University of Twente Student Theses

Login

Prioritizing mapping areas in OpenStreetMap using quality metrics for improving disaster preparedness

Saleem, Muhammad (2020) Prioritizing mapping areas in OpenStreetMap using quality metrics for improving disaster preparedness.

[img] PDF
6MB
Abstract:According to UN’s Sustainable development Goal (SDG) # 11 there exist challenge in measuring the indicators for disaster risk resilience is the availability of data. Most of the developing countries lacks basic spatial data that is required for measuring risk related to disasters. during the disaster unavailability of data causes delays in rescue operation which puts lives in danger. OpenStreetMap (OSM) is often used for disaster rescue operation to produce and exploit spatial information. This is because OSM is free, open and editable and contains millions of spatial features combined with non-spatial information in the form of tags. Due to its freedom the quality of OSM is often questionable and it needs to be assessed. The quality assessment should be application specific to be able serve best to the users. This study will focus on how spatial data can be produce in OSM to enhance the flood preparedness by prioritising mapping areas and practically implement this by organising coordinated mapping event. This study will analyse the first who are the users, what spatial data do they require for disasters and what are their struggles while dealing with this spatial data. Humanitarian organisations like Red cross are found to be users and they struggles with information about critical infrastructure. Their struggles include completeness, reliability and trustworthiness of information about critical infrastructure and availability of satellite imagery. Two most essential spatial data requirement includes (a) buildings and (b) road networks. Taking this into account four quality metrics (QM) were developed to prioritise the mapping area which are: (1) Tag completeness focuses on the how complete the information about critical infrastructure is in OSM. (2) Reliability checks the last date and frequency of edits of the features representing critical infrastructure. More recent and frequent the edit is more reliable it is. (3)Experienced mappers check the mappers experience. (4) Application specific QM considers the specific end application and prioritise the areas for mapping. Mapping areas are prioritised using this quality metrics approach in selected study area (Dar es Salaam). Results shows that reliability and experienced mapping have correlation with tag completeness. So, the tag completeness and application specific quality metrics will be used in prioritising mapping areas in coordinated mapping event. Second important spatial data requirements are road networks. Keeping in mind its end application routing algorithm is developed which uses the highway tags to assign speed to the edges. Results shows that highway:unclassified is considered as inadequate tag because speed cannot be assigned to edges with such tag. Next three scoring criteria was developed to assess the quality of available satellite imagery. Combination of QMs, findings from routing algorithm and satellite imagery scoring two mapping areas, Dar es Salaam and Tadjoura were selected to measure the success of this approach. AOI-1 is prioritise for highway tags for better routing and AOI-2 is prioritise for mapping buildings for better flood exposure mapping. For comparison and measuring success of this study, mapathon is divided into two groups. First group mapping area is prioritise using QM approach and second group without. The impact of the mapathon is measured by calculating service areas for hospitals in AOI-1 and flood exposure for AOI-2. The mapathon results in significant improvement in quality of the information for both AOI. Although the comparison approach found not be successful in comparing both groups because both groups were in a different geographical location in the same AOI. This creates thecomparison and measuring of success of this approach over traditional one difficult. For this reason, hypothetical mapathon is designed which is divided into two mapping tasks in same area instead of two mapping groups in different locations. First mapping task follows the QM approach and second one without but before the second mapping task all the features would be deleted from the first tasks. In conclusion, this approach is found to be useful in improving the quality of information in OSM. This either by improving the quality of information about geographical accessibility of hospitals in Dar es Salaam or information about number of buildings exposed to floods in Tadjoura, Djibouti. Furthermore, combination of two or more QM found to be more effective than using only the single QM.
Item Type:Essay (Master)
Faculty:ITC: Faculty of Geo-information Science and Earth Observation
Programme:Spatial Engineering MSc (60962)
Link to this item:https://purl.utwente.nl/essays/84937
Export this item as:BibTeX
EndNote
HTML Citation
Reference Manager

 

Repository Staff Only: item control page