JSON file in USJ needs revision

[mlmaskey]

There exist the following discrepancies among the nodes or links defined in the pywr_model.json file for Upper San Joaquin River and openagua:

• At least 20 nodes or links that appeared in the openagua are not included in the JSON file. Below Big Creek 2 Tunnel Inflow, Below Kerckhoff 1 Tunnel Inflow, Below Portal Penstock Outflow.2, Below SJN_09a Headflow, Below SJN_09a Inflow, Big Creek 2 Tunnel, Friant-Kern Canal.1, PH 2A Aqueduct, SAN-L-RIV4.1, SAN-L-RIV4.2, SJN_01_Inflow.1, Big Creek 2 Tunnel Inflow, Big Creek 8 Penstock Outflow, CVP Madera Canal Demand, Millerton Lake Flood Release, IFR No. Fk. Stevenson Creek above Shaver Lake, SAN J-N-JUN1, SAN J-N-JUN3, SAN J-N-JUN9, SJN_09a Inflow, USGS 11238400 BIG C PH NO 2A NR BIG CREEK CA, UPPER-L-CON3, UPPER-L-CON4, S Fk San Joaquin Inflow

• Because of missing nodes, at least 30 links are no properly defined. For instance, the link [ "Willow Creek.1", "SJN_08 Inflow" ] should be [ "Willow Creek.1.2", "SJN_08 Inflow" ] according to the openagua. Likewise, ["Below SJN_25 Inflow", "Big Creek 5 Diversion Reservoir"] does not exist in the JSON file, however `[ "Below SJN_25 Inflow", "Big Creek 2 Tunnel Inflow" ].

• It seems that the node Big Creek 5 Diversion Reservoir is not defined in the JSON file.

After fixing all the links, the model run becomes infeasible. So, it ought to stick to the default case. @rheinheimer , it seems we need to fix the network or JSON file to get the appropriate result.

[drheinheimer]

@mlmaskey, thanks for this. A few points to note:

1. The naming is indeed a mess. This is a result of pulling from the WEAP models (in which there is no “Big Creek 5 Diversion Reservoir”, for example, just a diversion from the river), and having some auto-naming scheme in OpenAgua (rather than having to name every single diversion/junction/split link manually).
2. Absolutely none of this affects model performance/results. All needed facilities are included, even if they don't currently align 1:1 with what is in the OpenAgua schematic.
3. While we could manually change the Pywr JSON file to map 1:1 with what is in OpenAgua, I'm not sure it's a good use of time, since a) see 2. above, and b) there is an additional processing of the original model JSON file to simplify the network, effectively removing most of these nodes that we'd be adding in.

HOWEVER, it might be worth renaming those key facilities that are currently mis-named in the JSON file, such as the Big Creek 5 Diversion Reservoir (possibly all diversion reservoirs are so misnamed)