arcgis vs python preprocessing toolkit

[Jana Corominas]

Hello everyone,

I have always used ArcGIS Pre-processing Toolkit to prepare input data for WRF-Hydro model. However, I don't always have an ArcGIS license so I would like to have another way to prepare static data.

I found wrf_hydro_gis_preprocessor and managed to use it with same data previously used with ArcGIS Toolkit, however the output is not identical. 
Github python code is in BETA release, which makes me think it might not be up to date with the latest code. 

As an example, I have attached two CHANNELGRID maps that show some differences, one with python code based and the other with arcgis toolkit. 

Does anyone know what could be happening?

Thanks a lot. 

[Kevin Sampson]

Hi Jana,

These tools use different underlying geospatial libraries, and thus the result cannot be expected to be identical. The ArcGIS version uses Spatial Analyst hydrology tools, while the open-source python version uses WhiteBox tools. They implement and support different hydrological processing algorithms, which is why there may be slight differences between them. I have extensively tested both suites of pre-processing tools and have tried to ensure that they have similar capabilities. I would say that they are both current and updated, even though the open-source tools have not been taken out of Beta status. My opinion is that I prefer the open-source tools as they will run in many environments, including docker/linux, and you can modify them to use alternative algorithms that are available in WhiteBox Tools. Another benefit is that they typically run about 15-20x faster than the ArcGIS tools.

Cheers,

Kevin