K-means clustering of a soil sampling scheme with data on the morphography of the Ogosta valley northwestern Bulgaria
Published 2019-01-01
Keywords
- Spatial clustering, K-means clustering, River pollution, Digital Terrain Model (DTM), Floodplain, Ogosta River
How to Cite
This work is licensed under a Creative Commons Attribution 4.0 International License.
Abstract
The spatial distribution of 665 soil sampling sites in the arsenic contaminated floodplain of the Ogosta River in the Northwest of Bulgaria is analysed against geomorphological parameters computed from a precise digital terrain model. The study aims at partitioning and
classifications of hidden patterns of the morphographic features of the river floodplain, which to be used for the explanation of the arsenic dispersal in the polluted soils at a further stage. The field sites are split into 4 clusters using K-means algorithm with the following variables: elevation, distance to the river, vertical distance to channel network, multiresolution index of valley bottom flatness and a modified topographic SAGA wetness index. It is found that each cluster is related to a distinct area in the valley and is in good agreement with the distribution of the previously determined geomorphological units, as well as with the extent of a simulated historic flood