Finally: visualized OCO2 satellite data showing global carbon dioxide concentrations

From the “if the government won’t visualize it, a climate skeptic will” department.

Guest essay by Erik Swenson

In July of 2014, NASA launched its most advanced carbon dioxide monitoring satellite, The Orbiting Carbon Observatory (OCO-2). The first OCO burned up on launch. There has been a lot of anticipation regarding the data from this instrument. However, over a year after it launch, there has been little public information presented about its results. The only data made available by NASA has been images showing CO2 from an AGU14 session.

These images are shown below.

clip_image002

Figure 1: NASA-provided OCO-2 data for Oct 1 – Nov 11, 2014

http://www.nasa.gov/sites/default/files/thumbnails/image/mainco2mappia18934.jpg

clip_image004

Figure 2 NASA-provided OCO-2 data for Nov 21 – Dec 27, 2014

Source: http://svs.gsfc.nasa.gov/vis/a010000/a011700/a011788/F5a.png

Back in May 2015, there was a release of some visualized data showing mixing ratios of CO2 over the oceans:

JpGU_OCO2_glint_data

For some reason, NASA has not chosen to publish any recent updates of the OCO-2 satellite data. Many people are interested in the data from OCO-2, but have not been able to access the information. NASA has now provided access to the raw data from OCO-2, but the data is in the HDF file format. No common commercial programs such as Excel can access this data in this form.

I have created a program to parse this data, and attempt to graph it in a form that closely matches the output of the NASA images. The data is available from 9/20/2014 – 9/22/2015 as of this writing. I have generated the plots in approximately 6 week intervals. It takes about that much data to cover most of the globe with observational data. You can see how the orbit path is from this NASA visualization story:

OCO2-orbits
From NASA: “It takes 16 days and 233 orbits for the satellite to produce a complete global picture of carbon dioxide.”

A few implementation notes.

The data from each sample is put into an array. Each point is added to the array as a circular blob. The center point of the circle has a weight of 1 for the averaging function. The remaining points in the circle are weighted in a decreasing manner from the center. This choice is based on the images from NASA which show circular artifacts.

All of the images use the same min/max scale of 380 – 415 ppm. This does not give the best dynamic range for each image, but it does present a good range over all of the images.

The NASA images are chopped beyond 60 degrees N and S latitude. I have chosen to show whatever data is there.

All data points are plotted from the OCO-2-Lite files regardless of warn_level. Warn_level is used to judge the quality of the sample. The OCO-2-Lite files say they are the “high-quality” samples, so I chose to use them all.

The data used for these images is from the OCO-2-Lite v7 data set. It can be accessed here:

https://co2.jpl.nasa.gov/#mission=OCO-2

Finished visualizations

The data here is presented without comment. I will leave it to others to decide what this data means. So, without further ado – here is the data I have processed.

clip_image006

Figure 3: Processed data from Oct 1 – Nov 11, 2014

Figure 3 is an attempt to match the first NASA image from Oct 1 – Nov 11, 2014 to see how closely my algorithm matches. Note that NASA has adjusted the data set multiple times since the release of the NASA image. The current version is v7. I am not sure what changes have been made in the data.

clip_image008

Figure 4 : Processed data from Nov 16 – Dec 31, 2014

clip_image010

Figure 5 : Processed data from Jan 1 – Feb 15, 2015

clip_image012

Figure 6 : Processed data from Feb 16 – Mar 31, 2015

clip_image014

Figure 7 : Processed data from Apr 1 – May 15, 2015

clip_image016

Figure 8 : Processed data from May 16 – Jun 30, 2015

clip_image018

Figure 9 : Processed data from Jul 1 – Aug 15, 2015

clip_image020

Figure 10 : Processed data from Aug 16- Sep 12, 2015

UPDATE: Eric Swenson provides this map in comments showing CO2 over the entire year from From September 2014 to October 2015 – Anthony

OCO2-1year-co2-globalmap

Also, reader “edimbukvarevic” provides this map of anthropogenic CO2 emissions for comparison:

co2_map_global_anthro-emissions

5 1 vote
Article Rating
464 Comments
Oldest
Newest Most Voted