SDR- Error: exception found while running natcap.invest.sdr

Hello NatCappers!

I'm running the SDR model and I am getting the following error. And I see someone had the same questions as mine, but i'm pretty much sure that the type of biophysical table is correct.
Is there any other possible reason? Would you help me to slove this problem? Thanks a lot ! 

10/20/2017 18:55:10 pygeoprocessing.routing.routing_core INFO calculate_flow_weights 87.6% complete

10/20/2017 18:55:17 pygeoprocessing.routing.routing_core INFO calculate_flow_weights 93.4% complete

10/20/2017 18:55:23 pygeoprocessing.routing.routing_core INFO calculate_flow_weights 99.2% complete

10/20/2017 18:55:24 root ERROR ---------------------------------------------------

10/20/2017 18:55:24 root ERROR ---------------------- ERROR ----------------------

10/20/2017 18:55:24 root ERROR ---------------------------------------------------

10/20/2017 18:55:24 root ERROR Error: exception found while running natcap.invest.sdr

10/20/2017 18:55:24 root DEBUG

10/20/2017 18:55:24 root DEBUG Build details

10/20/2017 18:55:24 root DEBUG Interpreter

10/20/2017 18:55:24 root DEBUG Current temp dir: M:\workspace\tmp

10/20/2017 18:55:24 root DEBUG tempfile.tempdir: M:\workspace\tmp

10/20/2017 18:55:24 root DEBUG

10/20/2017 18:55:24 root DEBUG System

10/20/2017 18:55:24 root DEBUG OS : Windows-7-6.1.7601-SP1

10/20/2017 18:55:24 root DEBUG Processor architecture: AMD64

10/20/2017 18:55:24 root DEBUG FS encoding : mbcs

10/20/2017 18:55:24 root DEBUG Preferred encoding: cp936

10/20/2017 18:55:24 root DEBUG

10/20/2017 18:55:24 root DEBUG Python

10/20/2017 18:55:24 root DEBUG Version : 2.7.9

10/20/2017 18:55:24 root DEBUG Build : ('default', 'Dec 10 2014 12:24:55')

10/20/2017 18:55:24 root DEBUG Compiler : MSC v.1500 32 bit (Intel)

10/20/2017 18:55:24 root DEBUG Implementation : CPython

10/20/2017 18:55:24 root DEBUG Architecture : 32bit

10/20/2017 18:55:24 root DEBUG Linkage format : WindowsPE

10/20/2017 18:55:24 root DEBUG

10/20/2017 18:55:24 root DEBUG Packages

10/20/2017 18:55:24 root DEBUG Cython : ?

10/20/2017 18:55:24 root DEBUG Numpy : 1.11.2

10/20/2017 18:55:24 root DEBUG Scipy : 0.16.1

10/20/2017 18:55:24 root DEBUG OSGEO : 1.11.3

10/20/2017 18:55:24 root DEBUG Shapely : 1.5.5

10/20/2017 18:55:24 root DEBUG InVEST : 3.3.3

10/20/2017 18:55:24 root DEBUG

10/20/2017 18:55:24 root DEBUG

10/20/2017 18:55:24 root DEBUG Exception not environment-related

10/20/2017 18:55:24 root DEBUG Printing traceback

Traceback (most recent call last):

File "C:\Users\natcap-servers\jenkins-home\workspace\natcap.invest\label\GCE-windows-1\exe\build\invest\out00-PYZ.pyz\natcap.invest.iui.executor", line 620, in runModel

File "C:\Users\natcap-servers\jenkins-home\workspace\natcap.invest\label\GCE-windows-1\exe\build\invest\out00-PYZ.pyz\natcap.invest.sdr", line 172, in execute

File "C:\Users\natcap-servers\jenkins-home\workspace\natcap.invest\label\GCE-windows-1\exe\build\invest\out00-PYZ.pyz\pygeoprocessing.routing.routing", line 127, in flow_accumulation

File "C:\Users\natcap-servers\jenkins-home\workspace\natcap.invest\label\GCE-windows-1\exe\build\invest\out00-PYZ.pyz\pygeoprocessing.routing.routing", line 88, in route_flux

File "src\pygeoprocessing\routing\routing_core.pyx", line 2712, in pygeoprocessing.routing.routing_core.route_flux (src/pygeoprocessing/routing\routing_core.cpp:33736)

File "src\pygeoprocessing\routing\routing_core.pyx", line 253, in pygeoprocessing.routing.routing_core.calculate_transport (src/pygeoprocessing/routing\routing_core.cpp:6805)

MemoryError



10/20/2017 18:55:24 root ERROR Exiting due to failures


Tagged:

Comments

  • RichRich Administrator, NatCap Staff
    Hi, there's a long standing bug in the routing core of InVEST that makes it difficult to route VERY large DEMs. It's possible that this is your case. Do you have a raster that's on the order or larger than 20k X 20k pixels? If so, do you think you could coarsen the raster to a larger cell resolution? We've done some work to show that SDR model results are well within a reasonable range of uncertainty with 30 meter pixels. Mentioning in case you have very small pixel sizes and are hoping for more accuracy.
  • RichRich Administrator, NatCap Staff
    The other option is to wait less than 6 months until we can patch this issue. Then you can run SDR on the whole planet for all the software will care.
  • Thank you for your kind reply! I think you're right. Because when I resampled the DEM to 90 meter pixers the model was successful. 
    There is another issue  confused me that when I run the model in a very large region,like 500000 square kilometers, some areas would appear null value in the results but the model didn't show any error information and run successfully. And if the region is smaller, the null value would reduce, but still exist.
    Could you help me to solve this issue, the null value? Thanks again!

  • RichRich Administrator, NatCap Staff
    Is it possible that there are hydrological pits on your DEM? The effect you're describing sounds like what we encounter if the DEM doesn't drain and you miss whole upstream sections. You could try a pitfill algorithm like Wang & Liu in SAGA GIS?
Sign In or Register to comment.