For my study I have six different DEMs of a glacier. The DEMs are all from different years and I want to calculate the volume changes over a timeperiod. Including the six DEMs I have one subglacial DEM with low resolution (100 meter). In ArcMap, I use raster calculator to subtract the DEMs from the subglacial DEMs to get the volume results. The problem is that the DEMs with high resolution gets converted to 100 meter resolution. Is there any way to derive results with better resolution?
Generally speaking you maintain the lower resolution that you have during an overlay (or other gis operation)
Thank you for answering.
Yes, I know. One thought though; when producing pixelvalues after subtracting the subglacial DEM with the other DEMs, I multiply the cell length and width (100 m) to convert each pixel into a volume. Then I use the Zonal Statistics tool in ArcMap to sum up the volume across all of the glacier. Regardless, is it possible to calculate systematic errors of the glacier volume series I have in my investigation? Any suggestions?
Thanks
Another thought I have.
Do you roughly know the thickness of the ice sheet? And how rough is the subglacial DEM and the surface of the ice?
Maybe it would be useful to take the 3D surface into account. There are tools available in ArcMap (or extensions) to calculate first the surface of each cell based on the topography. During the volume calculation you multiply then the surface of the cell by the thickness of the ice instead of using 100x100m.
This approach could be more accurate depending of the terrain roughness and the relation between thickness and cell size.
I agree with Matthias Hunziker in that the terrain roughness has a huge influence. Consider the following: If you are multiplying the cell length and width you get a planimetric area. It is not the real area since you are neglecting the influence of the slope. A simple (but overestimating) way is to divide the area by cos(slope) to get the surface area.
Check the literature, too. I am sure there are a lot of papers dealing with your problem (especially error calculations on different sources of DEMs).
For the surface area calculation, you can use the algorithm from Jenness, for example. It calculates the surface of the eight segments divided by the two perpendicular bisectors and the two diagonals.
http://www.bioone.org/doi/abs/10.2193/0091-7648(2004)032%5B0829:CLSAFD%5D2.0.CO%3B2
Hi Marius,
isn't it possible to artificially increase the resolution of the subglacial DEM to the resolution of the other DEMs using the resample tool? When doing so, it's probalby a good idea to go to the "Environment Setting" where you can choose the other DEM as "Snap Raster" in the "Processing Extent" options; this will align the raster cells properly to each other.
This should not add any new data but the raster calculator should then be able to obtain more detailed results.
Best,
David
- Badges
- Science topic
- Similar topics
- Geoscience
- Geoinformatics