ktoddbrown / PrecipOnSOC

Effects of precipitation frequency on SOC distribution
BSD 2-Clause "Simplified" License
0 stars 0 forks source link

Soil data indentification #2

Open ktoddbrown opened 6 years ago

ktoddbrown commented 6 years ago

ISCN3 has several variables to choose from for soil data including organic carbon as a mass-percentage and bulk density. Most data also has lat-lon-depth-time. What other variables are we interested in for this study?

vlbailey commented 6 years ago

definitely OC mass% and Db. How is texture identified? By category (e.g., SiL) or with actual %clay and %sand? Ideally we'd get both. Db plus percent clay would tell me a lot.

Can we get any kind of information about water contents at critical thresholds (Field capacity, PWP)?

bpbond commented 6 years ago

Agreed, Db + OC + %clay seems like a good place to start if possible.

ktoddbrown commented 6 years ago

We can do clay, sand and silt percentage. Keep in mind that everything we add a new variable we reduce the size of the database. That being said, I think we still have a few thousand profiles to work with!

Field capacity is sadly not in ISCN3 though I can check the NRCS data product to see if there is anything there if we want. We might be able to pull electronic conductivity if that is useful.

bpbond commented 6 years ago

Let's not pull unnecessary data. If %sand or %silt reduces the dataset size, I'd drop them–let's focus on clay, which we have a priori reason to think might be important. The other two, not so much.

ktoddbrown commented 6 years ago

bulk density, organic carbon, and clay percentage gives us anywhere from 16K to 5K soil cores to work with from ISCN depending on how strict we want to be with our bulk density. If we only want to use sieved oven-dried soils then we are stuck with 5K. If we are ok with un-sieved and some amount of moisture then we have more.

My inclination is to go with the 5K sample but I wanted your thoughts.

bpbond commented 6 years ago

5K seems like a healthy-sized data set. 👍