Help with georeferenced point file importing where it should be

[PNWPaul]

Would anyone have tips on how to get a geo-referenced .las point file to import into its correct position? I have downloaded the file from the Washington State Lidar portal, and it is in Washington South (same coordinate as if it was downloaded from USGS). I set my Vectorworks file to the exact same coordinate system

EPSG:4152

NAD83(HARN)

 It imports it nowhere near the proper insertion point (not in the same country or even on the map). The same file imported into ArcGIS Pro lands precisely where it should be. Does anyone have any pointers as to what I'm doing wrong?

I have attached a Google Drive link with the .las file for reference and scene shots from Arc with the .las files coordinate system information and a picture of where it should end up.

I have tried importing in both mill and feet, neither with success; the base file is in US survey Feet.

Any help would be greatly appreciated.

I do a lot of concept design, and being able to create site models with accurately placed lidar data is critical. I'm sure I must have something wrong on my end when I import.

Thanks,

Paul

 

https://drive.google.com/drive/folders/1HVY0p959jP2O_hHDf-0ZQ9xFxG3-XTOn?usp=sharing

[Jeff Prince]

9 minutes ago, PNWPaul said:

Does anyone have any pointers as to what I'm doing wrong?

 

You are not doing anything wrong, Vectorworks has a bunch of issues importing LiDAR that have been going on for years unfortunately.

The work around to positioning the file correctly is to note the bounding coordinates of the LiDAR tile and then move it manually.

 

Yes, it's silly we still have to do this compared to just about every other software that imports georeferenced LiDAR.

[Vlado]

@PNWPaul It looks like EPSG:4152 is not the correct coordinate system, as it looks like this is only spherical coordinate system. The images you provide shows a projection added to it, but it is not clear what is the EPSG if that projection.

 

I managed to pull the data out from the image and setup this WKT that is based on EPSG:4152 spheroid plus the projection parameters from your images. See below for the text of it.

 

However, I'm still not able to align your file with it, as it's not clear what units the LIDAR file is using. Yes, it should be automatic, but it should be possible to be done manually. I've tried several settings but shouldn't guess it.

 

Do you know the latitude and longitude of the location where the file is supposed to be? That can give us a clue what are the correct unit configuration.

 

Regards,

Vlado

 

PROJCS["custom",
    GEOGCS["NAD83(HARN)",
        DATUM["NAD83_High_Accuracy_Reference_Network",
            SPHEROID["GRS 1980",6378137,298.257222101,
                AUTHORITY["EPSG","7019"]],
            AUTHORITY["EPSG","6152"]],
        PRIMEM["Greenwich",0,
            AUTHORITY["EPSG","8901"]],
        UNIT["degree",0.0174532925199433,
            AUTHORITY["EPSG","9122"]],
        AUTHORITY["EPSG","4152"]],
    PROJECTION["Lambert Conformal Conic"],
    PARAMETER["latitude_of_origin",45.33333333333334],
    PARAMETER["central_meridian",-120.5],
    PARAMETER["standard_parallel_1",45.83333333333334],
    PARAMETER["standard_parallel_2",47.83333333333334],
    PARAMETER["false_easting",1640416.667],
    PARAMETER["false_northing",0],
    UNIT["foot",0.3048,
        AUTHORITY["EPSG","9002"]],
    AXIS["Easting",EAST],
    AXIS["Northing",NORTH]]

 

[PNWPaul]

hi Vlado, 

Here is the lat lon that arc shows for the north West corner of the las bounding box

[Jeff Prince]

2 hours ago, Vlado said:

Do you know the latitude and longitude of the location where the file is supposed to be? That can give us a clue what are the correct unit configuration.

 

The center of the subject area appears to be at or very near 

48.906562°, -122.693548°

[Vlado]

@PNWPaul I was able to import it somewhat where it is supposed to go, but unfortunately it is not perfectly fitting. And I'm not sure what's wrong. Anyway, here is what I did:

ok, so these are called cartesian coordinates (i.e. easing and northing).
1110896.65E
1313006.9N
Note they are measured in linear coordinates, where lat/long are measured in degrees.

 

After I setup the Vectorworks drawing geo-referencing with the WKT above, these cartesian coordinates gave me:
-48.91273585 lat
-122.070105424 lon

These checks are very easily done in Vectorworks using the Stake tool, and the .Set Geolocation" button.
Then with the Geoimage, it all seems to line up to your image from the original post:

 

Then I was able to import the .las file, making sure the center checkbox is off, and select Feet for units as this coordinate system uses feet (note, it doesn't really matter what units the Vectorworks document is in)

 

Then the image didn't fit at all, it was way to far, and comparing the location of where it should be to where it is (in cartesian coordinates) it seems like it's off by scale of 1000.

 

So then I just scaled the Point Cloud object by factor of 0.001:

 

And that put it at the stake point, but it is still slightly off from the geo-image background:

 

 

I have no idea why this is happening. I'll keep poking it, and see if something comes out.

