I love our table on the source of event reports versus number of events reported in 2019. In looking through these, I think that there might be some general categories we could define, and then that would give us another layer of structure in talking about general types of sources.
The ones that I was thinking of as a first pass (and the specific sources that might make sense within each) were:
Automated observation systems:
Mesonet
ASOS
AWOS
River/Stream Gage
Drought Monitor (we should check that this is from automated equipment, otherwise it should go in a different category)
Buoy
SNOTEL (we should check that this is from automated equipment, otherwise it should go in a different category)
RAWS
C-MAN Station
AWSS
Officials or emergency response/agency:
Emergency Manager
Law Enforcement
911 Call Center
Department of Highways
State Official
Fire Department / Rescue
NWS Employee
Other Federal Agency
County Official
Park/Forest Service
Local official
Lifeguard
Post Office
Tribal Official
Coast Guard
Trained volunteers:
Trained spotter
COOP Observer (I wasn't sure here---is this a trained volunteer similar to a trained spotter?)
CoCoRaHS (I think this is a network from trained volunteers collecting precipitation, right?)
Storm Chaser (I think we would consider them "trained"? If not, could be move to general public)
Media:
Broadcast Media
Newspaper
WLON (I'm guessing this really should have been coded in the data as "Broadcast Media")
Company or general public:
Public
Social Media
Amateur Radio
Utility Company
Mariner
Airplane Pilot
Insurance Company
Ones I wasn't sure where to put (maybe we could get some more details on them?):
NWS Storm Survey
Official NWS Observations (are these by equipment or by people?)
SHAVE Project
With these categories, we could provide not only the plot of specific sources, but also estimate what proportion of events of a certain type were reported in each of these general categories. I would think that the automated networks would be more consistent from event to event and would be unlikely to miss an event if it happens (you don't need someone there to "see" the event, because the monitor is always in its spot and on), although if you have a spatially sparse network, I guess an event could slip in between monitors without you seeing it.
If we decide to do this split, then we could organize the table with number of events reported by source, as well as the heat map with percent of events of a certain type reported by each source, to be "blocked" or grouped by these general types.
For the table, there's more on how to achieve this with the kableExtra package here.
For the heat map, we might want to look at some of the resources here:
I love our table on the source of event reports versus number of events reported in 2019. In looking through these, I think that there might be some general categories we could define, and then that would give us another layer of structure in talking about general types of sources.
The ones that I was thinking of as a first pass (and the specific sources that might make sense within each) were:
Ones I wasn't sure where to put (maybe we could get some more details on them?):
With these categories, we could provide not only the plot of specific sources, but also estimate what proportion of events of a certain type were reported in each of these general categories. I would think that the automated networks would be more consistent from event to event and would be unlikely to miss an event if it happens (you don't need someone there to "see" the event, because the monitor is always in its spot and on), although if you have a spatially sparse network, I guess an event could slip in between monitors without you seeing it.
If we decide to do this split, then we could organize the table with number of events reported by source, as well as the heat map with percent of events of a certain type reported by each source, to be "blocked" or grouped by these general types.
For the table, there's more on how to achieve this with the
kableExtra
package here.For the heat map, we might want to look at some of the resources here: