Raster GIS Lab: Map Algebra

Deliverables, in a single Word document:

• Which do you find most useful, the DSM, DTM, or NVS grids? 
• Assess the quality of each survey in terms of point density, or complete coverage of the coast.
• Consider a large scale map from the most complete lidar survey.
  • What is the difference between creating a derivative grid for slope and reflectance, versus symbolizing the DEM differently?
  • Which maps (elevation or reflectance) have the most missing data, and why is this the case? 
  • How well does the TIGER line data overlay the LIDAR?  Which do you think is more accurate (Lidar or TIGER), and why?
• Consider just the oldest and most recent DEMs, and a difference map between the two.  What has happened in this area over the time period?
• Consider all of the maps
  • What do the statistical maps from the DEMs (envelope and standard deviation) tell you about coastal change during this decade?  How do they compare with the simple difference map from the oldest to newest surveys?
• Based on what you see overall, does this area appear to be eroding or accreting (depositing)?

Go to the NOAA Coastal Lidar site (link and directions), and get at least 6 lidar data sets for a beach region near your home.

• Pick a small area about 3 km long along the coast (you will wind up with a thin strip because of the way lidar is collected).  Do not get greedy in the size of the area.
• Try to get a stretch with sandy beaches.
• If possible, get at least one survey with topobathy data, which should be mentioned in the survey name.
• Pick the following dates:
  • The two oldest surveys.  The oldest surveys are frequently of lower quality compared to more recent lidar hardware, and hopefully one of these will allow change detection over a longer time interval.
  • Surveys before and after the largest storm to impact this part of the coast.
  • The two most recent surveys.

LAS file prep, and grid creation

• You want to create a DTM, DSM, and NVS grid for each survey. 
• You cannot get a DTM if the LAS files do not have a classification (check the options available for coloring the display).  Do not attempt to do this classification.
• You want the map to be tightly centered around the region you want.  The new grids will cover the same area as the map, and bigger area means larger files and longer processing time.
• Once you have the first new grid, you can use it as the base and load the other point clouds on and they will all match exactly.
• For some of the older data sets, there might not be ground classifications, in which case you cannot create a DTM.  To create a DTM you would have to install a program that can do the classification, and then use the appropriae Class tab option.  If there is no point classifiction, that option to color the point cloud will not be enabled, and you would get a blank grid if you tried to create it.

You will create a number of new grids.  You can save these to reopen later (name them carefully, and remember where you saved them), or just recreate them as needed.

New DEM and Open all the grids.  You need to decide:

• Which year's survey has the best data.
• Which type of grid (DSM, DTM, or NVS) is the best.

New DEM and Open just the best grid.

• Create a derivative grid with Slope in percent. 
• Create a derivative grid with Reflectance, using the default sun illumination.  Display this as "elevation", and a grayscale coloring.
• New map windows for slope and reflectance.  These do not create a new grid, but just symbolize the DEM differently (Display Parameter).  You can also just recolor the existing map.

New DEM.  Close everything, and open just the newest and oldest grids.  If one of those is of particularly low quality, pick the next oldest or youngest.

• Create a Difference map between the two grids.  The order in the operation is important, because you want to differentiate erosion and deposition in this 10 year period.  Do you want positive to be erosion or deposition?
• Export this map to Google Earth, to see what the imagery says about where the differences are.

Open all the DEMs.  Insure that you only have the DEMs open, and not any of the derived grids.  If you were unable to create a reasonable number of grids, you can download grids from Duck, NC, and use them for this part.

• Multiple Grid statistics.  You want particularly the envelope and the standard deviation.
• Multiple DEM profiles  Note that you can get a legend.  You should run the profiles perpendicular to the shoreline and the dunes.

You might want to look at this area in Google Earth, or try the WMS map overlays to help with your interpretations.  Export options on the Right click on the map popup menu, which can show the location in Google Earth, or USGS TNM viewer high resolution imagery.

Last revision10/4/2018