INTRODUCTION

Welcome to the CO2 Virtual Science Support page.
Here you can access necessary resources throught the process of downloading your data.

WHICH MISSION SHOULD I CHOOSE?

ACOS

GOSAT Satellite

ACOS GOSAT

BIO:

Methane and CO2

ACOS data focuses on the transport mechanisms of Carbon Dioxide (CO2) and Methane (CH4). GOSAT seeks to observe sources and sinks of greenhouse gases, particularly CO2, over time in an effort to make future reductions in greenhouse gas emissions.

AIRS

AQUA Satellite

AIRS AQUA

BIO:

Water Vapor Impacts

AIRS primary scientific impact is improved weather prediction and insight into the Earth's water and energy cycle, including the interrelated processes and the relationship to Earth system changes. AIRS products include several important greenhouse gases (O3, CO, CH4, CO2) in the free troposphere, which facilitate the study of atmospheric circulation processes in the free troposphere.

OCO-2

OCO-2 Satellite

OCO-2 OCO-2

BIO:

High-Resolution CO2

OCO-2 is the only mission focusing solely on the behavior of atmospheric carbon dioxide. Data from OCO-2 will provide a complete picture of human and natural carbon dioxide sources and sinks. OCO-2 strives to map the global geographic distribution of CO2, and study the CO2 changes over time.

TES

AURA Satellite

TES AURA

BIO:

Tropospheric CO2

TES data focuses on the Earth's troposphere. Measurements from the TES instrument will further understanding of long term variations in the quantity, distribution of minor gases and the resulting effects on climate and the biosphere.

TCCON

TCCON Stations

TCCON TCCON

BIO:

Validation on the Ground

TCCON data structures are ground truths. Often times, researchers use TCCON establishments for validation purposes with their satellite models. There are several TCCON sites to choose from globally.

MISSION SPECS

MISSION SPECS
Deployed
Swath
Altitude
Spatial Resolution
Repeat Cycle
Mass, Power
Data Rate
Thermal Operating Range
Physical Size (L,W,H)
Vertical Coverage

ACOS

GOSAT Satellite

AIRS

AQUA Satellite

OCO-2

OCO-2 Satellite

TES

AURA Satellite

TCCON

TCCON Instrument

January 23, 2009 May 4, 2002 July 1, 2014 July 15, 2004 Varies by Location
1000 km 1650 km 10.6 km (nadir) 5.3 km x 8.5 km --
666 km, 98° 705 km, A train 705 km, A train 705 km, A train --
TBD 13.5 km nadir horizontal,
1 km vertical
1.29 km × 2.25 km TBD --
12 days 16 days 16 days 16 days --
1750 kg, 4 kW 156 kg, 256 W 140 kg, 786 W 385 kg, 334 W --
120 mbps 1.44 mbps ~1 mbps for 54 min 4.5 mbps --
TBD 20 - 25°C TBD 0 - 30°C --
2.0 m x 1.8 m x 3.7 m 139.7 x 151.2 x 76.2 cm 2.3 m × 1.4 m 1.0 m X 1.3 m X 1.4 m --
TBD TBD TBD TBD --

DURATION

Currently, CO2 Virtual Science Data Environment can access data products
from four satellite missions. We offer some information on TCCON sites, to
download TCCON data, you must visit their web service.

Duration graphic

DATA LEVELS

DATA PRODUCT LEVELS

Products are processed at various levels ranging from Level 0 to Level 4.
CO2 Virtual Science Data Environment offers Level 2 data products, and the tools to
generate a Level 3 data product.

Level 0

LEVEL 0

Reconstructed, unprocessed instrument and payload data at full resolution, with any and all communications artifacts (e.g., synchronization frames, communications headers, duplicate data) removed.

Level 1A

LEVEL 1A

Reconstructed, unprocessed instrument data at full resolution, time-referenced, and annotated with ancillary information, including radiometric and geometric calibration coefficients and georeferencing parameters (e.g., platform ephemeris) computed and appended but not applied to Level 0 data.

