Error and Uncertainty
Representing our world through maps has always brought problems. As Longley et al. (2008) state, “it is impossible to make a perfect representation of the world, so uncertainty about it is inevitable”. A simple example is the world map; there is no possible way of representing our spheric world in a flat surface without compromising the size, the shape of the land or the distance between two points. Nevertheless, these uncertain or innacurate representions of our world may still be useful to us.
As Longely et al. (2008) depicts in figure 6.1, as we move from the “real world” through each of the “filters” of research, the representation of the world becomes more distorted. In my experience in my Remote Sensing and Land Use class, we dealt with issues in the “conception” when trying to define what a forest meant for different communities in different locations. In Burkina Faso, dry areas with dispersed trees are considered forrests, while in Rwanda forests are only those with the highest densities of canopy cover. However, deciding the threshold percent of canopy cover that we would use to classify a pixel as forest became a very subjective task. Thus, the forest areas calculated from that classification may have been over or underestimated depending on what we conceptualized as a forest in our code.
Geographers, and any scientist really, should strive to reduce uncertainty as much as possible. The use of new technologies and methodologies may aid in more and more accurate depicition of our world. However, since uncertainty will always be present, geographers have the responsibilty to communicate the sources of uncertainty in their reasearch as well as to clearly outline their methodology and data sources. In an ideal situation, they would follow the principles of Open Science and make available both their methodology and data for further reproduction, replication and scrutiny.
References
Longley, P. A., M. F. Goodchild, D. J. Maguire, and D. W. Rhind. 2008. Geographical information systems and science 2nd ed. Chichester: Wiley. Chapter 6: Uncertainty, pages 127-153.