https://www.youtube.com/watch?v=6g75Ce3ZQwQ

Czesc Pawel,

You are right, the results are very method-dependent:

http://gis4geomorphology.com/topographic-index-model/

Kind regards, Balázs

Thanks so much for the answer.

According to what is in the link you gave, in the DTM streams should be burnt in the model and the DTM should be filled ("fill sinks").

For example it is written:

- "In Spatial Analyst, create a flow direction raster from a filled, projected DEM (use meters or feet as units)."

- and also here "The DEM used should be assessed for sinks, and if appropriate, the sinks should be filled (...), filling sinks is appropriate because very few are "real", and the ones that are, when filled, should not affect the result of the CTI calculation".

Particularly, we know that flow accumulation on DTM without fill sink works on average.

Do you know the publications that write about the TWI set with filled DTM and non-filled DTM, and about the differences?

Dear Pawel,

No, I don't have any in mind. However, (important: this is a personal choice/opinion) I don't like filled DTM-s, because filling can considerably modify the scene. If it is a karst region then you should rather forget any fill sink algorithm, anyhow.

Kind regards, Balázs

Pawel

As has been discussed, the conventional "fill" technique within ArcGIS can be a bit of a blunt instrument in that it can smooth and remove the detail from possibly large areas of the terrain model.

A better technique for drainage enforcement is to do a bit of analysis of the terrain model first and identify the properties of the sinks within the model. Properties such as their depth, surface area and catchment area. Those sink points that then meet a selection criteria can then be removed from the model (set to “no data”) prior to the drainage enforcement.

There are what I think are better drainage enforcement tools available than the conventional "fill" technique within ArcGIS. I would suggest you review the “Optimised Pit Removal” tool available from the University of Texas.

http://tools.crwr.utexas.edu/OptimizedPitRemoval/CRWR%20Tools%20Optimized%20Pit%20Removal.html

The tool has a processing limit of 50,000,000 cells, so if your terrain model is larger than that you will need to divide it into a number of overlapping tiles and then merge/fill the results.

The tool requires a .ASC array file of the terrain model as input.

I hope this helps.

Hey! There are free toolboxes for ArcGIS, where You can find ready-to-go tools for calculating TWI (sometimes it calls wetness index or compound topographic index). Look in the internet for these: 1) Geomorphometry & Gradient Metrics; 2) MNDNR Hydro Tools; 3) Morphometric Tool; 4) TauDEM Tools.

All the best!