pipeline for timesteps

[aappling-usgs]

timesteps interact with other datasets but deserve their own champion because they're complex and essential.

• timesteps are determined by the availability of precip data (or they have been, anyway)
• timestep processor should create an .rds file (perhaps a data.frame containing columns for two formats, POSIXct and character string?)
• timestep .rds file should be referenced by everything downstream - aligning the storm track timesteps, displaying the date-time on the plot, and so on. Make sure this is the case.
• timezones will differ by vizzie (because storms occur in different timezones). Ensure that this information only gets entered/determined in one location (parameters?) and is consistent throughout the vizzie (e.g., shows up in these 6 files in Maria: https://github.com/USGS-VIZLAB/hurricane-maria/search?utf8=%E2%9C%93&q=America%2FPuerto_Rico&type=).

[jordansread]

@aappling-usgs good thinking. Flexibility that could allow us to, say, have a vector of times that is every 2 hours for mobile and every 1 hour for desktop might be useful in the future. I assume this would work w/ what you are proposing (esp having a champion for it).