matt on 19 Jan, 2019 10:15 PM
For various reasons, I am not in a position to incorporate piecemeal elevation data from disparate sources into a single coherent layer. It's a tremendous amount of work, but I also want to note that I can't just plunk in a new dataset as there's also a significant amount of vetting required - for example, while the data may work great above treeline, a lot of LIDAR datasets pickup on trees and produce very noisy results below treeline (Google's Terrain layer being an example in many locations).
I don't think it's really accurate to say that the accident was attributed in part to inaccurate slope data. Not understanding the limitations / micro-interpretation of a dataset beyond its actual resolution, yes. Higher resolution data certainly wouldn't hurt but it also wouldn't solve the problem entirely, e.g. cornices forming along ridges and dramatically changing the surface slope angle vs summertime.
I am considering ways to to make the existing slope angle shading layer's limitations more visually apparent.
I was not trying to imply in any way that this website bears any responsibility for this accident (if that was how my comment was received). Users should be aware of the limitations of these tools. From personal experience though, it is tempting to try to use this type of data to predict the avalanche danger on very specific slopes and terrain features.
I do like the idea of making the limitations of the slope angle layer more apparent though. This can still be a very useful tool for planning trips into avalanche terrain, as users we just need to be smart about how we use it. The more we know about its limitations, the better.
matt on 22 Jan, 2019 06:00 PM
I didn't interpret it that way, I just see a difference between being inaccurate and just being low resolution and this is a case of the latter not the former. Although there are places where the NED is probably flat-out wrong as well, I think addressing those two issues requires different approaches.