Re-analyzed Tair for Tsoil, Pm for SWC surrogates in soil respiration modeling
More model statistics (slope, RMSE, residual, EF and d) have been evaluated
Results showed Tair is a good surrogate for Tsoil in soil respiration modeling
However, Pm is not a good surrogate for SWC in soil respiration modeling
But Pm can explain a similar amount of soil respiration variability as SWC
Thus, the mechanism of precipitation or SWC affect soil respiration should be different
Following up analysis of surrogates
We could separate SWC vs Rs regressions into different groups according to p(slope), EF, and d, and using a decision tree analysis to see when SWC can well predict Rs and when not.
Same process for the precipitation vs Rs relationship (see figure below).
Alternatively, we could do something like this: within a site, if Pm vs Rs linear regression has an opposite slope as SWC vs Rs linear regression, or Pm vs Rs showed significant linear relationship (p(slope)<0.05) but SWC vs Rs showed no linear relationship (p(slope)>=0.05) or SWC vs Rs showed significant linear relationship (p(slope)<0.05) but Pm vs Rs showed no linear relationship (p(slope)>=0.05), we set Pm is not a good surrogate of SWC in Rs modeling, otherwise, we say Pm is a good surrogate of SWC.
Then we use a decision tree approach to see when Pm is a good surrogate of SWC, and when it is not a good surrogate of SWC in Rs modeling.
SRDB project
Time and date data input and check finished.
Checked Darlin and Jason's data input.
Global C project
Ready for AGU poster presentation.
Have some idea about the color we used in Figure 1-3 of AGU poster.
Done last week
Surrogates manuscript
Following up analysis of surrogates
SRDB project
Global C project