Atmospheric boundary condition with evaporation
Recently I'm trying to build an simple model for rainfall filtration in vadose zone with TOUGH2 EOS3. I have read Anthony and Stefan's discussions on this topic four years ago, but still feel confused.
According to TOUGH2 training material, it seems that to set up an atmospheric boundary condition with evaporation needs the following steps:
1) specify an single atmospheric element with large volume and small nodal distance at land surface.
2) Use single phase gas model to simulate atmospheric boundary with less than 100% RH.
3) Use ICP (1) in EOS3 in atmospheric element in which a very strong Pc is estimated based on Kelvin's law.
4) Specify evaporation and infiltration rate in row of elements below the atmospheric boundary element using the GENER Block.
When I follow these steps to set up an atmosphere bc and try to simulate the filtration into vadose zone for one year with annual climate data, I met following problems:
1. How to give an reasonable daily evaporation rate in the model according to the climate data? Do I need to give a negative evaporation rate when rainfall is zero?
2. when I use single phase gas model for atmosphere elements, the model doesn't work. Instead if I use two phase model at the atmosphere elements with an extremely large porosity and low water saturation, then it works.
3. If I need to simulate the filtration in a long term (50 years for example) by assuming every year has the same annual climate data, how could I give an reasonable rainfall and evaporation rate?
So Any ideas? I also attached the climate data for reference. Thanks!
I only have time to answer part of the many issues your raise.
(1) Your Points (2) and (3) combine two different approaches to simulate evaporation. Just go with Option (3) and reread my related posts and/or the paper by Ghezzehei et al. (Vadose Zone J., 3, 806–818, doi:10.2136/vzj2004.0806, 2004).
(2) As you know, infiltration is essentially rainfall minus evapotranspiration minus runoff. I don’t know what type of data you have. If you have (rainfall - runoff) data AND evaporation estimates, just specify the total in the layer below the atmospheric element (and forget about the Kelvin equation, as evaporation is already accounted for). If you want, you can actually put all this into a single GENER block, i.e., both positive and negative values (use COM1 as the component; it automatically switches to MASS if the sum becomes negative). However, it would be cleaner to have two blocks, one for net infiltration being positive, and one for dry-out due the strong evaporation, i.e. a negative net infiltration rate.
(3) I can never help with a statement saying “the model doesn’t work”. The large porosity in a large-volume element is definitely inconsequential.
(4) Sorry, I don’t understand your final question. Implement the “same climate data” in the way described above and repeat 50 times (iTOUGH2 would have a short-cut for repeating rate cycles).