UW-Hydro
`MAX_SNOW_TEMP` and `MIN_RAIN_TEMP` might could be independent for gridcells?
Feature Requests
I suggest that MAX_SNOW_TEMP and MIN_RAIN_TEMP might be better to be the gridcell parameters e.g. can be determined in new columns in the soil parameter file of classic driver or new variables in parameter file of image driver.
In my application I found it is should be better to calibrate these two parameters when observed snow depth or SWE data are provided. Some study (Ding et al., 2014) also found the two parameter would be impact by elevation, vapor pressure and other factors that might should be different values for different gridcells. I think this improvement should improve the capability of snow simulation of VIC.
References: Ding, B., Yang, K., Qin, J., Wang, L., Chen, Y., He, X., 2014. The dependence of precipitation types on surface elevation and meteorological conditions and its parameterization. J. Hydrol. 513, 154–163. http://dx.doi.org/10.1016/j.jhydrol.2014.03.038.
Another more recent reference: Jennings, K. S., T. S. Winchell, B. Livneh, and N. P. Molotch, 2018: Spatial variation of the rain–snow temperature threshold across the Northern Hemisphere. Nat. Commun., 9, 1148.
(I mention it because I believe they have a gridded dataset that could be added as another field in the VIC parameter file)
I would support this change. Someone would just need to put together a PR and testcase.
I would support this change. Someone would just need to put together a PR and testcase.
It might could be test in my recent project. I would try to test it by amend my R package. It might be implemented by adding two parameter to soil_con struct.
In addition, some studies published in China suggest that single threshold outperforms double threshold in northwest China. I think this might would be similar in some other area worldwide. The function to calculate rainfall might be revised by that MAX_SNOW_TEMP can equal to MIN_RAIN_TEMP.
@jhamman In my latest examination, I found it might be better to make "snowfall fraction" as an optional forcing, as I found many snowfall thresholds are temporal-varying. This modification might be difficult for code level for you so I can try it on my R package.