Level 1B

LEVEL 1B

Level 1A data that have been processed to sensor units (not all instruments have Level 1B source data).

Level 02

LEVEL 2

Derived geophysical variables at the same resolution and location as Level 1 source data.

Level 03

LEVEL 3

Variables mapped on uniform space-time grid scales, usually with some completeness and consistency.

Level 04

LEVEL 4

Model output or results from analyses of lower-level data (e.g., variables derived from multiple measurements).

DOES THIS SITE CARRY LEVEL 1A or 1B DATA?

The purpose of this site is to provide Level 2 and Level 3 CO2 satellite data through a "data product as service" model that allows users to define their own data products that match their science needs. Providing Level 1a and Level 1b data products would be redundant to services available elsewhere and lies outside of our mandate.

LINKS TO LEVEL 1A and 1B:

Official Level 1a and Level 1b data for NASA missions are stored and managed by the NASA Distributed Active Archive Centers (DAACs) and can be found there.

DATA VERSIONS

WHAT ARE DATA VERSIONS?

Each mission uses software to process the raw data from the spacecraft into usable products. These algorithms are often upgraded as understanding of the instrument, measurements, and physical system change. The raw data from the spacecraft is then re-run by the new algorithms, generating a new data version.

WHY ARE THERE DATA VERSIONS?

There are different versions of missions’ of how best to process the raw data into a more usable form. Higher version numbers (almost always) represent deeper understanding of the entire measurement system and therefore more accurate products.

WHAT IS THE DIFFERENCE BETWEEN ACOS DATA VERSIONS?

The number of algorithmic changes between ACOS data versions can vary from minor modifications of spectral calibration to dozens of bug-fixes, major re-gridding strategies, and fundamental model re-parameterizations. Access the data user guides to discover the specific changes between versions.

WHEN SHOULD I PICK AN EARLIER VERSION OVER A NEW VERSION?

Short answer: You almost never want an earlier version. Use the latest. It is sometimes desirable to compare before and after a retrieval algorithm change. Hence, multiple data versions are kept for this comparison purpose. This is mostly for those interested in the algorithm development that generates the user data.

Additionally, papers are often published using a particular version of the data. To verify their precise results, previous versions of the data are needed.

Also, if an anomaly is found in a particular data version, it is sometimes possible to go back through previous versions to find if it was present in all data versions or appeared after introduction of a particular model type or code change.

WHEN SHOULD I MAKE THE SWITCH TO A NEW VERSION?

Short answer: Immediately.
This is a personal decision. Newer versions are (almost always) more accurate and therefore superior, encouraging immediate switch. However, the file formats will often be slightly different and the number of observations may vary due to differences in filtration. This will require a data user to potentially modify their data loaders and re-examine previous findings to ensure they are equally present and robust in the new data version.

NASA OFFICIAL VS. NON-OFFICIAL PRODUCTS

WHAT IS A NASA OFFICIAL PRODUCT?

NASA Official products have undergone the certification process at NASA. NASA has certified the Level 2 Full products for all satellite missions on our web service. The ACOS 3.5 Lite product as well as the OCO-2 Lite product are NASA certified products.

NASA Official

When you see the above symbol on our Build page, your product is NASA official.

WHAT IS A NON-NASA OFFICIAL PRODUCT?

Any time a user edits a Full or Lite product, s/he will develop a new data product. This edited product will not undergo NASA certification, thus it will not be a NASA Official product. Edits that influence the validity of NASA official products include subsetting spatially or temporally, subselecting data variables, specifying Warn Levels for OCO-2 and ACOS products, and developing a Level 3 product.

NASA Non-official

When you see the above symbol on our Build page, your product is not NASA official.

HIGH QUALITY POINTS

WHAT IS CONSIDERED HIGH QUALITY?

