Hanasaki implementation

[Ivanderkelen]

First working version of the Hanasaki implementation

Some issues/thoughts:

• the part where the inflow and demand values are saved in the memory is not yet included (let me know if I should give it a try)

• the monthly mean inflow parameters (RPARAM_in(segIndex)%H06_I_Jan, RPARAM_in(segIndex)%H06_I_Feb, etc) are zero in the module (while they have values on network_topology_Dickson.nc and this file is correctly specified in the Hanasaki_lake.control). I didn't investigate why (yet) but just added hardcoded values for the time being to test the code.

• the release coefficient E_release, needs to be communicated if it is determined at the start of the operational year: do we need to include this as a parameter? Also, a condition to determine the initial release coefficient still needs to be finetuned to only include the first months before the start of the first operational year.

I still need to review my code and check the results themselves, as wel as do some testing with the other reservoir data before integration.

[Ivanderkelen]

I did some testing. The code runs but the results are not fine. It seems that the inflow (q_upstream and RCHFLX_out(iens,segIndex)%REACH_Q_IRF) values are not corresponding to the daily inflow values from input_Dickson.nc. (in the code they are ~10^7; while they should be 10^2 order of magniture. Could this have to do with units or unit conversion?

[ShervanGharari]

Hi, Inne, I have merged your pull; the issue with the Inflows are fixed. It was a mistake that I have done in read_control file and it was reading demand as inflow. so now it is fixed and the inflows are passed to the model.
Re: E_release, I think we have enough parameters to calculate this internally... in the pseudo code we have:
Par_release_ini = storage_obs[0]/(Par_alpha*Par_Smax) # inital value assuming the storage of the first time step
so basically we have all the needed parameters to calculate that but if you feel we can add an extra parameter... if that makes the modelling more flexible and if this assumption on the initial E_release is not logical...
Re: q_upstream; I think I have put the wrong units in the control file it should be mm/day and it becomes reasonable...
you will get a stop of the code at the first time step as the water balance of the lake is not closed yet... so you have to make it work and refine the water balance so the code goes to the next time step. I would suggest you continue with the coding of Hanasaki and see if you can get it working...
one last note, as Naoki mentioned this to me when I started, make sure your text editor thematically removes the tailing spaces in your code. that would be helpful...
looking forward to your new pull!