[GRASS-user] r.terraflow crashes
Anna Petrášová
kratochanna at gmail.com
Sun Jan 3 12:46:09 PST 2016
Hi,
On Sun, Jan 3, 2016 at 2:23 PM, Ken Mankoff <mankoff at gmail.com> wrote:
> Hi List,
>
> I'm a grass newbie. I'm trying to route some water through a DEM and think
> r.terraflow is the right tool. It is crashing. Can someone point out what
> I'm doing wrong? I have not found much when searching for this problem on
> google or list archives. I think there are several issues occurring here.
> I'm having the same problems on grass64 (installs easily on OS X) and
> grass71 (on a linux machine).
>
>
> I don't have any experience with r.terraflow, but try using r.watershed.
I'm working in the following region:
>
> > g.region -p
> > projection: 99 (Stereographic)
> > zone: 0
> > datum: wgs84
> > ellipsoid: wgs84
> > north: -657600
> > south: -3349500
> > west: -638000
> > east: 864700
> > nsres: 150
> > ewres: 150
> > rows: 17946
> > cols: 10018
> > cells: 179783028
> >
>
> 1) r.mapcalc produces file of 0 size
>
no idea, have you tried testing it on a smaller region?
Anna
>
> When I run
>
> > r.mapcalc 'phi = (1000*9.8*bed + 917*9.8*(surf-bed))'
>
> The file has 0 size. I now do the same calculation using nctools and then
> load the netcdf file, but I'd rather do this in grass if I could. FYI, bed
> is 589 MB and surf is 190 MB. When I calculate phi with nctools (ncap2 -O
> -v -s "phi=1E3*9.8*bed+917*9.8*thickness" in.nc phi.nc) the resulting
> file is 1.4 GB. Clearly the 0 size grass-produced phi is due to overflow
> int/float/double issues? But a) there is no warning/error and b) wrapping
> the variables above in float()'s doesn't seem to help. The file still has 0
> size if I do a similar but simpler computation:
>
> > r.mapcalc 'phi = (bed + 0.917(surf-bed))'
>
>
> 2) r.terraflow crashes with 'Bad address' error.
>
> > r.terraflow --o --v elevation=phi fill=f direction=d sw=s acc=a tci=tci
>
> The results of that command:
>
> > cell phi header compatible with region header
> > Elevation stored as FLOAT (4B)
> > WARNING: raster phi is of type CELL_TYPE --you should use
> r.terraflow.short
> > Region size is 17946 x 10018
> > STREAM temporary files in
> /home/user/data/grass/GR/PERMANENT/.tmp/host/26900.0 (THESE INTERMEDIATE
> STREAMS WILL NOT BE DELETED IN CASE OF ABNORMAL TERMINATION OF THE PROGRAM.
> TO SAVE SPACE PLEASE DELETE THESE FILES MANUALLY!)
> > Memory manager registering memory in MM_WARN_ON_MEMORY_EXCEEDED mode.
> > Reading data from <phi> to stream
> > </home/user/data/grass/GR/PERMANENT/.tmp/host/26900.0/STREAM_baaaaX>
> > 100%
> > total elements=179783028, nodata elements=18914292
> > largest temporary files:
> > FILL: 9.38G (10067849568) [179783028 elements, 56B each]
> > FLOW: 11.99G (12869498880) [160868736 elements, 80B each]
> > Will need at least 23.97G (25738997760) space available in
> > /home/user/data/grass/GR/PERMANENT/.tmp/host/26900.0
> > MM warning: limit=314572800B. allocating 156610272B. limit exceeded by
> 3706B.
> > EMPQUEUEADAPTIVE: starting in-memory pqueue
> > EMPQUEUEADAPTIVE: available memory: 297.937MB
> > EMPQUEUEADAPTIVE: desired memory: 297.937MB
> > sz_stream: 270400 buf_arity: 200 mm_overhead: 8666496 mm_avail:
> 312409214.
> > EMPQUEUEADAPTIVE: memory overhead set to 8.26501MB
> > EMPQUEUEADAPTIVE: pqsize set to 37967839
> > EMPQUEUEADAPTIVE: starting in-memory pqueue
> > EMPQUEUEADAPTIVE: available memory: 297.163MB
> > EMPQUEUEADAPTIVE: desired memory: 297.163MB
> > sz_stream: 270400 buf_arity: 200 mm_overhead: 8666496 mm_avail:
> 311597910.
> > EMPQUEUEADAPTIVE: memory overhead set to 8.26501MB
> > EMPQUEUEADAPTIVE: pqsize set to 37866426
> > EMPQUEUEADAPTIVE: starting in-memory pqueue
> > EMPQUEUEADAPTIVE: available memory: 296.389MB
> > EMPQUEUEADAPTIVE: desired memory: 296.389MB
> > sz_stream: 270400 buf_arity: 200 mm_overhead: 8666496 mm_avail:
> 310786606.
> > EMPQUEUEADAPTIVE: memory overhead set to 8.26501MB
> > EMPQUEUEADAPTIVE: pqsize set to 37765013
> > MM warning: limit=314572800B. allocating 155798940B. limit exceeded by
> 3706B.
> > rusage/gettimeofday: Bad address
>
>
>
> Thanks for any help,
>
> -k.
> _______________________________________________
> grass-user mailing list
> grass-user at lists.osgeo.org
> http://lists.osgeo.org/mailman/listinfo/grass-user
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/grass-user/attachments/20160103/86db57be/attachment.html>
More information about the grass-user
mailing list