Import County of Santa Clara buildings & addresses

[impiaaa]

The county publishes a building footprints layer, generated from LiDAR collected in "early 2020." It includes the elevation and height of each building in feet at 1/100 precision. They also publish a parcel layer, which includes the street address of each. We have used this layer as a secondary source in the past, for example in #23. Parcels themselves shouldn't be imported into OpenStreetMap, but the address can be tagged on a building from the buildings layer, or as a node at e.g. the parcel centroid.

Compared to data sets available from the individual cities, neither of these are as high quality. The buildings have less resolution than what were imported in #4; the buildings don't have any extra tags like floor count, usage, or build date; and only one address is available per parcel. However, they are more comprehensive, covering the entire county, and the buildings data is more recent than many of the other buildings layers.

Most of the county's buildings have already been mapped in OSM, and a few cities have done address imports as well. We'll need to consider how to conflate intersecting or duplicate objects: when to replace the OSM building or address with imported geometry, when to keep the OSM geometry but add imported tags, and when to leave the OSM data untouched and skip the imported object; and whether (and how much) to consider the age of the OSM object, the relative level of detail, and the distance or difference in shape.

Since the layers are less detailed, this does not obsolete #37.

[1ec5]

#37 is just for the City of Santa Clara, not the whole county. I think the city dataset would be better for OSM, given its richer attributes and higher resolution. Maybe we could exclude the city from this import if we want to tackle it first. The county data is newer, so we could follow up with county buildings within the city once a city-based import concludes. Does sound reasonable?