Opened 15 years ago

Last modified 15 years ago

#264 closed defect (fixed)

Documentation regarding SDTS DEM is incorrect

Reported by: mcoletti@… Owned by: warmerdam
Priority: normal Milestone:
Component: default Version: unspecified
Severity: trivial Keywords:
Cc:

Description

The documentation found at the above URL states <i>USGS DEMs are always returned
with a data type of signed sixteen bit integer."</i>  SDTS DEMS can also be 32
bit floats, particularly in the case of decimeter DEMs.  There is also rumor of
centimeter DEMs, which will also employ 32 bit floats.  The DDSH module will
have a record denoting what binary type the ELEVATION field in the CEL module
will be using.

Change History (2)

comment:1 Changed 15 years ago by warmerdam

I have corrected this.  Can you point me to a "decimeter" DEM? Does this imply
that an elevation of 1 meters is represented as 10 or just that the floating
point values will vary by decimeter increments? 

comment:2 Changed 15 years ago by warmerdam

mcoletti@lychnobite.org wrote:
>>I have corrected this.  Can you point me to a "decimeter" DEM? Does this
>>imply
>>that an elevation of 1 meters is represented as 10 or just that the floating
>>point values will vary by decimeter increments?
> 
> 
> Try Kauai County, Koloa, HI 1:24k SDTS DEM.  It has a "z resolution" of "0.1".
> It has a file name of 1660083.dem.sdts.tar.gz.
> 
> Try downloading it from:
> 
> http://dl1.geocomm.com/download/6928/1660083.DEM.SDTS.TAR.GZ
> 
> You should see elevation values like this:
> 
> CVLS : Cell Values
>         ELEVATION :  = 282.50000000 (FP32)
> 
> CVLS : Cell Values
>         ELEVATION :  = 283.10000610 (FP32)
> 
> CVLS : Cell Values
>         ELEVATION :  = 283.60000610 (FP32)
> 
> CVLS : Cell Values
>         ELEVATION :  = 284.20001221 (FP32)
> 
> CVLS : Cell Values
>         ELEVATION :  = 284.50000000 (FP32)
> 
> CVLS : Cell Values
>         ELEVATION :  = 284.89999390 (FP32)
> 
> CVLS : Cell Values
>         ELEVATION :  = 285.10000610 (FP32)
> 
> Mind, the server is ver-r-r-r-ry slow.  :'(


Mark,

The test file is already missing from dl1.geocomm.com.  Anyways, I tried
another old test file I have (abcdcatd.ddf, etc) and it reports floating
point values with one decimal place of accuracy so it would appear that
(as above) the decimeter datasets are simply measured in meters with one
decimeter of accuracy.   No problem.  I will update the docs accordingly.


Note: See TracTickets for help on using tickets.