Remote Sensing in Ecology and Conservation (Sep 2019)

Estimates of landscape composition from terrestrial oblique photographs suggest homogenization of Rocky Mountain landscapes over the last century

  • Julie A. Fortin,
  • Jason T. Fisher,
  • Jeanine M. Rhemtulla,
  • Eric S. Higgs

DOI
https://doi.org/10.1002/rse2.100
Journal volume & issue
Vol. 5, no. 3
pp. 224 – 236

Abstract

Read online

Abstract While orthogonal (i.e., aerial or satellite) imagery has become the more conventional source of land cover data because it can yield spatially accurate land cover maps, terrestrial oblique photographs present a valuable, relatively untapped source of raw optical data for studies of land cover change. We present a case study contrasting how these two types of imagery sample landscape composition and using repeat oblique photographs to evaluate long‐term land cover change in a remote region of the Canadian Rocky Mountains. We classified 46 historical oblique photographs and their corresponding modern repeats using the same discrete land cover classes employed in a Landsat‐based map of the same area. We compared landscape‐level composition estimates from both sources and regressed the land cover proportions from Landsat against the modern oblique images, hypothesizing a linear relationship for most classes. We found that the two sources sampled the landscape in broadly similar ways, with near‐concordance for dominant land cover classes, yet that oblique photographs more frequently detected narrow landscape features and estimated higher proportions of rock compared to satellite imagery, possibly due to the higher spatial resolution of the oblique photographs, and to their angle of incidence against steep slopes. We then evaluated land cover change from corresponding historical and repeat photographs and found that the landscape has homogenized over the past century via increased coniferous forest cover. Our work shows that terrestrial oblique photographs can be used to estimate landscape composition, particularly in mountain environments. This is helpful for analyzing past landscape conditions in historical photographs, monitoring decadal‐span landscape change and assessing habitat to model biodiversity through time.

Keywords