A GIS-Based Methodology for Extracting Historical Land Cover Data from Topographical Maps: Illustration with the Nigerian Topographical Map Series

Abstract

The availability of historical land cover data is a major challenge to long term land change analysis. This is more so in developing countries like Nigeria with weak land information systems and poor inventories of long term land cover data. This situation is to some extent ameliorated by the existence of topographic maps which represent encryptions of historical snapshots of land condition based on primary sources like aerial and land surveys. Topographic maps are however not easily amenable to analysis for land cover data extraction, given their inherent characteristics. This paper presents a GIS-based digitization, symbols analysis, pattern recognition, and polygonization methodology for the extraction of land cover information from topographic maps. The methodology is demonstrated with sheets in the Nigerian topographic map series covering the Idemili River basin. Results show that indigenous settlement types, derived Savanna, and residual forests occupied 27%, 35% and 24% of the basin area, respectively, during the period. An internal data validation approach showed a significant correlation (p = 0.000; r = 0.975) between base topographical map and extracted cover data. There is a need to apply the methodology to other topographical sheets in country’s inventory to build up a national digital database of historical land cover.