Include some basic mapping functionality?

[richardsc]

I made an Rpubs that shows how to do some basic mapping using facilities from the oce package (for which I am a co-author), found here:

https://rpubs.com/clarkrichards/86727

@rich-iannone expressed an interest in including some mapping functionality in stationaRy, so this discussion is to narrow down what that would entail and how it would look. Some questions are:

1. What functionality could be added that would differ from directly connecting with and using other packages that provide mapping (e.g. oce, sp, etc)?
2. What kinds of maps would be general enough to include? Some examples could include:
   • Station locations (as in the Rpub)
   • Met values (e.g. temperature or whatever) for a certain date at a variety of stations
   • Gridded (i.e. spatially interpolated) met fields
   • ... ?

[rich-iannone]

I think the global view of stations using oce would be nice and options can be exposed for different coloring of stations by groups (e.g. by countries, by position/height, by years of consecutive data, by types of met data, by network, etc.).

Another nice implementation of mapping would be to show a map of the filtered stations when filtering the list of stations. Since that's likely to be zoomed in a bit, and perhaps have less stations overall, maybe a leaflet map might be good (with metadata bound to each of the stations).

I like the idea of gridded met fields. I suppose for that we could extract from already available global data (e.g. GDAS, reanalysis files, etc.) or develop those with something simple like CALMET.

[pssguy]

I think the leaflet idea is good and pretty easy to implement with drilldown options I'd be interested in an indicator by stations that are closes to say a country or state capital city . Is there such a dataset anywhere out there already?