Integrative Mapmaking – a Mapping Approach for the Humanitarian Community

Mapmaking in many African countries is still facing substantial challenges on different levels, encompassing technical aspects like fragmentary and insufficient baseline datasets, missing data standards as well as political issues like awareness of territorial sovereignty and unsettled border conflicts [1]. In contrast to these hindering factors the demand for up-to-date spatial data and maps is increasing, especially in peripheral regions like Sudan, where the humanitarian community is heavily depending on updated spatial information to achieve its tasks. From that aspect the increasing endeavors initiated by the UN to foster spatial data infrastructures lead to promising activities among national mapping agencies in many African countries [2].
In the case of Sudan the situation in terms of exchange and management of spatial data is still poor. The manifold mapping activities initiated by UN organizations and local NGO’s in consequence can hardly satisfy the requirements for a concerted and updated spatial database. This was the toehold for the South Sudan mapping project funded by the Swiss Federal Department of Foreign Affairs (FDFA), with the major aim to provide maps and the underlying spatial database to support humanitarian and reconstruction work.
The encompassing spatial database, compiled from various data sources, attempted to bridge the gap between thematic data layers captured on local level and the needs for countrywide standardized spatial information. Therefore, efforts were necessary fostering integrative aspects like the stimulation of data sharing policies and the elaboration of workflows to forward corrections from the field to the database editor. It is the degree of integration of these workflows in the whole mapmaking process that determines the final quality and harmonization level of the spatial database, giving the name to the proposed mapping approach.
In the meanwhile, the Sudanese UN- and affiliate organizations have setup an internet based exchange platform ensuring the update and exchange of locally generated data [4].
To compile the database, baseline data have been vectorized from Russian military topographic maps, from a map scale of 1:200’000 or 1:100’000. Since these map series are dating back to the 70ies and 80ies of the last century, recent objects had to be updated or added based on recent satellite imagery acquired between 2000 and 2005. Due to the sometimes poor topographic information on the old Russian map series, the more recent, freely available SRTM3’ [3] digital elevation dataset was adopted to produce a topographic backdrop for modeling and mapping.

Further data sources for most recent thematic information were locally acting organizations (UNJLC, UNICEF, UNMIS, OCHA, OLS, FAO), which use the internet to disseminate and share their information [4].

All information layers are derived from free or low cost sources, ensuring the practicability of the workflows under budget constraints.

Most data processing steps and data compilation workflows were based on standard routines implemented in GIS software packages like ESRI’s ArcGIS.
Basic tasks like digitizing and updating of features like roads and villages were mainly carried out manually, the update of topographic information like identification of spot heights occurred algorithm based. The update of the mostly poor hydrologic information was a major challenge, since extraction from satellite imagery based on spectral features is virtually impossible in arid environments. Eventually a hydrologic modeling approach was adopted outputting a hierarchical drainage network based on topography.
The land cover information was obtained from satellite imagery processing through rule based classification procedures implemented in the LEICA image processing suite. The extraction of poorly visible agricultural schemes was achieved by the inclusion of textural information in the classification process.
The average geometric accuracy of the spatial database is estimated to be better than 100m, however serious ground truthing and ground control points were never applied so far.
In view of the poor institutional and professional setting the presented integrative mapping approach proved to contribute effectively to a concerted spatial database compilation and dissemination. Meanwhile, the compiled spatial database reached a level justifying its distribution to local organizations and governments in form of printed maps and a spatial database. Further customized dissemination is planned for the near future using web mapping applications.