For some data sets, the concept of a "good/bad" data flag may make sense. However, most data is neither good nor bad but some mixed state in between. In these cases, such as ACOS and OCO-2, the mission team has brought together experts to create a very high quality dataset that is believed to be entirely made up of valid data.

These datasets are very small compared to the entire record (~20%), as only the most trusted of data is admitted. If the user can accept the reduced data quantity, this is an excellent way of immediately selecting a vetted subset of the data. Warn Levels offer an alternative method of data quality estimation that is tunable to any data volume but does require more user interaction than a pre-prepared subset of high-quality data.

WARN LEVELS

WHAT IS A WARN LEVEL?

Users of scientific data from any modern instrument are familiar with the concept of bad data. These are observations where something has gone very poorly, for example unmodelled winds were very strong, a mathematical instability was reached in a retrieval, clouds obscured the entire scene, or a communications error scrambled the bits. These are easily removed, usually with the concept of a Data Quality flag.

However, in planetary and Earth-based missions, the majority of observations cannot be cleanly ascribed to Good or Bad groups. A scene with a very thin, flat cloud (extremely common) slightly perturbs the retrieved results. Is this too affected to be used or not? Some scientists are desperate for data and will take such contaminated scenes. Others want only the most pristine scenes, regardless of the decimation of data that results.

WARN LEVEL BINS:

Warn Levels attempt to address these needs by ordering the data into bins, labeled 0-19. Warn Level 0 data correspond to the most trusted, most pristine observations, while Warn Level 19 represents totally corrupted data useless to nearly anyone. Between 0 and 19 is a gradient of decreasing likelihood of being useful. This takes the place of a Data Quality flag; there is no assignment of good/bad here.

Warn Levels

HOW TO SET YOUR WARN LEVELS

To use Warn Levels, start at Warn Level 0 and observe how much data you will receive and in what places and times. If that is enough for your purpose, stop. You can now be confident your data is quite high quality. If it isn’t enough (likely), slowly increase the Warn Levels you are willing to tolerate. The data will populate in space and time, and the scatter and bias in your data will increase as slowly as possible because you are using Warn Levels.

Set Warn Levels

The warn levels that are roughly defined add 5% of the
data per Warn Level accepted, slightly deviated due
to data distribution characteristics. Examples below:

CREATING CUSTOM QUALITY FILTERING

This act of slowly increasing the Warn Levels until the desired data quantity is obtained is the creation of your own custom filter. Instead of downloading all the data and playing until you make your own filter based on any number of observation qualities, you can simply allow the Warn Levels to sort and filter the data quality for you and "dial in" the right setting for your application.

APPENDIX

Internally, the Warn Levels are based on a small number of observation estimations of known confounding issues. The ratio of co2 estimates determined by different bands, when different from 1, is an excellent example. A large optimization problem was solved to find the most useful, most informative observation characteristics that could predict when the data is likely to be driven from true. Warn Levels must be calculated separately for each data revision, to ensure optimality. ACOS and OCO-2 warn levels, for instance, will be based on entirely different observation characteristics and should not be compared directly except in the "percentage of the data accepted" manner.

GENERATING LEVEL 3 DATA

Generating a Level 3 Product

So, you want a Level 3 product? If the user chooses to apply a Level 3 (L3) algorithm, then the Level 2 data would be aggregated into a customizable latitude-longitude rectangular grid where the latitude, longitude, and temporal bin width are specified by the user.

If the user chooses to download the Level 2 (L2) data, then the data are at the original footprint level. That is, the Level 2 datasets include longitude and latitude information along with the corresponding CO2 values (and other auxiliary and geophysical variables of interest) at the footprints over which the instruments collected data. These data are not pre-gridded to a latitude-longitude rectangular grid.

CITATIONS

ACOS CITATION:

Users of ACOS Data products should cite the following key ACOS references:

Wunch et al (2011a and 2011b), O'Dell et al (2011), Crisp et al (2012), and Frankenberg et al (2014).

