

i.e. basically none. So what's going on here? Was there a huge spike in CO? I'm right in the middle of that black area, I did not notice anything on Feb 25th (last Thursday).

It turns out it was just a glitch in the data. Here's NASA's statement:

http://gmao.gsfc.nasa.gov/news/geos_system_news/2016/incorrect_CO_concentrations.php

March 1, 2016


IMPORTANT NOTICE: Elevated carbon monoxide concentrations over California in the GEOS-5 products since February 25, 2016, are incorrect. They are a consequence of unrealistic CO emissions computed by our biomass-burning algorithm, which is based on satellite observations of fires. Several data points over California were incorrect and these led to unrealistically high emissions of CO, causing the elevated atmospheric CO concentrations.

GMAO is working to correct this problem.

Content from External Source
The important thing to realize here is that the earth.nullschool.net display is not actually showing real measurements of carbon monoxide. It's a computer model, basically an estimate or forecast of what carbon monoxide is likely to be, give a variety of input data. In this case there was incorrect input data about fires in California, and this led the system to predict high CO levels, which were then later rendered as a graphical CO forecast.

The glitch happened on Feb 26th. The other thing that happened around then is on Feb 25th when the Terra satellite came back on line after being in "safe mode" for a few weeks. Terra is responsible for the "Fires and Thermal Anomalies" part of the equation. Here's the Feb 24 input:
20160301-115136-do90s.jpg
(i.e. no data at all)

And here's Feb 25th:
20160301-115240-m9mwk.jpg

So basically the input data went from zero fires worldwide, to lots of fires. The forecasting algorithm was unable to handle this, and it resulted in the odd spikes over California and other areas in the next day's CO forecast. As new data came in, things settled down to normal.

Of course the actual "ground truth" of this was the levels of CO on the ground on 2/26. The faulty data shows levels of over 30,000 ppb in the California Central Valley:

30,000 ppb (parts per billion) is 30 parts per million. (30 ppm)

But actual ground level monitoring stations showed 1 ppm or lower. And there was no increase from the previous day.
http://www.valleyair.org/aqinfo/d-CO.htm (Archived: http://archive.is/nUAvw)


There are multiple reading here: http://www.arb.ca.gov/aqmis2/aqdselect.php

In Sacramento, near where I live, the story is the same, normal levels, of under 1 ppm (http://archive.is/LqPqo ) with no spike on 2/26

 
Last edited:
Hey Mick,

Long time reader, first time poster. To start, thanks for all your work, from Contrailscience to here, it's been a fascinating, and inspiring lesson in thoroughly researched debunking. You have a great community too.

But enough brown-nosing :) ...

The carbon monoxide meme in the OP has been all over my Facebook feed, where I'm a member of a couple geoengineering/skywatch type groups (you never know what hijinks they'll be up to next). They also posted dismissive responses to NASA's official explanation, for example from Paul Beckwith ("a pretty mainstream scientist, University of Ottowa"), here:
http://paulbeckwith.net/

Beckwith says "NASA’s explanation below makes no sense whatsoever", and offers these five points:

1) They only mention CO, and say nothing about CO2 and SO2.
2) The emissions are spatially located over the fault-lines; how could fires aggregate in these regions?
3) There was an Earthquake in New Zealand around Feb 29th that showed CO release.
4) They are trying to say the error was just over California. Really? Was the data wrong over the whole globe?
5) I discussed the idea of “instrumentation error” in my second video yesterday. In my opinion, the explanation from NASA does not address any of the points in my video, which I stand by 100%.
Content from External Source
(Beckwith links to the referenced Youtube videos on his blog.)
 
Last edited by a moderator:
Hey Mick,

Long time reader, first time poster. To start, thanks for all your work, from Contrailscience to here, it's been a fascinating, and inspiring lesson in thoroughly researched debunking. You have a great community too.

But enough brown-nosing :) ...

The carbon monoxide meme in the OP has been all over my Facebook feed, where I'm a member of a couple geoengineering/skywatch type groups (you never know what hijinks they'll be up to next). They also posted dismissive responses to NASA's official explanation, for example from Paul Beckwith ("a pretty mainstream scientist, University of Ottowa"), here:
http://paulbeckwith.net/

Beckwith says "NASA’s explanation below makes no sense whatsoever", and offers these five points:

1) They only mention CO, and say nothing about CO2 and SO2.
2) The emissions are spatially located over the fault-lines; how could fires aggregate in these regions?
3) There was an Earthquake in New Zealand around Feb 29th that showed CO release.
4) They are trying to say the error was just over California. Really? Was the data wrong over the whole globe?
5) I discussed the idea of “instrumentation error” in my second video yesterday. In my opinion, the explanation from NASA does not address any of the points in my video, which I stand by 100%.
Content from External Source
(Beckwith links to the referenced Youtube videos on his blog.)


NASA has also removed the statement they made from their website!
 
Beckwith says "NASA’s explanation below makes no sense whatsoever", and offers these five points:

2) The emissions are spatially located over the fault-lines; how could fires aggregate in these regions?
Content from External Source
It seems Beckwith doesnt understand the explanation. The reason given was the there were fires detected that were not there and then the computer model predicted CO levels based on this erroneous info. Seems pretty straight forward explanation. Unless I am missing something which is entirely possible.

They flipped open to doors on satellite - turned on the computer- and all this info flooded in and caused erroneous results...that then smoothed out as more data came in. At least thats I how I read it.
 
They have a setting for Carbon Monoxide. Sorry, the link didn't go to that. You can also change the date. http://www.baaqmd.gov/about-air-qua...aView=tech&StartDate=2/26/2016&ParameterId=13

These readings on this page are presented in parts per ten million (pptm) rather than parts per million (ppm) on the OP links. I don't see any spikes @mawrite (unless I am using my "boy-eyes"?), could you clarify which part looks out of the ordinary?

Either way:
There are two federal standards for carbon monoxide: 90 parts per ten million averaged over 8 hours, and 350 pptm averaged over one hour.
Content from External Source
And none of the readings violate (or even come close) to those standards.
 
It was also found on this site which uses different satellites. http://macc.copernicus-atmosphere.eu/ Hopefully it gets debunked. Bye.

The MACC project, which is a simulation similar to GEOS-5, actually gets Fire Radiative Power (FRP) data from the MODIS instrument on the Aqua and Terra satellites. It was bad FRP data from Terra that led to the bogus CO calculations in GEOS-5 so it should be no surprise that MACC would produce similar bogus results from the same bogus data. Ground monitors did not detect excess CO beyond what would be expected under the meteorological conditions at the time.

http://macc.copernicus-atmosphere.eu/about/project/macc_input_data/
 
The MACC project, which is a simulation similar to GEOS-5, actually gets Fire Radiative Power (FRP) data from the MODIS instrument on the Aqua and Terra satellites. It was bad FRP data from Terra that led to the bogus CO calculations in GEOS-5 so it should be no surprise that MACC would produce similar bogus results from the same bogus data. Ground monitors did not detect excess CO beyond what would be expected under the meteorological conditions at the time.

http://macc.copernicus-atmosphere.eu/about/project/macc_input_data/

GIGO in action:)
 
Back
Top