GPSFileDepot.com
 

News:

Welcome to GPSFileDepot!

Main Menu

Contour interval for non-US maps

Started by maps4gps, March 16, 2010, 04:20:55 PM

Previous topic - Next topic

maps4gps

I have been doing some tests contouring of the world using the 90m SRTM data.

I expect the display scale for the detailed contours would be about .3 mi (.5 mi on the OR300) which is about 1:150,000 or 1:125,000. 

1.   Would a good series of contour intervals in meters(depending on slope) be:
       50 - 25 - 12.5? - 6.25??
   or 40 - 20 - 10? - 5?
   or 30 - 15 -  7.5? (which is close to 100 - 50 - 25 feet)

2.   Should the contours not be 'touching' at the zoom level they first display at, or would zooming-in one level for 'clarity' be acceptable? 

3.   Any value to including physical features like summits, etc - most of the coordinates in the NGA data base only record location to the nearest minute (6100 ft / 2000 m)?

 

maps4gps

90meter version 4.1 from CGIAR
Expect this would be about correct for full contours at the 21 bit level - scale around 125k or 150k.

Using a 20' CI where the USGS source was 80' or 40' does make a tan display until zoomed in beyond the 'resolution' of the source data.  Seems many users think a smaller CI is 'better'.

What CI value to use from a series for a given area would depend on the slopes in that area.  I did not want to use a series which was unusual for metric users.  I would guess that 50 - 25 - 12.5 - 6.25 is not used; perhaps 50 - 25 - 5 is; or maybe 40 - 20 - 10 - 5 ?

maps4gps

I have seen 80m or 40m on some of the European 250ks, but also 100m or 50m on others.

Not sure on 1/2 for 100ks. For 15 or so quads I checked in various locations in AK, the 63360s were only 1/2 of the 250ks.

Agree on the void filling.  An alternative would be to use the non-void filled set from NASA and have GM contour with void-fill off and create discontinuous contours.

I was thinking of doing a visual on each 'quad' (whatever size that might be) and rerunning at 1/2 or 2x, etc. for the visual appearence.  USGS indicates they use visual appearence at the printed scale to determine contour interval; therefore visual appearence on the GPSr at the zoom display  the full set of contours first appear (within the limit of what the source data supports).  Yes, no ??

Perhaps the currenly available SRTM data would not be useful.  I was looking for something better than those in VMAP0; possibly to be used the OSM road data and some source of hydro, etc.  If not feasable, that would be one less item on the 'to-do' list.

arash


maps4gps

#4
The maps in the international map collection at a local University were using: 100m, 100m/sup50m/ 50m, 50/sup25m, 20m, and 20m/sup10m.  Seemed like some western European countries are more uniform than USGS is and would have some quads with very flat terrain with only a few small closed areas using a CI of 20m rather than use a 10m or 5m CI.

Last 'test' run at 50m in area of high slopes and 20m elsewhere.  I may do 10m in very flat areas (which might exceed the SRMT data quality - overall 95% confidence level is 13 meters)