Numerical modelling of soil-landscape relationships using diversity indices and conditional probability: a case study from an Iranian arid region

Context. Quantitative and numerical modelling of conceptual and qualitative concepts in the soil-landscape relationship is of great interest for soil mapping. Aims. We quantified some conceptual and qualitative concepts concerning soil-landscape relationships by numerical analysis of landforms in soil identification using diversity indices (DIs) and conditional probability (CP). Method. The geomorphology map was prepared based on the method of Zinck (1989) and used as a basic design for soil sampling. Finally, 200 soil profiles were excavated and described. The DIs and CP were calculated based on soil taxonomic and geomorphological hierarchies. Key results. The DIs increased from landscape to landform level. The lowest and highest DIs were obtained for the soil order and soil family at each geomorphic level. The geomorphic diversity based on the soil taxonomy hierarchy showed that soil orders, including Entisols and Inceptisols, were observed in various landscapes and landforms. In contrast, some soil classes, such as Mollisols and its lower levels, did not have geomorphic diversity. The CP based on the geomorphological hierarchy indicated that the present possibility of a specific soil at the higher level (landscape) was less than at the lower level (landform), indicating more soil homogeneity at lower geomorphic levels. However, the probability of observing a certain geoform increased according to the soil classification hierarchy, consistent with the DI results. Conclusions. The efficiency of DIs and CP in showing the distribution and possibility of soil separation depends on the alignment of soil and geomorphological processes and the diagnosis of these processes.