Update NDR figure

[dcdenu4]

We had a NDR deep dive discussion focused on nutrient load, but also touched on other things. Notes from that discussion can be found in this google doc.

From that conversation we thought the figure shown in this issue could use an update. This is related to natcap/invest#1044, where we'll be adding in the ability to specify nutrient load type of either measured runoff or application rate.

Some notes from the doc about updating the figure (it was mentioned that Hector would update the figure):

• The legend should explain what all the variables are
• Do we have a name for the nitrogen that will be captured between the application and the load? That sink effect. “On-pixel retention?”
• We want to show that you have the option of putting in the application rate or the load directly. The difference between those two is the flag that the user provides in the biophysical table.

[dcdenu4]

We added two columns to the biophysical table: load_type_p and load_type_n. These new columns can have values of application-rate or measured-runoff. If application-rate is given the model will adjust load such that: load_n = load_n * (1 - eff_n).