Comparing Global DEM with Lidar

This is currently replaced with DEMIX wine contest.  Lidar may be incorporated there "soon"

Compare Point clouds to global 1" DEMs (and potentially 3" DEMs).  This documentation is not completely up to date

• Get Key Global Regional DEMs.  This is geared to work with SRTM, NASA, ALOS, ASTER, and COP[ernicus] 1" data
  • Index in c:\mapdata\indexed_data\dems\ with series names that are less than 6 characters (these will be used for database field names and need space for the Z or SLP)
• Lidar or ICESat-2  photon data point cloud
  • Input data
    • ICESat-2 photon data
      • ICESat-2 photon data to LAS (doing this limits the output to the desired confidence levels)
      • You must have "icesat2" in the file name to be able to filter the data by beam and confidence.
    • Icesat 2 elevation (terrain and CHM) data. 
      • Very sparse coverage compared to photon data converted to point clouds. 
      • Much less noisy compared to photon data converted to point clouds. 
    • Lidar point cloud already in LAS
  • LIDAR Point clouds Open LAS file on DEM
  • Create a point density map for the LAS file at 1" spacing (or other spacing of the DEM), name it "point_density.dem" and put in the same directory as the DEM
  • Pick the Slice option on the LIDAR Point clouds
  • For ICEsat-2, Set options (which beams to use, the minimum confidence for the point to use it, but you might already have limited the confidence values that were exported).
  • Pick Canopy button option on Slice window control form
  • Requires SRTM, ASTER, and ALOS DEMs
  • The program can automatically open them and gets the elevations where needed
• Analysis of the point cloud data
  • Filter to remove very large CLOUD_HT, or very high values of FR_SRTM or other DEM
  • Double click in a record
    • You must still have the point cloud open
    • Select database with results of the Canopy button
    • Filter to remove very large CLOUD_HT or large FR_xxxxDEM values
    • Double click on one of the records
      • Point cloud statistics
      • Point cloud points: creates a DBF with all points in a 1"x1" box, and see if the problems are with one track, individual beams, confidence levels, or the GPS Time.

• Analysis of the comparision statistics with the results of the Canopy button in a DBF file, Lidar or ICESat-2  photon or elevation data data.  For ICESat-2 you might want to filter on date, beam, and track_id.
  • Stats, Lidar to analyze global DEMs
    • DEMs within point cloud range. Filter for one track, and one beam.
      • Lat profile, current filter (generally the better choice because of the orbital track geomtry)
      • Long profile, current filter
    • Distribution histograms (must be from a point cloud)
      • All DBs
      • This DB
    • Dirt and air shots: makes a diagram with the over, under, within for the point cloud (or ground/canopy top) for all points in the DB and each of the 3 global DEMs, and a table with the results
      • All DBs
      • This DB
      • This DB, landcover categories.  The landcover must be in an integer field; when you add it from the landcover grid, pick the nearest value as an integer
      • This DB, slope categories
    • ICESat-2 beam splitter: makes a LAT profile and a distribution summary for the current filter, adding each of the 6 beams to the current filter
    • Cloud summaries
      • All DBs
      • This DB, landcover categories.  The landcover must be in an integer field; when you add it from the landcover grid, pick the nearest value as an integer

ICESat2 anomalies:

• In Gabon, there are a number of geoid elevations below sea level.  These appear to occur where the canopy estimates exceed 100 m.

Open point cloud

• From Icesat-2 to LAS. 
• From airborne lidar, which will have much larger point densities.

Canopy button results

Last revised 1/5/2021