Carbon: Sample Data?

Does anyone know if the included sample carbon pool data is from a legitimate source? Or is it just random data for comparison?


  • RichRich Administrator, NatCap Staff
    If you're referring to the carbon pools in the sample InVEST biophysical table, I have heard that those were mostly made up and it might be a problem to ship them like real data.  

    In general if we ship real data, we'll attach a source to it.  I apologize for how confusing that might seem!
  • OK, thank you for the clarifications. I was finding different values in other sources, so that makes me feel better.
  • AbdelGISAbdelGIS Member
    hello, can you give me valid sources of the different land use classes in the sample data ? thank you 
  • swolnyswolny Member, NatCap Staff
    As Rich said, we do not have a listing of the sources for the carbon values in the sample data, they really are to be used as samples only, not for applying to your own analysis. To find carbon values for your landcover types and area of interest, we usually do a literature search. For global values, you can start with Ruesch and Gibbs IPCC carbon pools data.

    ~ Stacie
  • Hi,
    If we can't find for the values of carbon pool, are we allowed to use the sample carbon pools InVEST sent us in our research? Thanks
  • swolnyswolny Member, NatCap Staff
    Hi @mssoriano2 -

    No, do not use the InVEST sample data values for your research, they really are only intended for testing.

    If you cannot find carbon values for some of your specific land use/land cover types, and they've not included in the Ruesch and Gibbs global data listed above, you'll have to use values for a more generic land cover type. For example, if you can't find data for a specific tree species, use values for "deciduous forest" or "evergreen forest" or "tropical forest".

    It does take a lot of time to do these literature searches, but get the best that you can now, document the sources, and try to improve them over time if you can.

    ~ Stacie

  • RichRich Administrator, NatCap Staff
    Hey Stacie, do you know who created that original table?  I think it's time to give him a good old "Shame!  For shame on you."

    Along these lines, I've got a low priority task on my list to overhaul/simplify/compress the InVEST sample data.  I suspect that'll get done this year.  We'll take care of the fake carbon pools in that pass.
  • swolnyswolny Member, NatCap Staff
    All of the original tables were made by NatCap alumni a loooooong time ago. The values may be based on something reasonable, but without any supporting documentation (yes, shame on us for that) we don't know for sure, and simply cannot recommend using them for real world applications. What I would probably do is use something like the Ruesch and Gibbs values for new sample data, and if we keep using the Willamette data, group the large number of LULC classes into a more generic slack handful.

    ~ Stacie
  • phifercphiferc Member
    Hi Staci and Rich and et al.

    For my work, I ended up punting to IPCC and/or Gibbs dataset, plus one. The instructions are available on the Carbon how-to page on how to find a value from the IPCC datasets -- it is not a straight forward process!

    Also, try to find a paper similar to what you want to do, and see if they published the values they used for the Carbon module. Often times, these are available in supplemental online information. Goldstein et al. 2012 (PNAS), for example, has the carbon values they used in the supplemental files. Always remember, the more LULC types, the more data you have to find!

    To summarize: carbon values provided are not real and are for testing purposes only AND finding "real" carbon values is time consuming but important.

  • RichRich Administrator, NatCap Staff
    Awesome, thanks for that @phiferc.  Carbon values are a pain.  So much we sometimes dare to dream of an InVEST remote-sensed alternative.  
Sign In or Register to comment.