SUN1999 albedo implementation in snow_utility.c about vic

[bartnijssen]

Comment from Ted Bohn per email:

I agree, it appears to be an incomplete implementation. I took that code from Keith Cherkauer many years ago. I didn't read the Sun et al paper very closely, so I missed the mismatch between the code and the paper. I should have been more cautious; however fortunately I implemented it as an option.

I suspect few people have used that option. At the time, the US ACE option was hard-wired to switch from accumulation to ablation on March 1, which made little sense in the Southern Hemisphere. So, the Sun et al option appeared to be a more process-based scheme that would be guaranteed to work globally. But in the intervening years, we've updated the US ACE scheme to use October 1 for the cut-off date in the Southern Hemisphere, so the Sun et al scheme no longer has that advantage.

I personally don't see much benefit to keeping the Sun et al option. To keep it, we would need to implement it correctly and then validate it against observations. This would take time and effort that could be spent elsewhere, for a questionable amount of improvement in the albedo scheme.

[bartnijssen]

Agreed - since the method is incompletely implemented I would mark it for removal. This will require a few changes:

• remove SUN1999 as an option (in the function that reads this option)
• remove code in snow_utility.c that implements this option

I suggest that we do this as part of 4.2 as a cleanup item. If someone wants to take on implementing this correctly, we can bring it back in a future version.

I remarked it as a bug, which can be closed when the option is removed

[bartnijssen]

Closed per pull request #45

[jhamman]

I don't think these changes made it into the develop branch. I have removed the option in my development branch and will do the merge once the testing is complete.