What interpolation algorithm does clm5 apply when the spatial resolution of forcing data is different from the resolution of model domain ?

[Xueli Huo]

Hi

I am running clm5 with the configuration of datm. Four experiments are set up to investigate the effect of different domain resolution on model outputs.
Each of the experiments is run with a different set of domain and surface data in resolution, but with the same forcing data which has the resolution of 0.9x1.25.
The resolutions of all four sets of domain and surface data are higher/finer than the resolution of the forcing data. And the results show the advantage of providing finer domain and surface data set is overwhelmed by the corse resolution of forcing data. My question is
1. What kind of spatial interpolation does datm apply when the forcing and model domain resolution is different ?
2. The overwhelming result implies that if we want to do high resolution simulation, we have to make sure both the forcing and domain/surface data are high-resolution ?

Sincerely,

 

[oleson]

1. Bilinear interpolation (mapalgo = "bilinear", ... in datm_in)
2. If you want to imitate what might be obtained from a high-resolution coupled model, then yes.