Informations et ressources scientifiques
sur le développement des zones arides et semi-arides

Accueil du site → Master → Pays Bas → 2014 → Mapping soil variability with a decision tree modelling approach in the northern highlands of Ethiopia

Wageningen University (2014)

Mapping soil variability with a decision tree modelling approach in the northern highlands of Ethiopia

Hof, S.

Titre : Mapping soil variability with a decision tree modelling approach in the northern highlands of Ethiopia

Auteur : Hof, S.

Université de soutenance : Wageningen University

Grade : Master of Science (MS) 2014

Descriptif
In contrast to many other soil mapping surveys, which use a ‘pure’ conventional or digital soil mapping strategy, this is a study in which a more qualitative approach to digital soil mapping is used. The objective is to evaluate the use of decision tree models in order to map soil variability at a detailed resolution in poorly accessible landscapes using limited observations and available auxiliary information. A study area in the Northern Highlands of Ethiopia, is used to implement the conceptual method. The locally developed decision tree model is extrapolated to predict soil type for other representative areas in the region. The results of this study show that the spatial heterogeneity and complexity of the landscape in Northern Ethiopia can be sufficiently mapped with use of decision tree modelling. However, the limited amount of samples and auxiliary information clearly constrained the prediction accuracy.

Mots clés : soil surveys / mapping / decision models / ethiopia

Présentation

Version intégrale (3 Mb)

Page publiée le 15 mars 2015, mise à jour le 14 octobre 2018