For modeling purposes, I'm currently creating a list of potential explanatory variables to map the probability of my event at a very large scale, and at a coarse spatial resolution (1 to 50km pixel side, not fixed yet). As fragmentation is important to understand how a natural event spread in the landscape, and by extension in a large area (ecoregion for example), is it relevant to compute some metrics and to upscale them at a coarser resolution?
For instance, computing a connectivity index of streams at 30m resolution and upscale a pack of 30m pixels at a 1km single pixel and average input values.
Just to be clear, I don't want to compute landscape metrics at coarse resolution (that is definitely not relevant!).
It depends on your metric and why you are measuring fragmentation/connectivity. Are measuring structural pattern for descriptive/comparative purposes? Or are you linking to a functional response (ecological process, animal movement)? Scale will be more impotant in the latter case. Some metrics are more scale sensitive that others. The DCI for stream connectivity is designed to be scale-independent but we've never tried calculating it at a fine resolution and then saling it up through averaging. I'd be interested to see how that works. For a good paper that outlines issues with metrics and scale see Lu et al. 2013 Physical Geography 34 (1) , pp. 40-49
Scaling behaviours can be highly variable too - there are some characteristic patterns in change behaviour depending on the spatial resolution of the data and when certain thresholds are crossed. There are some papers on this, including one I co-wrote using very fine spatial resolution (Corry and Lafortezza, 2007).
There two papers that may provide some useful information to your question. Andrew Millington did a paper upscaling landscape metrics in tropical forests ( Millington AC, Velez-Liendo XM, Bradley AV (2003) Scale dependence in multitemporal mapping of forest fragmentation in Bolivia: implications for explaining temporal trends in landscape ecology and applications to biodiversity conservation. ISPRS Journal of Photogrammetry & remote sensing 57: 289-299). There is paper we published in 2009 that explores how different scales affect landscape metrics for species with limted range, yet widely distributed (Perotto-Baldivieso et al. 2009).
Article Spatial distribution, connectivity, and the influence of sca...
https://www.researchgate.net/publication/229826816_Empirical_patterns_of_the_influence_of_Spatial_Resolution_of_Remote_Sensing_Data_on_Landscape_Metrics
https://www.researchgate.net/publication/230738894_Peri-Urban_to_Urban_Landscape_Patterns_Elucidation_through_Spatial_Metrics
https://www.researchgate.net/publication/301609558_Geospatial_analysis_of_forest_fragmentation_in_Uttara_Kannada_District_India
Article Empirical patterns of the influence of Spatial Resolution of...
Article Peri-Urban to Urban Landscape Patterns Elucidation through S...
Article Geospatial analysis of forest fragmentation in Uttara Kannad...
Article Geospatial analysis of forest fragmentation in Uttara Kannad...
- Badges
- Science topic
- Similar topics
- Biological Science
- Bioecology
- Systems Ecology
- Landscape Ecology