Why are cartographic skills essential when working with UAS data?
Without proper display, UAS data is useless. In order to ready UAS data for presentation, especially to those who were not at the scene working on the UAS data collection, extra contextual information needs to be presented in order for the UAS data to be interpreted correctly.
What are the fundamentals of turning either a drawing or an aerial image into a map?
Many contextual items must be included. UAS data should be displayed possibly with a basemap to give locational context or a locator map such as the one seen in this lab, a scale bar, a north arrow, data sources, and a personal watermark. In addition to these cartographic elements being included, metadata should be included for more background information. This means information about where the drone was in altitude when taking the image, the drone's sensor's characteristics, GPS precision level, other tools used to collect the data, coordinate system and projection information, and date and time of data collection.
What can spatial patterns of data tell the reader about UAS data? Several examples?
Data that shows less area (when this is the area of interest the data was collected specifically for) can indicate a lower UAS elevation of data collection. Data showing more area (if the area of interest that the data was collected for was a larger area) can be assumed to have been collected from a higher elevation as one would want to collect more data with less images and with less geometric correction needed.
What are the objectives of this lab?
This lab is meant for practice of purely the cartographic elements of the UAS data process. Instructor supplied UAS data already processed in Pix4D is to be turned into a map with all proper cartographic and metadata information.
Methods:
After copying the appropriate folders of data that were supplied to my personal folders I opened ArcGIS Pro and made a new project folder in which a new file geodatabase for the project was made. I now added the appropriate orthomosaic and DSM for the middle school track field from the connected folder to the file geodatabase. From here I ran a the hillshade tool using my DSM file. I now made two maps. In one I put the orthmosaic, and in the other I layered the DSM and hillshade layers so that the hillshade was on top of the DSM and was partially transparent. I now changed the DSM layer's color ramp in the symbology settings to the "Elevation 1" color ramp. I opened ArcScene and imported my orthomosaic layer. In the layer properties window I selected the "Base Heights" tab from which I selected "Floating on a custom surface" and selected my DSM layer to be this custom surface. I now selected "calculate from extent" for the vertical exaggeration in the scene properites window. Selecting "Edit" and then "Copy Scene to Clipboard," I copied the scene which I then pasted into a new 8.5 x 11 landscape layout in ArcGIS Pro. In this layout I now added all of the necessary map elements, metadata, and maps.
What is the difference between a DSM and DEM? The difference between a Georeferenced Mosaic and an Orthorectified Mosaic (This was discussed in the demo, but please do a quick search and cite your sources in your description)
A DSM (Digital Surface Model) shows the surface of the earth with the above ground features such as buildings, trees, and other structures included. This is made using first return data filtered from a LIDAR point cloud data set. The DEM (Digital Elevation Model) usually refers to the display of the bare ground (LIDAR last returns filtered from a point cloud dataset)
Georeferencing of images is simply putting those images together and into a a correct coordinate system reference frame. An orthomosaic image is made from multiple images and attached to a coordinate system but also has had the geometric errors added by changes in elevation and placement of sensor corrected for in the orthorectification process. This process uses ground control points, tie points, and also a DSM. On large scale images taken from lower elevation UASs these errors are magnified.
After performing the necessary layer creation and symbology changes for I looked at the statistics in the item properties for the DSM.
What are those statistics? Why use them?
These statistics, along with the unit information found in the same window provide simple contextual information about the elevation change data. The maximum and minimum show the extents to which there is change in the map and the standard deviation shows how relatively flat or rugged the area is.
Hillshade the DSM images. How did you do this? Delineate regions of the DSM, thinking of each region in terms of topography, relating that to the vegetation.
The hillshade, as told before was made with the hillshade tool using the DSM. In viewing the DSM and hillshade together as told about above, there is a clear shallow slope going downward toward the south-west of the data. This is the field, covered by concrete or another surface for the track and also grass. This surface is interrupted by trees which sharply contrast the flat area.
Results:
Figure 1 depicts the maps made from the orthomosaic and from the DSM and Hillshade Models and Figure 2 depicts the data quality issue.
What types of patterns do you notice on the orthomosaic?
In the orthomosaic, both on the map and the scene view patterns of vegetation and land cover can be observed. Trees, grass covered areas, track, roads, and snow can all be differentiated. Elevation of trees and buildings can be observed in the scene view.
What patterns are noted on the DSM? How do these patterns align with the DSM descriptive statistics? How do the DSM patterns align with patterns with the orthomosaic?
In the DSM the elevation of buildings, of the track, and of the trees can be observed, as well as the general trend to slope to the south-west. The small standard deviation is evident in the relatively flat nature of the field, and the maximum and minimum elevations are also evident in the north-east and south-west corners of the image. The general trend observed in this image is not seen or noticeable in the orthomosaic, but the acute changes in elevation such as trees and buildings are.
Describe the regions you created by combining differences in topography and vegetation.
There are two obvious areas when looking at the DSM. There is the area including all of the grass and field, the track, the street, and other surrounding areas, and then there is the areas with trees and a shack which have acute elevation differences. When you add the vegetation data on top of this, the two areas become further divided. There is the field center area of tended grass, there is the paved track and paved roads, there is the sidewalk, there are the areas with trees and buildings creating natural barriers around the track, there is the snowy covered area, and there is all of the lawn surrounding the track. The land is further divided when you examine the sharp drop off on the east side of the DSM. This shows that the field is a separate graded surface.
What anomalies or errors are noted in the data sets?
Between the curb and the sidewalk only some of the trees' elevations are reflected by the DSM. The ones that are registered on the DSM distort the immediately surrounding area with the interpolation method used.
Where is the data quality the best? Where do you note poor data quality? How might this relate to the application?
Data quality is best in the detection of the slow change in elevation of the field. It is not as good in the detection of the trees' forms, and should not necessarily be used to make a DSM, perhaps only being of high enough quality to make a DEM.
 |
| Figure 1 |
 |
| Figure 2 |
Conclusion:
UAS data can be extraordinarily helpful to a cartographer or GIS user. Data, as long as it has proper metadata, can be mapped and used to interpret areas of land in new and exciting ways. Now, without having to commission an expensive manned flight, as long as certain skills (and FAA regulations and airspace safety knowledge) have been ascertained, UAS offers an easier solution to capture image data that can then be processed like stereoscopic imagery to create DSM files. This data can be used to differentiate land uses, grades, and observations such as the sharp drop off on the east side of the data that would have not been otherwise visible can be seen with the correct UAS data.
What limitations does the data have? What should the user know about the data when working with it.
The data is limited by the quality of the data. A user should know that the accuracy of the elevations and forms of the trees are not accurate. A user should also note the resolution (radiometric, spatial, etc) of the data which could affect data analysis in so far as size of different areas or ability to differentiate different areas of vegetation health or other very slightly different areas.
Speculate what other forms of data this data could be combined with to make it even more useful.
One other type of data that could be used would be GPR (ground penetrating radar) data. This type of data could be used in conjunction with the data already collected for data on digging and what could be found and where bedrock and other problems or below ground objects could be found when doing new grading of the area for land repurposing. This could include construction of a building, regrading the area if it has been sinking or eroding, or even installing a geothermal heating system.
Sources:
http://gis.stackexchange.com/questions/5701/what-is-the-difference-between-dem-dsm-and-dtm
https://imageryspeaks.wordpress.com/2012/01/24/georeferencing-vs-georectification-vs-geocoding/
http://gisgeography.com/dem-dsm-dtm-differences/
https://apollomapping.com/blog/g-faq-orthorectification-part
