COMPASS-DOE / grossMethane

analysis code for paired data sets of laboratory-incubations of gross methane flux and associated field net methane and CO2 flux rates
1 stars 1 forks source link

Location and sign of FRAC_K adjustment #16

Closed bpbond closed 1 year ago

bpbond commented 1 year ago

@kendalynnm In 376543c86b01b33faa3c2e0a5930a09ccfe963d7 you put this back into the linear model:

Screen Shot 2022-10-03 at 5 52 31 AM

Let's discuss this morning!

bpbond commented 1 year ago

Hahaha and now the AP fits are, um, degraded 😄

ap_pred

kendalynnm commented 1 year ago

Discussed this morning -> turns out to be not so much the return of FRAC_K to the log (although current thinking dictates that this should be excluded), rather extreme sensitivity of the model to order of magnitude of the input data, which went from 10^1 to 10^-4 with recent dilution factor correction