Fiji Sigatoka MCDA-GIS Data

Published: 14-01-2021| Version 1 | DOI: 10.17632/txn9pj4nn4.1
John Lowry


The research objective of this paper was to present a set of methods that use a combination of GIS and MCDA to evaluate land suitability for a selection of target crops for agroforestry spatial planning. The first objective is to show how commonly available edaphic and environmental cultivation criteria for agroforestry crops (through websites such as can be used within a GIS to identify the most suitable areas for target crops. Because suitable land areas for target crops overlap, there is the issue of deciding which locations are best suited to which crops. The second objective is to present ways of dealing with the problem of deciding where a suite of target crops should be planted. We first look at a method that deals with this problem using an equitable distribution approach—one that distributes the crops to locations based on the proportion of suitable land available to each crop. We then examine a method that uses expert knowledge and MCDA to quantify the relative benefits of each crop and determines the spatial distribution of crops most suitable for different agroforestry strategic goals. In this repository you will find the ArcGIS Pro v 2.5 project with accompanying GIS and and models for our case study. In addition you will find the Excel tables used for AHP calculations. Land Suitability model: The default cultivation parameters are for Tropical almond. See the manuscript for other crop parameters. Weighted Maximum model: The default weights are for the Economic benefit. See the manuscript for AHP weightings of other benefits. See README.txt file for more details


Steps to reproduce

Viewing files in these folders requires MS Excel and ArcGIS Pro from ESRI. ModelBuilder models and study area GIS data can be found the the Sigatoka ArcGIS Pro project: Land Suitability model -- creates binary crop suitability raster based on input cultivation criteria. Requires user specified range parameters. Weighted Maximum model -- creates weighted maximum output. Requires binary suitability rasters as input. Requires user specified weights.