Let simulate_day do Pre-Columbian calculation.

[jamesmcclain]

The simulate_year function already had this ability. The definition of "Pre-Columbian" has also been changed: previously, anything other than "water", "herbaceous wetland", or "woody wetland" would be transformed into "mixed forest". Now, only built-types are transformed into "mixed forest".

My interpretation of earlier instructions was that anything other than the three types mentioned was to be turned into mixed forest. However, the statement of card #36 lists a number of built-types and says that all of them should be transformed into mixed forest -- I took the liberty of generalizing this to all built types.

Connects #36

[mmcfarland]

I didn't anticipate the previous use of any but water/wetlands being converted to mixed forest. I'll clarify the approach with Stroud and see if they really would rather have only built use types converted instead.

[jamesmcclain]

Okay, sounds good.

[mmcfarland]

Barry recommends:

all "non-natural" land cover types [changed] to mixed forest, including all "developed", "agricultural", and "disturbed (e.g., mining, quarries, etc.)" land. Anything like wetlands, shrub/brush land, prairie, grassland, forest, etc., I would leave as is.

[jamesmcclain]

Okay, I made a change to implement Barry's recommendation.

I have also moved the precolumbian parameter back onto simulate_water_quality and simulate_modifications, undoing an earlier change. It was an error to make that change because it did not allow the Pre-Columbian state to be applied to the water quality calculation. I have expanded the tests to catch this.

[mmcfarland]

Looks good, can you add a section to the README that describes the parameter and the technical meaning of it in our implementation?

Update: Looks like it exists, but we just need to update the technical description which is now outdated.

[mmcfarland]

+1