Chappell, Adrian; (1995) Geostatistical mapping and ordination analyses of 137CS-derived net soil flux in south-west Niger. Doctoral thesis (Ph.D), UCL (University College London).

Text Geostatistical_mapping_and_ord.pdf Download (22MB)

Abstract

Soil loss by wind and water varies considerably in time and space making it difficult to quantify. Monitoring methods are limited by accuracy, precision and replication. The 137Cs technique overcomes many of these problems and can provide net time- integrated soil flux measurements for the last 30 years, but the technique is time- consuming and expensive. This study uses cheap and rapidly assessed soil properties as surrogate variables that reflect the 137Cs concentrations as an alternative method for gaining information on soil loss. The distribution of 137Cs and other related soil properties were sampled over a 600 m2 area. A principal components analysis showed that variation in many of the soil properties is associated with the soil flux status. A non-hierarchical multivariate procedure established six classes of sampling sites according to differences in 137Cs and the other soil properties. The net soil flux for each class and the study area as a whole was then established. Results suggest that the average soil redistribution rates for each class may be predicted using several of the surrogate soil properties. Periodic variograms of soil spatial variation in the study area were found for most soil properties. Large periodic structure was interpreted as the influence of natural horizontally bedded ferricrete terraces, separated by more steeply sloping sections. Smaller periodic structure was interpreted to be due to the accumulation of soil beneath vegetation islands between bare areas. Parameters of the models fitted to these variograms were used in block-kriging property maps for interpreting the processes of net soil redistribution and for subsequent analysis. Block-kriging estimates of properties were used to elucidate soil-landform relations using redundancy analysis, thus avoiding problems with missing data. Compound topographic attributes for susceptibility to fluvial and aeolian activity are surrogates for modelling spatial processes and were also included in the analysis. The results identified the importance of topography-limited surface wash processes in some areas and surface cover limited aeolian processes in others. Redundancy analysis with a forward selection procedure identified the most important soil properties and topographic attributes for predicting net soil flux. Models of the processes of soil redistribution were created for the two main parts of the study area.

Download activity - last month

Export as JSONXMLCSV

Download activity - last 12 months

Export as CSVJSONXML

Downloads by country - last 12 months

Export as XMLJSONCSV

Archive Staff Only

View Item