Tree Line Change Detection Using Historical Hexagon Mapping Camera Imagery and Google Earth Data

The role of spatial data and geomatic approaches in treeline mapping: a review of methods and limitations

“…elevation, slope, aspect, etc. ), that are useful to characterize treeline ecotones, can be easily obtained from different available DTMs (Digital Terrain Models) [Stueve et al, 2009;Kharuk, 2010;Groen, 2012]. Pixel-based [Tattoni et al, 2010] and object-oriented [Sitko et al, 2008] classification approaches, both supervised [Bader et al, 2008;Kral, 2009] and unsupervised [Klasner et al, 2002;, of multispectral aerial and satellite imagery, are also used.…”

“…Some studies base mapping on screen image-interpretation [Zhang et al, 2009]. Ground surveys, mainly performed by GNSS (Global Navigation Satellite Systems) [Groen et al, 2012;Ropars et al, 2012] are widely used to collect ground truth samples that are in turn used to train classifiers and validate classification results. Essentially, almost all subdisciplines of Geomatics can effectively support treeline detection.…”

A Review of Methods Used to Measure Treeline Migration and Their Application

“…While most satellite records only stretch a few decades back in time, there are instances where high-quality satellite imagery have been captured by some of the earliest satellites placed in orbit. One of these data sources is declassified spysatellite imagery, which for example was utilised by Groen et al (2012) to track treeline changes along the border of Macedonia and Bulgaria using images captured by the United States Hexagon mission in 1976.…”

“…A common characteristic of remotely sensed data used to map treeline position is high spatial resolution, as the spatial resolution of the dataset significantly influence the accuracy of the treeline mapping. In order to map changes that are occurring to individual trees, spatial resolutions often need to range from 0.5 to 4 m (Groen et al, 2012;Luo and Dai, 2013;Mathisen et al, 2014). There are several earth-orbiting satellites which are capable at offering data with adequate spatial resolution, including Quickbird (0.6 ~ 2.4 m resolution) (Hofgaard et al, 2013;Luo and Dai, 2013;Mathisen et al, 2014), WorldView (0.5 m resolution) (Mathisen et al, 2014), and Google Earth (1.8 m resolution) (Groen et al, 2012).…”

“…In order to map changes that are occurring to individual trees, spatial resolutions often need to range from 0.5 to 4 m (Groen et al, 2012;Luo and Dai, 2013;Mathisen et al, 2014). There are several earth-orbiting satellites which are capable at offering data with adequate spatial resolution, including Quickbird (0.6 ~ 2.4 m resolution) (Hofgaard et al, 2013;Luo and Dai, 2013;Mathisen et al, 2014), WorldView (0.5 m resolution) (Mathisen et al, 2014), and Google Earth (1.8 m resolution) (Groen et al, 2012). Some earth-orbiting satellites which provide a lower spatial resolution, but greater temporal coverage include Landsat MSS (60 m resolution) and Landsat TM (30 m resolution) (Hofgaard et al, 2013).…”