Blue tab
Green tab
Brown Tab

Preliminary results (examples)



Test site 1. Sokolov - Hierarchised drainage pattern computation from DEMs



Surface drainage is a major contamination pathway in the source – pathway – receptor paradigm commonly used in mining-related risk assessment. It is hence important to make available a comprehensive drainage pattern that provides information about the most probable downstream flow, perennial or not, from a potential contamination source.
The identification of drainage pathways and runoff contributing areas based on DEMs is hence important in assessing the potential downstream contamination of rivers and soils from a given contamination source, being a pit, a waste heap, a mining facility or even a contaminated soil or area.


Several flow accumulation algorithms can be used to compute a drainage pattern from a DEM grid or image (Wilson et al 2008).
The ArcGis Spatial analyst algorithm uses a deterministic approach (Jenson and Domingue, 1988) based on the maximum slope, i.e. the algorithm directs flow from each grid cell to one of eight nearest neighbor cells based on slope gradient (selecting the greatest slope).  This algorithm tends to produce parallel flow paths on planar surfaces.

The stochastic algorithm used here, developed with J. Fairfield (University of Virginia) introduces a degree of randomness to break up this unwanted parallel flow paths; This algorithm starts by identifying all the neighboring downslope cells, then calculates the slope gradients in each of these directions, and finally extracts random numbers from a table to direct the flow to one of these candidate cells. The random numbers are allocated on a slope-weighted basis such that the potential flow paths with the steepest gradients have the greatest probability of being selected, and the overall flow pattern more or less matches the one produced with the deterministic approach.

Illustration of determination process for flow patterns
Illustration of determination process for flow patterns



Attempts over Sokolov using and ASTER DEM

A problem encountered when computing flow accumulation from DEMs are "pits". Most the algorithms fill these pits to avoid flow "falling into the pit" and not flowing downstream.

In the case of Sokolov, this leads to errors in the flow computation: the computed drainages flows out of the mine open pits as if the water was filling the open pit before flowing out (see Figure 1). If the "no fill" option is activated, the water flowing from outside the pits remains inside (Figure 2 below).

Click on image to see it full-sized.

Figure 1. Flow accumulation image Figure 2: Flow accumulation image
Figure 1: Flow accumulation from ASTER DEM using deterministic approach, open pits filled.
Figure 2: Flow accumulation from ASTER DEM using deterministic approach, open pits not filled. The difference in drainage patterns using deterministic and stochastic approach can be seen from Figure 1 and Figure 3
Figure 3: Flow accumulation image  
Figure 3: Flow accumulation from ASTER DEM using stochastic approach, open pits filled.