Using Historical Vertical Aerial Photography to Map Landscape Change Through Time in the Central Subarctic

Abstract

In the era of climate change, landscape features and their systems are becoming altered in unpredictable ways. In sensitive areas such as the arctic, it is important to monitor these landscape features and develop protection strategies to keep the natural systems they are connected to in balance. However, the first step in this process is to quantify the extent to which these features are changing. One of the ways in which we can do this is through historical vertical aerial photography. This study looks at the feasibility of distinguishing landscape features from 20 historical vertical aerial photographs located in the central Northwest Territories and their feasibility of quantification through time. By narrowing in on the vegetated margins of ponds, tree islands between the forest-tundra ecozone and shrub boundaries this project discerned the feasibility of measuring change on the arctic landscape through time. This was done by comparing the historic photographs to contemporary satellite imagery, and were quantified through ArcMap v 10.5. The study found 110 ponds increased in area over time and 265 declined. 17 of those ponds decreased and fragmented dramatically, but it was not enough to attribute these outliers to landscape wide change. The number of trees found in the 10 plots surveyed on the landscape saw growth, with densities doubling through time, however areas of low density in 1946 were subject to low densities in 2010. Shrub boundaries were not measured, as through careful consideration, the orthorectification of the photographs did not prove to be reliable enough to produce conclusive results. The manipulation of the historical vertical aerial photographs proved to produce a number of opportunities for inaccuracies, and these were critically observed and a set of best practices was developed to help with the maneuvering and feature discursion of these remote photographs. Future research would include the ground validation of these results and a set of intermediate data between the two time periods.