Questions on Reproducing Model from "Simultaneously determining..." Paper

[chooron]

Hello,

I am writing to you after reading your paper, "Simultaneously determining global sensitivities of model parameters and model structure." I am trying to implement your work using a different hydrological modeling framework, but I have encountered several issues during the reproduction process. I would be grateful for your assistance with the following questions:

1. In the SNOBAL_HBV model, it is not specified which x[?] in your study corresponds to the Tf parameter. The parameter group # [[12,13,14,15,16,17],[],[17,18]], # 5:parameters of snow balance options (structure) only seems to list parameters for ka and SNOW_SWI.
2. The potential snowmelt calculation depends on Mcumul (the cumulative melt since the last period of zero snow depth). I suspect this can be approximated as an intermediate state. When calculating snowmelt, we first need to calculate the potential snowmelt, and this result is then used to update the cumulative melt, Mcumul. However, what puzzles me is that this calculation does not seem to be part of the main snowbalance calculation but is instead treated as a separate, independent process.
3. When performing the weighted sum of the three different snow balance options, the paper's equations seem to combine refreeze (from SNOBAL_HMETS), melt (from SNOBAL_SIMPLE_MELT), and refreeze (from SNOBAL_HBV). The meaning of the second process (melt) is different from the other two. What does the resulting weighted sum represent? My current hypothesis is that a weighted sum of the overflow from each of the three snow balance calculations would be more reasonable, as the meaning of overflow is consistent across all three (i.e., the amount of water flowing out of the snow water storage). I am unsure if this is what the author intended.
4. Each of the three snow balance models has its own state variables, such as SWI, liquidwater, cumulmelt, etc. Do these state variables also need to be combined via a weighted sum, or are the state variables for the three models independent and calculated without interfering with each other?
5. In the infiltration calculation, R represents the "rainfall/snowmelt rate." I assume this is the sum of the two: rainfall + snowmelt. In the context of the paper, should "snowmelt" be interpreted as the overflow?
6. The equation for maximum soil storage is H * n * (1 - SF). I understand from looking at the rvp file that n defaults to 1, but the paper does not seem to mention which x[?] corresponds to SF (STONE_FRAC).