The study area for this assignment is a 700 meter long reach of a braided river located in the Cairngorm Mountains of Scotland. In this exercise, I used several sets of 1-meter riverbed elevation data to detect erosion and deposition between data capture events.
"The GCD software was developed primarily for topographic change detection in rivers, but will work for simple, raster-based change detection of any two surfaces. The volumetric change in storage is calculated from the difference in surface elevations from digital elevation models (DEMs) derived from repeat topographic surveys.
As each DEM has an uncertain surface representation (which might vary in space and time), our ability to detect changes between surveys is highly dependent on surface representation uncertainties inherent in the individual DEMs. The fundamental problem is separating out the changes between the surveys that are due to topographic change as opposed to noise in the survey data. GCD provides a suite of tools for quantifying those uncertainties independently in each DEM and propagating them through to the DEM of difference. The program also provides ways for segregating the best estimates of change spatially using different types of masks. The overall suite of tools is more generically applicable to many different spatial raster-based change detection problems."
The change between 2005 and 2006 (1400 - 4500 cubic meters) was lower volume than the change between 2003 and 2006 (7100 - 10700 cubic meters). Overall, the system appears to be degrading, with more erosion than deposition for both analyses.
When delineating zones of bar development, I looked for distinctive bar shapes in the DEM that were highlighted in a stronger blue color.
When delineating zones of channel bed rising, I looked for the absence of distinct bar forms in areas that were highlighted in a pale blue color.
When delineating zones of channel bed lowering, I looked for areas highlighted in pale red where the change between the earlier and later DEM did not include retreat of a defined bank.
When delineating zones of bank erosion, I looked for areas highlighted in dark red where the change between the earlier and later DEM included the retreat of a defined bank.