AIRS CITATION:

Users of AIRS Data products should cite the following key AIRS references:

Chahine, M. T., et al., "On the Determination of Atmospheric Minor Gases by the Method of Vanishing Partial Derivatives with Application to CO2", Geophys. Res. Lett., 32, L22803, doi:10.1029/2005GL024165

Chahine M. T., et al., (2008), "Satellite remote sounding of mid-tropospheric CO2", Geophys. Res. Lett., 35, L17807, doi:10.1029/2008GL035022

Olsen, E.T., (2015), "AIRS Version 5 Release Tropospheric CO2 Products", http://disc.sci.gsfc.nasa.gov/AIRS/documentation/v5 docs/AIRS V5 Release User Docs/AIRS-V5-Tropospheric-CO2-Products.pdf

Example Citation AIRS:

Bibtex references:
@Techreport{olsen_licata_2014,
AUTHOR={Olsen, E. T. and Licata, S. J.),
TITLE={AIRS Version 5 Release Tropospheric CO2 Products},
INSTITUTION={Jet Propulsion Laboratory, California Institute of Technology},
YEAR={2014}
}
@Article(pagano_2014,
AUTHOR={Pagano, T. S., Olsen, E. T., Nguyen, H., Ruzmaikin, A., Jian, X. and Perkins, L.},
TITLE={Global variability of mid-tropospheric CO2 as measured by the Atmospheric Infrared Sounder},
JOURNAL={Journal of Applied Remote Sensing},
VOLUME={8},
YEAR={2014},
DOI={10.1117/1.JRS.8.084984}
}
@Article(chahine_2005,
AUTHOR={Chahine, M., Barnet, C., Olsen, E.T., Chen, L. and Maddy, E.},
TITLE={On the Determination of Atmospheric Minor Gases by the Method of Vanishing Partial Derivatives with Applicaiton to CO2},
JOURNAL={Geophysical Research Letters},
VOLUME={32},
DOI={10.1029/2005GL024165}
}

OCO-2 CITATION:

Users of OCO-2 Data products should cite the following key OCO-2 references:

O'Dell, C. W., et al. "The ACOS CO 2 retrieval algorithm–Part 1: Description and validation against synthetic observations." Atmospheric Measurement Techniques 5.1 (2012): 99-121

We suggest that users of OCO-2 Level 2 data should include following statement in their publications:

"These data were produced by the OCO-2 project at the Jet Propulsion Laboratory, California Institute of Technology, and obtained from the /OCO-2 data archive maintained at the NASA Goddard Earth Science Data and Information Services Center."

We recommend sending courtesy copies of publications to the OCO-2 Project Scientist. Contact information: oco.jpl.nasa.gov

TES CITATION:

Users of TES Data products should cite the following key TES references:

Kulawik et al. (2013 and 2010).

WARN LEVEL CITATION:

Did you select a custom Warn Level? Users of with a custom Warn Level should cite the following reference:

Mandrake, L., Frankenberg, C., O'Dell, C. W., Osterman, G., Wennberg, P., and Wunch, D.: Semi-autonomous sounding selection for OCO-2, Atmos. Meas. Tech., 6, 2851-2864, doi:10.5194/amt-6-2851-2013, 2013.

Examle Citation Warn Levels:

BibTex reference:
@Article{amt-6-2851-2013,
AUTHOR = {Mandrake, L. and Frankenberg, C. and O'Dell, C. W. and Osterman, G. and Wennberg, P. and Wunch, D.},
TITLE = {Semi-autonomous sounding selection for OCO-2},
JOURNAL = {Atmospheric Measurement Techniques},
VOLUME = {6},
YEAR = {2013},
NUMBER = {10},
PAGES = {2851--2864},
URL = {http://www.atmos-meas-tech.net/6/2851/2013/},
DOI = {10.5194/amt-6-2851-2013}
}