Clipping Regions for Comparison

For a first comparison, I decided to extract four regions that represent different topographic domains.  In the image below, the three light blue polygons show the domains for D1 (furtherst south), D2 and D3 (furthest north).

Below are descriptions and LiDAR slope-shade imagesfor the first three domains.

• D1: a mountainous, rocky surface with high slopes

• D2: a tundra-covered hilly area with good fluvial network and diffusive hillslopes

• D3: a low relief area with many coalessed thaw lakes, north of the moraines

• D4: a small region with known thermokarst activity between 2009 and 2012
• (this area is not yet defined...considering Horn Lake or Itkillik Riverside features)

??

For each domain I will:

• Extract a subset of elevation data from the tiles with the most coverage.  If there are overlapping tiles from the same data source, I will mosaic those two together (best method? mean? max? largest area?).
• D1
• Aerial LiDAR Data:  
• (will need to be merged before analysis)
• m4_5_6
• m7_8_9_11
• ASP_Par_4m: 
• WV01_12JUN212145259-P1BS_R2C1-102001001B04AA00_WV0
• ASP_Bay_4m_W1:
• WV01_12JUN202207366-P1BS-102001001C17A600_WV01_12J
• WV01_12JUN202207377-P1BS-102001001C17A600_WV01_12J
• ASP_Bay_4m_WV2:
• (No Data in Domain)
• SETSM:
• SETSM_ArcticDEM_46_18_5_3_DEM
• D2
• Aerial LiDAR Data: 
• m18192922
• ASP_Par_4m: 
• WV01_12JUN152224058-P1BS_R4C1-102001001C555C00_WV0
• ASP_Bay_4m_W1 
• (two tiles, same half-coverage, different dates):
• WV01_12FEB132219064-P1BS-1020010017AD8900_WV01_12F
• WV01_12OCT212212221-P1BS-102001001EB61400_WV01_12O
• ASP_Bay_4m_WV2:
• WV02_12OCT022230092-P1BS-103001001B4F9F00_WV02_12O
• SETSM: 
• (strange merge of 2 datasets, one rough, one smooth?)
• SETSM_ArcticDEM_47_18_1_5_DEM
• D3
• Aerial LiDAR Data: 
• m2728293031
• ASP_Par_4m: 
• WV01_12JUN032207416-P1BS_R3C1-102001001CE78000_WV0
• ASP_Bay_4m_WV2:
• WV02_12OCT202307414-P1BS-103001001C901700_WV02_12O
• WV02_11APR062258334-P1BS-103001000A751700_WV02_11A
• SETSM:
• SETSM_ArcticDEM_47_19_1_2_DEM
• D4
• ????
• After extracting the datasets within each domain, I will run a standard set of calculations, assuming that the airborne lidar is the 'most accurate' dataset.
• Straight difference (LiDAR - stereoDEM)
• output map of differences
• look for spatial coherence in differences, suggesting georeferencing errors
• output histogram of differences
• good for comarison against other stereoDEM datasets
• output bulk statistics of differences
• good for comparison between domains