Visualization of uncertainty interviews for Masters' study
Uncertainty is inherent in all spatial data. Many studies have been done around uncertainty and much is known
about uncertainty. However, this information does not always find its way into the foreground when spatial
data is used.
My name is Sven Christ. I am doing a study on uncertainty and the visualizing of uncertainty for my Master’s
degree in Geography and Environmental Studies at Stellenbosch University. My study has three major tasks:
establishing a baseline of the South African perception on uncertainty, developing a tool for uncertainty
visualization and finally getting external validation on the usefulness of the tool.
The first two tasks have now been completed and it is time to start with the final task. I hereby call on your help
as users and producers of spatial data to help me achieve my final task. I need individuals who are willing to
participate in my study through interviews.
These interviews will start off with an introductory chat about uncertainty. A piece of data will also be
introduced and discussed briefly. This then leads into a demonstration of the developed software tool. Final
remarks on the tool as well as the implications it has on the data will then be discussed.
These interviews will be arranged at a time and location convenient to the interviewee. All identifying
information will be kept confidential and no names will be mentioned in the research paper or anywhere as
per the Stellenbosch University Research Ethics Committee requirements.
If you are interested in being a participant in the study please contact me via this sheet
http://goo.gl/forms/E12s8CNtbc. If you require more information about the study you may contact me as well.
The software tool will be available as an open source plugin for QGIS at the completion of the study.  The
complete study paper will also be freely available through the Stellenbosch University Library online thesis
programme.
Example:
Below is two sample datasets that have been run through the tool. Dataset 1L was treated as a reference
dataset for dataset 2S. The two datasets represent basins that were modelled from two DEM datasets, both 5m
resolution however producing slightly different results. The map on the right shows the output of the
uncertainty visualization tool as an overlay showing the areas of higher uncertainty in the original DEMs and
how/if they correlate to differences in the datasets modelled from the original data.
Thank you to BEATS-SA for allowing me to use this space.