A question about DEM generated terrain

[Jananton]

Hello All,

After building some fantasy routes and some content, I thought to have a go at some "real" terrain. I went through the enourmous list of filipsatwork and found a medium size route that I will attempt. It doesn't exist anymore, so how prototypical it gets is the question, allthough I found an interesting website about it. I presume that the vegetation will not have changed that much in 60 to 90 years so that will be google earth to the rescue. Situated partly in a national park these days will in this case help I suppose.:hehe:

And now for the question. I know that rivers are indicated by dark blue lines, or, if they're wider, dark blue borders. What I would like to know however, is the depth of the river also modulated, or is the surface of the water the surface of the map. In other words, do I need to dig those rivers out myself, so to speak, in the generated terrain.

If anybody has any other hints for the budding surveyor don't hesitate to share them here.

Greetings from nighttime Amsterdam,

Jan

 

[gfisher]

It has been my experience that the DEM indicates the approximate surface of the water level of the river, so you have to dig them yourself.

 

[davesnow]

Just don't dig them very deep. Since you can see all the way to the bottom, deep rivers don't look very realistic. I try to make mine as shallow as possible, then still give them a dark bottom so the eye will be fooled.

 

[cascaderailroad]

Treetops in forests, and tall urban buildings throw off the DEM satelite data heights. DEM's can be off in all of the x-y-z axis +/-20' in all dirrections. Inland small rivers usually are usually shallow, @25-50' deep. Gradients of trackage are generally less than 2.5% max, and 1.25% is a rather steep grade that the PRR found as a max acceptable incline. Trainz measures heights in the metric system. The Horseshoe Curve is @ a 250 m radius, which is an acceptable average curvature for mainline traffic. You can dig rivers by using track, and lowering it, and pressing the "smooth spline" tool button.

Since you are from Europe, I am taking for granted that you use the metric system, and don't need this english/metric equivilent chart:
http://forums.auran.com/trainz/showthread.php?t=44283&highlight=metric/english
Hopefully it will give you some pointers.

 

[angelah]

DEMs

I use DEMs a lot and find they can be out by a considerable amount depending on what is in the surrounding terrain. For instance if the area is fairly flat then you stand a chance of being reasonably accurate, but if you're working in certain areas of say, the UK, where there are deep gorges or large rivers with hills nearby, then expect to have to sort out problems.

The best way is to use good maps that show contour lines, such as Ordnance Survey (OS) Landranger ones. It depends where you are of course, but wherever it is it is likely that good maps will be produced.
This way you can realistically check gradient and land heights all along your route and adjust the DEM accordingly.

It is exactly what I am doing around Frome and Bruton, England, where there is a steep gradient, hills, deep valleys and slops, all bad news for an accurate DEM!

Angela

 

[Jananton]

Hello everybody,

First thanks for all the hints and tipe, these will come in handy when work progresses.
From the remarks on accuracy aspects of DEM maps several of you made, I think I haven't choosen the easiest one, but from the long list I went through this was the one that spoke most to my imagination. :hehe:

@cascaderailroad; Talking about accepteble curve radii, I presume this also depends on the era one wants to depict. Wat would be an acceptable radius for normal gauge passenger and lumber/log traffic round 1910 / 1930?
And yes, since Napoleon we're all metric over here.

@angelah; I indeed allready noticed that even in seemingly flat terrain "flat" isn't as flat as one expects.:hehe: Depending on wich side you start offcourse, the route roams from quite flat into a mountainous gorge area along a river, so all what you mentioned will be the case, I suppose.
Luckely the website I mentioned does provide some maps, a milage table between stations and a height profile. I'll have a look at the map manufactorers you mentioned, since that will probably be a must when reaching the mountains.

I think I will start first by plotting the stations en route to see what a rolercoaster ride this results in.:hehe:

Greetings from overcast Amsterdam,

Jan