[CF-metadata] new standard names for flood simulation

From: Jonathan Gregory <j.m.gregory>
Date: Tue, 11 Aug 2015 17:55:10 +0100

Dear Eizi

I have comments on some of your names.

> 5) ground_altitude

I appreciate the distinction you are making from surface altitude. Yes, we
should stick with surface meaning the bottom of the atmosphere, be it solid
or liquid. In some existing standard names, we have the phrase ground_level
(not just ground, like sea_level). For consistency with these, I'd suggest
ground_level_altitude.

> 1) flood_depth
>
> The flood_depth is the vertical distance
> between the surface of the flood water
> and the surface of ground,
> as measured at a given point in space.

In existing standard names, depth is a vertical coordinate, not a vertical
distance (i.e. a difference between two vertical coordinates). For the vertical
extent of a layer we generally use thickness, but I suppose that flood_water_
thickness would not be comprehensible. Can you think of any other alternatives
to "depth" in this case? Given the definition, a possibility would be height_
of_flood_water_surface_above_ground_level.

> 6) flood_arrival_time
>
> 7) time_at_maximum_flood_depth
>
> 8) time_when_flood_water_goes_below_threshold
>
> 9) time_span_with_flood_depth_above_threshold

Could you make these more consistent? For instance, is the threshold involved
in (6) as well? I guess it should be. In that case (6) and (7) could be
symmetrical

time_when_flood_water_rises_above_threshold
time_when_flood_water_falls_below_threshold

In (7) I think time_of would sound more natural than time_at. In (9) I would
suggest duration instead of span, since duration is more specific to time and
is used duration_of_sunshine. By analogy, yours could be simply duration_of_
flood, if the definition said that this referred to the threshold. Do you also
need to propose a standard name for the threshold?

> 10) hazard_area_flags
>
> The hazard_area_flags is a set of flags representing
> different types of hazards for a given point in space.

This seems rather vague to me. Can you spell out what sort of hazard they are?
Also, I think they should probably be string-valued, and not mention "flag"
in the standard name. The flag attributes are a way of encoding string-valued
data variables.

Best wishes

Jonathan
Received on Tue Aug 11 2015 - 10:55:10 BST