[Vlado]

... scaling problems are typically due to incorrect coordinate system, but the WKT looks the same.

Maybe you can getting more points and plotting them in Vectorworks to see how they relate to the imported image. This will give you more confidence that it's not the WKT.

 

Another way might be if you export a shapefile with prj file, and then import it in Vectorworks in a new file [so that it sets the Georeferencing]

[PNWPaul]

What I am going to try to do is set the coordinate system on vectoworks, and then set the file in ArcGis to that coordinate system and see if it will transform properly. When I share as a web layer in arcgis and then bind that layer to the vectorworks design layer it appears where it should. Unfortunately I can not use this as a base for creating a site model with as the integration does not share polling z data 

[PNWPaul]

Well, there seems to be no way I could get it to land where needed, which is a big downside. However, I did find a workaround for anyone who uses ArgGis. I created a contour file of the site in Arc and hosted that on a web layer. Then I used the ArcGis integration in vectorworks to bind that layer to the drawing. This creates 2D polys with contour height data in their records. I modify by record to turn them into 3D data and create a site model for that. It's a lot of extra steps, and Vectorworks should be able to read the data from the point cloud in the first place and set the coordinate system of the design to that. 

I am a new user of vector works, and I believe there must be a way for this to be accomplished as it can be done easily in Civil 3D. 

However, this workaround will work fine for now. 

Thank you guys for your help!

[Benson Shaw]

OK, Glad this is being addressed by @Vlado and others. The LIDAR surveys and subsequent contour shapefiles, point cloud LAS/LAZ files and xyz data files for this region of the USA are collected and created by the Pacific Northwest Lidar Consortium. These files are hosted and available for download at the NOAA Data Viewer Portal website: https://coast.noaa.gov/dataviewer 

Perhaps some of the info below about the PNLC data could help sort this. And I present a workaround.

 

PNLC "About" page:

http://pugetsoundlidar.ess.washington.edu/About_PSLC.htm

 

PNLC Tech Spec document - collection and data processing info (Projection, etc):

http://pugetsoundlidar.ess.washington.edu/Technical_Specifications_withAppendices.pdf

 

LAZ files from https://coast.noaa.gov/dataviewer imported into vwx2023

Same files, either new download or reused from local OS folders refuse to import into my vwx2024 - Corrupt data notice.  Try with the attached zips of folders from NOAA. One is .shp with contours (imports fine). The other is LAZ.

 

Here is a workaround for vwx2023:

To correctly fit the point cloud data into the Vectorworks drawing:

On the NOAA site, determine an Area of Interest and, in a series of requests, use the same capture rectangle (the website preserves it until you log out or redraw it) and receive via email the LAS, LAZ, Contours shp. May as well grab the XYZ, too while you are there:

 

0. Collect Contour and Point data from NOAA Data Viewer>Elevation.  Click the "Return to Viewer" button after each Cart entry to use same Area of Interest outline for each request. https://coast.noaa.gov/dataviewer

1. Import the contour .shp to vwx This set of contours is used as the standard for the point cloud. Contours port in at proper scale and location whether imported to a prior georef vwx, or by letting the .shp determine the georef. Adjust the vwx internal and user origins to suit, and geolocate an image to verify that all is in order.

2. Import the LAZ or LAS. Uncheck the Center on Import option.  Point Cloud will locate far from the contours. Group it and move it to, or near the contours via OIP. This makes both visible/selectable on screen. Separate layers is best, Show/Snap Others.

3. The scale will be wrong no matter the layer scale or which import scale. Make a ratio of the x value of the contour and point cloud groups.

4. Rescale the point cloud by the ratio of the groups. Point cloud will jump away to a new location.

5. Move the point cloud group near to the contour group (again). eg zero out the OIP location values of the Point Cloud group.

6. Lock the contour group so it cannot move.  In any view except, use the 3d Align command to center the point cloud group on the contour group.

7. Verify that the point cloud is correct scale and location:  Superimpose a DTM from the point cloud on a DTM made from the contours. They will be quite close.

 

I made a video a couple in vwx v2023:

https://www.dropbox.com/scl/fi/vd7x044kscx1t5l33hmz7/Vectorworks-GIS-2.mov?rlkey=kqejswd63jac1bjutlqjer4zg&dl=0

 

An alternative control is to use the ASCII XYZ file instead of the contour .shp file. Or, just use the XYZ data to make the DTM.

 

As noted by others, these LAS and LAZ files import correctly to Q and Arc, so wondering if vwx needs an adjust? I think, in general, LAZ and LAZ files port in correctly??? Perhaps these PNLC files are uniquely problematic for vwx?

 

Attached find a .shp and a .laz (GeoID 18) collected today and describing an area in the northwest Seattle, WA, USA.

Contours import as expected.  LAZ does not import into vwx2024. (My 2023 is expired, so not able to test)

Also tried a LAZ in GeoID09. Did not import.

 

-B

 

 

 

 

 

wa2016_pslc_king_Job978312.zip wa2016_pslc_king_Job978313.zip

Edited March 3 by Benson Shaw