Who is Quick Carbon?

Quick Carbon is a growing academic research initiative and a protocol for rapidly assessing soil carbon stocks across a landscape. Quick Carbon has been housed within Yale School of Forestry & Environmental Studies (F&ES) for the past three years, supported by staff, faculty, master’s students, and Ph.D. candidates. F&ES aspires to lead the world toward a sustainable future with cutting-edge research, teaching, and public engagement on society’s evolving and urgent environmental challenges.

Does Quick Carbon only work in grasslands or specific soil types?

So far, Quick Carbon has predominantly tested our protocol in grasslands in the continental US. However, we are expanding our work into new geographies, land uses, and soil types--with possible future work spanning novel applications for Quick Carbon in rangelands, row crops, permanent crops (e.g. orchards, vineyards), and forests, as well as in new climates and geographies.

Does the Quick Carbon protocol require local calibration or is it ready to go?

Quick Carbon first requires a local calibration process. The volume of samples needed for regional calibration can usually be collected in one field season. In the first year of model development for a given region, a small subset (~20%) of the regional soil samples are sent for traditional, highly accurate laboratory analysis, such as gas chromatography-mass spectrometry. This subset of data is then used to build machine learning models relating lab-measured soil carbon levels to the data collected with the field spectrometer.

Once a model is built, carbon content can be estimated in the field in subsequent years using only the reflectometer and remote-sensing data, dramatically reducing sample collection time and cost. Consequently, Quick Carbon users can rapidly collect hundreds of measurements across a landscape, allowing them to produce detailed maps of soil C that reveal distribution across space and robustly estimate soil C stocks. As users re-sample over time, they can improve estimates by evaluating the ecological and management drivers of soil C variability and refocus future sampling efforts on key areas. By learning from itself, the Quick Carbon system improves sampling efficiency as it is used, further reducing measurement and verification costs.

To date, Quick Carbon has collected training data in portions of WY, MT, CO, CA, OK, TX, NY, CT, NH, TN, and ME. We are working to scale up our training datasets so that they might be ready to go in new locations across the United States more rapidly.

Based on preliminary results from Quick Carbon’s work over the last three years, our method was able to estimate total soil carbon content across samples from +/- 0.2 % carbon (Oklahoma) to +/- 0.5 % carbon (California).

More broadly, the accuracy of our method depends on having adequate training data for the particular location you’re working in, as well as accounting for various soil factors that may impact reflectance results. Our ongoing work is focused on improving our training datasets and learning how best to account for those soil factors so that we can provide the most accurate estimates of soil carbon possible.


Quick Carbon does not measure the carbon content of an individual soil sample as accurately as traditional laboratory methods such as dry combustion and mass spectrometry. However, Quick Carbon is far less expensive and can be done in the field, meaning it’s possible to take far more measurements in a given area. With more measurements, we can build maps of estimated soil carbon content that reveal patterns in the landscape and help users better capture spatial variability of soil carbon stocks. Such landscape patterns and variation are not typically captured with traditional methods for measuring soil carbon due to their time-consuming and costly nature. Thus, Quick Carbon vastly improves the spatial accuracy of soil carbon stock estimates.

How does Quick Carbon compare to traditional methods for measuring soil carbon?

Both could impact our estimates. Inorganic C raises total C quite a bit, but it doesn't always produce much of a spectral signal in the visible to near infrared range we work in. If we can collect enough samples representing soils with high inorganic carbon content among our training data, we will have enough input data that the model we train on the backend can predict well for samples with high inorganic C. We are hoping to learn more about the impacts of inorganic C on reflectometer-based estimates of total C as we work with data collected Summer 2018.

Does inorganic carbon interfere with our measurements? Or non-carbonate salts?

Yes! Testing our method on archived soil that has associated soil carbon data is a priority for us, so long as those soil samples have location data and are not aggregated from points that are very far apart in space. If you have a sample archive you think might be useful, contact us per the instructions above.

Can this method be used on soils that were collected and then stored decades ago?

We at Quick Carbon have five main goals for our soil carbon work:

  1. Support efforts to draw down atmospheric carbon into agricultural soils — Quick Carbon can help overcome two major impediments to the broad-scale adoption of carbon-storing land management practices by: 1) Developing soil inventories with the spatial and temporal resolution needed to accurately quantify soil carbon stocks across large scales. 2) Identifying the range of site characteristics over which management increases soil carbon in particular landscapes.

  2. Enable regenerative, adaptive management practices by providing instant feedback on ecological indicators — Quick Carbon can serve as a decision support tool, providing managers with feedback on how the choices they make impact land and enterprises.

  3. Reinforce the science connecting soil carbon content to ecosystem services (e.g. water-holding capacity, growing season length) — Interest in soil and agriculture has extended research efforts beyond the academy. Farmer networks, non-profits, and ag data/information companies with fewer scientific resources than universities will need tools like Quick Carbon that suit their needs and help them co-learn with their client base.

  4. Reduce project transaction costs adequately to open doors for soil carbon markets — Quick Carbon can supply the necessary MRV data for carbon and ecosystem services markets that may develop in the near future.

  5. Engage diverse producers by getting them excited about measurement — Quick Carbon is financially and technologically accessible, built on affordable and everyday tools that bring measurement capabilities and exciting new information about the world below our feet to producers and researchers alike.

That said, we are committed to developing tools that can be used for a wide range of soil carbon applications, from land management to research to policy. We’re building the ‘shovel’the kind of ‘hole’ you dig is up to you!

What are Quick Carbon’s applications?

Quick Carbon makes use of reflectometers designed and sold by Our Sci; we do not produce or sell reflectometers ourselves.

The pocket-sized reflectometers we use measure reflectivity across nine wavelength bands and are much more affordable than a traditional benchtop spectrometer (~$65,000).

Does Quick Carbon sell reflecto-meters?

If you would like to receive more information about our work reach out to us through our contact page. We are always excited to talk about our work!

How can I get more information?

Quick Carbon is excited to build new partnerships and answer questions that further our research in new geographic regions.

How can my organization work with Quick Carbon?