The ABC method: Hannah E.E.S. Haga defended her MSC thesis
One of our students recently defended her MSC thesis, "A new method to quantify the accuracy of classification and spatial delineation in land cover maps". The method she developed is already being used by the GEco group to train students in vegetation mapping (BIO4120).
Hannah's thesis is the latest in GEco’s ongoing research into the quality of field-based maps. She used a new method (the 'ABC method') to separate two forms of inaccuracy in maps: inaccuracy in spatial delineation of polygons, i.e. where boundaries are set between patches of nature, and inaccuracy in classification, i.e. what ecosystem type is assigned to a polygon.
The goal was to explore inaccuracy in classification and spatial delineation, and develop a new method for training new mappers. First, GEco members mapped four study sites in parallel. These parallel maps were used to make ‘true’ consensus maps. Then, groups of students mapped the same sites. This process was dubbed the 'ABC' method because the study sites were split into three parts: in part A, the students classified ready-made polygons to ecosystem type. In part B, the students drew polygons around point coordinates that were already classified to ecosystem type. Lastly, in part C, both procedures were combined in wall-to-wall mapping.
Hannah's method is an exciting way to teach mapping to students, and holds promise for setting a baseline for map quality if experienced mappers can be tested with the same method.