SRTM 1-arcsecond global

[lomar]

https://www.usgs.gov/centers/eros/science/usgs-eros-archive-digital-elevation-shuttle-radar-topography-mission-srtm-1-arc?qt-science_center_objects=0#qt-science_center_objects

so It seems that recently(at least it's more recent than the WS2.0 release) USGS has released a global SRTM dataset with 1-arcsecond resolution(previous data was 1-arcsecond for USA only and 3-arcsecond for the rest of the world apparently)

so, is it of any relevance for us? can this new elevation data be used for building scenery for FG or would there be any issues(like licensing or bugs)

[frtps]

I tried out this for the Sydney (Australia) area and was disappointed that the central business district was very "peaky" presumably due to the tall buildings. I've gone back to SRTM-3 as it somewhat smooths out the topography in these areas. I've always wondered how this is dealt with in other city centres built using SRTM-1.

[wlbragg]

Wow, that's great. But really the difference between the 1 VS 3 isn't that much and If there is no editing of man made structures spikes then it may not work well That is why the "Panoramic" site data is used for the global scenery build, because they have cleaned up a lot of those issues.
I suppose if someone was really motivated to learn how to clean this data up, it would be cool to be able to use 1arc-sec for the entire world.

[xDraconian]

SRTMGL1 has been used for the Hawaiian Islands and Tennessee - not sure what has been used for Iceland, but likely SRTM-3.

I consider SRTMGL1 the preferred data source as there are no void areas. The void areas having been supplied from other non-proprietary data sources.
1-arc-second or better resolution requires careful decimation to reduce the vertices count to something reasonable. We do the same for 3-arc-second data sources, but it's even more important with the higher-resolution sources.

The issue with "peaky" terrain, most often found in urban areas, is often due to Digital Surface Model (DSM) data being utilized to fill the voids. A DSM represents the surface (including man-made objects - e.g. buildings). These areas can be flattened in GIS software.

[legoboyvdlp]

Viewfinder panorama is far far superior for our purposes (unless you have the skills to edit the data!) since it has been cleaned. The main advantage of SRTM-1 is when mountains etc are where it helps to have a hi-res data source.

Are those proprietary data sources compatible with GPL? Is the finished product still public domain?

It seems that terrafit has trouble with SRTM-3 on the next branch as it makes large 40000 foot spikes etc So using SRTM-1 or an older version would seem to be required at least on my PC. I will try and investigate the issues...

[vnts]

Viewfinder panorama is far far superior for our purposes (unless you have the skills to edit the data!) since it has been cleaned.

Would taking clean urban areas from SRTM-3 (giving them priority to override SRTMGL1 data) and the rest from SRTMGL1 be a shortcut that avoids cleaning? ..or is that also difficult

Kind regards

[legoboyvdlp]

I'm not so sure if the toolchain supports that, if it does it certainly would make sense to do that.

[xDraconian]

The toolchain does not support merging of different formats. A potential solution would be to pass a 'flattening of abrupt elevation changes' parameter for urban/town areas. A more flexible solution would be to also support a flattening shapefile defining a mask to be used to limit where the flattening was allowed.