pipeline for precip

[aappling-usgs]

Import and clean up the code for precip data fetching, calculation of cumulative precip, and any other munging

• ensure all precip pipeline functionality gets brought into the template.
• be flexible about spatial resolution (maybe even differing between mobile and desktop?)
• precip data has come from at least 2 different sources for us in the past. data might come from www.srh.noaa.gov/data/ridge2/Precip/qpehourlyshape (maria) or GDP (most others, e.g., https://github.com/USGS-VIZLAB/hurricane-harvey/blob/master/scripts/fetch/fetchPrecip.R, https://github.com/USGS-VIZLAB/hurricane-irma/blob/master/scripts/fetch/fetchPrecip_cells.R). offer both (all?) in this template, to be flexible about the location of future storms.
• it may be possible to simplify the precip-fetching code using sf and other insights; see https://github.com/USGS-VIZLAB/how-gdp/blob/master/scripts/fetch/fetch_utils.R#L14-L49 for our most recent iteration on this

[jordansread]

Creating a fine grid for precip is slow: https://github.com/USGS-R/geoknife/issues/230 but likely faster now in geoknife v1.6.0 We may need to come up with a better way of generating the post for fine grids, but we'll see.

[jordansread]

I see the second set of precip (www.srh.noaa.gov/data/ridge2/Precip/qpehourlyshape) as being a bigger lift to generalize and make robust. Should we table that part of this issue, or get out ahead of it so we aren't caught flatfooted in the future when Stage IV doesn't work for our needs?

[aappling-usgs]

How about table it and make a separate backlog issue for it so we can maybe get to it someday?

[dblodgett-usgs]

Yes. Table that 🗑 🔥.