Previous PageTable Of ContentsNext Page

Applying geostatistics, fuzzy logic and geographic information systems to soil quality assessment

Inakwu O.A. Odeh1, Budiman Minasny2, Alex B. McBratney3 and John Triantafilis4

1Australian Cotton CRC, Faculty of Agriculture, Food & Natural Resources, Ross St Building A03, The University of Sydney, NSW 2006, cotton.crc.org.au/ Email i.odeh@acss.usyd.edu.au
2
Australian Centre for Precision Agriculture, Faculty of Agriculture, Food & Natural Resources, McMillan Building A05, The University of Sydney, NSW 2006, www.agric.usyd.edu.au/acpa/ Email budiman@acss.usyd.edu.au
3
Australian Centre for Precision Agriculture, Faculty of Agriculture, Food & Natural Resources, McMillan Building A05, The University of Sydney, NSW 2006, www.agric.usyd.edu.au/acpa/ Email Alex.McBratney@acss.usyd.edu.au
4
Australian Cotton CRC, Faculty of Agriculture, Food & Natural Resources, Ross St Building A03, The University of Sydney, cotton.crc.org.au// Email johnt@acss.usyd.edu.au Currently, School of Biological Earth & Environmental Sciences, Biosciences Building, University of New South Wales, NSW 2052, www.bees.unsw.edu.au Email j.triantafilis@unsw.edu.au

Abstract

Geostatistics together with geographic information systems (GIS) offer a powerful tool for the spatial analysis of soil quality indicators for land management purpose. While the geostatistical method offers avenues of spatially modelling the soil quality indicators at their specified level of interest, GIS provide tools for querying based on classical Boolean logic, spatial manipulations and overlay of the resulting layers, especially in combining the effects of multiple soil quality indicators, and therefore consequent to land management. One other difficulty with soil quality assessment is the aggregation of multiple quality indices into a single index. But fuzzy sets and fuzzy logic offer this possibility.

Soil structure is one of the most important soil quality indicators that influence its productivity. Soil structural decline is related to many physico-chemical processes, the most important of which are sodification and salinisation. Electrical conductivity (EC) of the soil solution and exchangeable sodium percentage (ESP) are good indicators of salinisation and sodification of the soil. In Australia, sodic soils are defined as those with ESP ≥ 6 %. However, the use of ESP as the primary soil structural stability indicator becomes more complicated when, in certain conditions, it does not consistently relate to the actual aggregate stability and dispersion exhibited by the soil. Relatively high electrolyte concentration and low Ca:Mg ratio (< 2) of the exchangeable complex are the other factors reported to inhibit the impact of ESP on soil dispersion. Specifically, a given soil type with high electrolyte concentration (EC1:5), even at high level of ESP, may exhibits favourable soil structural conditions. This has led to the development of an index of structural stability termed the electrochemical stability index (ESI), a ratio of ESP to EC. The aim of this paper is to examine the critical levels of ESP, EC and Ca:MG, and ESI, at which significant soil structural decline may occur and use their critical levels to spatially model the potential risk of structural decline in the lower Namoi and Bourke Irrigation Districts of northern New South Wales (NSW), Australia. We applied the non-parametric geostatistical approach of multiple indicator kriging to derive the probabilistic risk of exceeding multiple thresholds of these indicators, and the associated uncertainty. We displayed these map layers in a GIS, and utilised the GIS query capability to decipher the spatial relationships among the indicators of soil structural decline and a measure of actual structural breakdown (based on Aggregate Stability in WAter Test― ASWAT). Having deciphered the appropriate critical levels of these soil quality indicators, we applied fuzzy logic to aggregate soil quality indices into a single index of soil structural quality. This index can be displayed readily in a GIS.

Previous PageTop Of PageNext Page