Six terrain attributes were subjected to K-means clustering performed in SAGA-GIS using the hill-climbing method (Rubin, 1967) using a terrain attribute combination of multiresolution valley bottom flatness, multiresolution ridgetop flatness, planform curvature, profile curvature, SAGA wetness index, and topographic position index. This process was intended to mimic the geometry of ‘fully developed slopes’ observed within the landscape. (Wood, 1942; King, 1957). Examination of 5-cluster map by overlaying it on the satellite imagery showed that clustered correspond to different slope positions viz summit, shoulder, backslope, toeslope, and footslope and was used to predict soil properties and disaggregated for the Busia area.

King, C. (1957). The uniformitarian nature of hillslopes. Transactions of the Edinburgh Geological Society, 17:81-102.

Rubin, J. (1967). Optimal classification into groups: an approach for solving the taxonomy problem. Journal of Theoretical Biology, 15:103-144.

Wood, A. (1942): The development of hillslopes. Geologists Association Proceedings 53:128-138.

Researchers should cite this work as follows:

• Minai, J. O.; Schulze, D. G. (2019). Busia K-Means Cluster Map. Purdue University Research Repository. doi:10.4231/4Q9T-FT90

  BibTex | EndNote

1. Agronomy
2. Agronomy (Dept)
3. Geographic Information Systems (GIS)
4. K-Means Clustering
5. Multiresolution Ridgetop Flatness
6. Multiresolution Valley Bottom Flatness
7. Planform Curvature
8. Profile Curvature
9. Slope Position
10. Terrain Attributes
11. Topographic Position Index