I have a file of points and lines with elevation data (z value) and I want to make an accurate Digital Elevation Model (DEM) using QGIS. The software allows me to make only a DEM with 12m cell size, but I want to make a more accurate elevation model, with 1m or 2m cell size.
Furthermore, the software allows to manage separately the points file and the lines file, while I need to consider them together in processing of the DEM.
How I can do? Is there a specific tool or procedure for creating DEM with other cell size?
Dear Marco,
Just a remark before you start:
Accuracy and resolution are two different things (though sometimes dependent). If you want to increase the resolution of your DTM (I guess this is rather a DTM as you mentioned contour lines) that is fine, but beyond a certain limit this will not increase the accuracy, therefore it does not make too much sense. There is always a trade-off. The proper choice of resolution is a key issue, and it depends on the (spatial) density of your input data. (There are some nice papers on that.)
Consider the following: now you have a grid of 12 m resolution. To get a 1 m resolution grid, you need 144 times more information! Normally, you don't have that.
To achieve an accuracy that is congruent with 1 m or 2 m resolution you should have at least ~1 point/m2 input data density and a contour interval of at least 1 m (or better). Considering a common contour line dataset (1:25000 or 1:10000 scale) and 2.5 m contour interval, the ~10 m resolution is typically optimal. (If you have special input like contours of an archaeological site with 10-25 cm (!) contour interval, that would be perhaps suitable to compute a 1-2 m resolution grid. In most of the cases for a 1 m resolution grid you need LiDAR data instead of contour lines.)
If you have to model a (very) low relief area like a floodplain out of contour lines, further special considerations should be taken.
I apologize if I misunderstood your problem.
Kind regards, Balázs
Dear Marco,
Just a remark before you start:
Accuracy and resolution are two different things (though sometimes dependent). If you want to increase the resolution of your DTM (I guess this is rather a DTM as you mentioned contour lines) that is fine, but beyond a certain limit this will not increase the accuracy, therefore it does not make too much sense. There is always a trade-off. The proper choice of resolution is a key issue, and it depends on the (spatial) density of your input data. (There are some nice papers on that.)
Consider the following: now you have a grid of 12 m resolution. To get a 1 m resolution grid, you need 144 times more information! Normally, you don't have that.
To achieve an accuracy that is congruent with 1 m or 2 m resolution you should have at least ~1 point/m2 input data density and a contour interval of at least 1 m (or better). Considering a common contour line dataset (1:25000 or 1:10000 scale) and 2.5 m contour interval, the ~10 m resolution is typically optimal. (If you have special input like contours of an archaeological site with 10-25 cm (!) contour interval, that would be perhaps suitable to compute a 1-2 m resolution grid. In most of the cases for a 1 m resolution grid you need LiDAR data instead of contour lines.)
If you have to model a (very) low relief area like a floodplain out of contour lines, further special considerations should be taken.
I apologize if I misunderstood your problem.
Kind regards, Balázs
Dear Balazs
Thank you for your answer. Obviously I fully agree with your considerations on accuracy and resolution of DEM.
The problem is not the accuracy. I have a topographic data with very high resolution, with 1 or 2 m contour lines (1:1000 or 1:2000), sometimes even more detailed. Generally I work for railways, highways and dams, so the topographic survey are expressly made with high detail.
The problem is not the input data (which are congruent with 1 or 2 m cell size), but the software processing that does not allow me to change cell size of DEM (or rather, I don't know how it can be changed).
Would be interesting which tool in QGIS you were using so far and which parameters you chose. Otherwise we cannot suggest alternatives :)
Hi Andreas,
I used the command "interpolation" from menu "Raster/analysis/interpolation". I don't know if there is other special tool to make DEM.
Regarding the parameters, I entered those required by the software.
Hey! If I understood - You want to change the resolution of your result DEM? In QGIS (2.18.15) in menu: Raster->Interpolation->Interpolation You'll see dialog window and on th eleft You have to add point/line layer (input data), and on the right You can choose interpolation method, output file and just resolution (cellsize x and cellsize y)
Hi Bartlomiej.
Previously I tried to interpolate the data with another tool (Raster/analysis/interpolation) and with this I had the problems described in the question.
Then I tried with the tool that you suggested (Raster/interpolation/interpolation) and the creation of DEM works very well. With this application I have only one problem, I can't load file of 3D polilines because the Z value is not contained in the attribute table (while the tool requires that the Z is a part of the table).
Do you know a method to analyze this type of file? Should I convert lines into points?
Hey! Maybe try convert 3D-polyline to points (in menu: Vector-> Geometry Tools-> Extract nodes), but I'm afraid it will not work. First - You have to convert Your 3D shapefile to 2D shapefile
- Badges
- Science topic
- Similar topics
- Geoscience
- Geoinformatics
- Geographic Information System