"left" and "right" sides of glacier boundaries

[Frank Paul]

Hi Bruce & Co.

I agree with using the hydrological convention for assigning left and
right, using the opposite direction for internal polygons and
follow the conventions given by the software. However, for practical
reasons I want to add two points:

(1) If we obtain glacier outlines automatically from satellite data
(what I thought would be the 'normal' way of getting them ?) we will
have no circulation direction of the outline. Using the 'gridpoly'
command in Arc/Info line segments with 500 vertices each are created
and the respective polygons have the value (label) from the original
Geo-tiff (which is 0 for 'glaciers' and 1 [or 255] for 'other'). In
order to allow calculation of glacier areas the value 1 is set to
NODATA and in the attribute table the item 'grid-code' is used for
glacier identification (-9999 for other and 0 for glaciers).

(2) The digitized Swiss glacier inventory from 1973 also did not use
any specific circulation. The discrimination of glacier polygons from
internal polygons (rock outcrops) is made by manual assignment of
a label to each glacier under investigation. As such, also 'problematic'
glaciers can be excluded from the statistical analysis.

Best regards, Frank