Contour line issues

[OvidiuFrantescu]

Hi all,

I am working with a raster file (.tif) that I have georeferenced, and now I would like to extract the contour lines of of it.

After using the contour tool I am left with multiple lines representing a single contour line from the raster.

Basically the tool does not recognize only the center-line of the contour line from the raster to generate an unique line (see attached image).

How can I condense these lines, or generate a single (center-line) one.

Thank you,

Ovidiu

[DanPatterson_Retired]

what coordinate system is the data in?  not the dataframe but the actual data?

what contour interval did you use?

What is the cell size?

I suspect you are contouring at too fine an interval relative to your cell size and the range of values in the raster

[OvidiuFrantescu]

Hi Dan,

The data is in WGS 1984 Web Mercator. 

In the contour tool I selected the interval to be 1, because the original contours are at 1 ft. intervals.

The cell size is X/Y 0.013888889, 0.013888889.

Now that you mention these, I am thinking that I should do a raster re-sampling to increase the cell size.

Any advice is greatly appreciated.

Ovidiu

[DanPatterson_Retired]

those cell sizes look like decimal degrees and I suspect the data may be in GCS WGS 1984 ... which is unprojected data

[SteveLynch]

what is the min/max of the input raster and how many columns/rows does it have?

[OvidiuFrantescu]

Here is a copy/paste set of details regarding the raster file:

<Columns_and_Rows> 30001, 28766

<Number_of_Bands> 3

<Cell_Size__X._Y_> 0.013888889, 0.013888889

<Uncompressed_Size> 2.41 GB

<Format>TIFF

<Source_Type>Generic

<Pixel_Type>unsigned integer

<Pixel_Depth>8 Bit

<Colormap>absent

<Pyramids>level: 7, resampling: Nearest Neighbor

<Compression>LZW

<Mensuration_Capabilities>Basic

<Status>Permanent

<Extent><Top> 399.527777778 <Left> 0 <Right> 416.680555556 <Bottom> 0 <Spatial_Reference>WGS_1984_Web_Mercator_Auxiliary_Sphere

<Linear_Unit>Meter (1.000000)

<Angular_Unit>Degree (0.0174532925199433)

<False_Easting>0 <False_Northing> 0 <Central_Meridian> 0 <Standard_Parallel_1> 0 <Auxiliary_Sphere_Type> 0 <Datum>D_WGS_1984

<Statistics><Band_1><Build_Parameters>skipped columns:1, rows:1, ignored value(s): <Min>0 <Max>255 <Mean>253.5990525616515 <Std_dev.>14.24670552247798 <Classes>0

<Band_2><Build_Parameters>skipped columns:1, rows:1, ignored value(s): </Build_Parameters><Min>6<Max>255<Mean>253.6721133398082<Std_dev.>9.882874895876389<Classes>0

<Band_3><Build_Parameters>skipped columns:1, rows:1, ignored value(s): </Build_Parameters><Min>6<Max>255<Mean>250.7431879087078<Std_dev.>24.38898611239873<Classes>0

Thank you for any input you might have.

[SteveLynch]

This is a multi band image and only the first band will be contoured. It has values between 0 and 255

[curtvprice]

If it is a scanned map with contours on it the workflow you should use is to first georeference the image (preferably to match the coordinate system of the map) and then use ArcMap digitizing to capture the contours using 'heads-up' digitizing, that is, using ArcMap edit tools with your scanned map in the background, to digitize the contours with your mouse. This can be fairly time consuming but if the contours have any complexity this is probably the best way to capture the data in the long run.

[OvidiuFrantescu]

@Curtis Price - this is what I have done, but since this map covers an 18 mile long lake, and the contour lines are at 1 ft intervals it will take a long time to do by hand. Any advice on how to solve the problem automatically, or at least to reduce my work time?

[DanPatterson_Retired]

If the contours were derived from the raster... then they aren't representative of the data.  The raster cell size is much larger than the contour spacing so you have spurious precision in your data.  

If the raster coordinates are in decimal degrees...which still remains unanswered... It should be projected to an appropriate coordinate system.  If you have to stick with feet for some reason make sure that the projected coordinate system has X,Y and Z in feet of if X and Y are in meters, make sure you account for the Z scaling if the elevations are in feet.

Once the data are projected using an appropriate cell size (ie. the foot equivalent of the cell sizes you report which are not in feet for sure), then you can use the contour tool.  The contour spacing should not be the same size as your cell size, otherwise, you will get the same artifact that you are currently seeing.