Constructing Maps with UAS Data

Introduction

Why are proper cartographic skills essential in working with UAS data?

      It is important to communicate the geographic information associated with UAS data to an observer in a way that they will understand what is being conveyed on the map.  

      What are the fundamentals of turning either a drawing or an aerial image into a map?

            There are several key elements that should be included in a map such as 1) a legend, 2) a title, 3) a north arrow, 4) scale, 5) citation information, 6) watermark, 7) a border, and 8) a locator map

      What can spatial patterns of data tell the reader about UAS data? Provide several examples.

          Examples:  

          Wetlands monitoring: Using orthophotos derived from photographs taken by a UAS in conjunction with radar data to determine phenology and land cover change in wetlands

    Farming: UAS can detect nitrogen levels, growth levels, and with infrared sensors can detect photosynthesis efficiency in plants. 

      What are the objectives of the lab?

           The objectives of this lab are to gain an understanding of the difference between an aerial image and a map and using UAS data to create proper maps that include the necessary elements previously discussed in a way that makes it easy to understand for anyone. 

Methods

For this exercise, Dr. Hupy provided UAS data from the Hadleyville Cemetery which was then copied and pasted into the personal student folder and renamed 'Hadleyville'. This data, and the map created from it, was analyzed using ArcMap. After establishing a connection to the student's personal folder that contained the 'Hadleyville' folder, both the DSM and the orthomosaic were added to the data frame by clicking and dragging each raster from the catalog. Next, statistics needed to be calculated on the DSM which was done by right-clicking on the raster file in the catalog and scrolling down and selecting 'Calculate Statistics'. 

What is the difference between a DSM and a DEM?

     A DSM is a Digital Surface Model whereas a DEM is a Digital Elevation Model. A DSM gathers elevation values of a surface using LiDAR and captures natural and built features such as trees and buildings. Applications using DSMs can include runway approach zone encroachment, vegetation management, and view obstruction. A DEM, on the other hand, gathers elevation values but cuts through the natural and built features resulting in a smooth, bare-earth elevation model. Applications using DEMs can include hydrologic monitoring, terrain stability, and soil mapping. 

What is the difference between a Georeferenced Mosiac and an Orthorectified Mosaic? 

     A georectified mosaic is one that has been digitally aligned with a map of the same area. A number of control points are marked on both the image and the map that are then used as reference points for further processing of the image. An orthorectified mosaic is one where the geometry of the image has been corrected so that each pixel appears to have been acquired from directly overhead and it uses elevation data to correct terrain distortion. 

Figure 1: Calculated statistics of the Hadleyville DSM

To view the statistics that were previously calculated, go into 'Properties' and under the 'Source' tab scroll down until the bold heading on the left-hand side read 'Statistics' (Figure 1).

What are those statistics? Why use them?

The calculated statistics for the Hadleyville DSM are as follows:

• Minimum: 283.36
• Maximum: 310.46
• Mean: 289.35
• Standard Deviation: 4.45 

These statistics reveal the surface elevation patterns where the lowest value is 283.36 meters, the highest value is 310.46 meters, and the average elevation value is 289.35 which is much closer to the minimum value as opposed to the maximum value, which suggests there are a greater number of lower elevations compared to higher elevations. Knowing these patterns can be very useful for the reader depending on the application of the DSM. 

Figure 2: 'Hillshade' tool window

After looking at the statistics of the DSM, the next step was to run the Hillshade tool on the DSM. Hillshade reveals shaded relief from a source of illumination. To do this, type "Hillshade" into the search bar and click on 'Hillshade (Spatial Analyst)' when the search results come through. This opens the 'Hillshade' tool window (Figure 2). Under the 'Input raster' tab, select the Hadleyville DSM and save it into the appropriate geodatabase for this exercise in the 'Output raster' box. Click 'OK' and let the 'Hillshade' tool run the analysis.

After completing the hillshade on the DSM, an oblique 3D image needed to be produced of the DSM. This was done in ArcScene where the DSM was added to the data frame, and under 'Properties' under the 'Base Heights' tab, the elevation features were set to a custom surface (the DSM). Then under 'Scene Properties', the vertical exaggeration was set to '0.1' to produce the exaggerated 3D image. The image was then exported as a 2D JPEG which could then be utilized in ArcMap. 

Results/Discussion

Figure 3: Hadleyville delineated 3D map (top) and
hillshaded map (bottom)

The result is a hillshaded map which is overlaid on top of the DSM and given a 50% transparency value (bottom of figure 3). Green values are low shaded relief while red values are high shaded relief. The green values appear to concentrate on the southeast side of the vegetation while the red values tend to concentrate on the northwest side of the vegetation. This suggests the source of illumination of the DSM was out of the southeast. Above the hillshaded map is the delineated 3D image brought in from ArcScene (top of figure 3). Green areas are areas of low elevation while red areas are those of high elevation. The image is broken up into five sections where 1 = trees, 2 = the cemetery, 3 = crop field, 4 = the road, and 5 = vegetation. 

For discussion purposes, the DSM and orthomosaic images have been included as well (Figure 4).

Figure 4: Hadleyville DSM (top) and
orthomosaic (bottom) images

What types of patterns are noted on the orthomosaic?

Patterns that can be noted in the orthomosaic image include the following:

• A long and narrow grey colored object with parallel boundaries and dashed lines cutting through the middle of it, indicating a road
• A rectangular shaped area of grass in the middle of the image that contains small, white and grey colored spots that are arranged in disjointed lines, indicating a possible man-made area of some sort
• Groups of trees in the southwest corner of the grassy area and along a stretch east and extending south from this area
• Vegetation to the south and west of the grassy area arranged in neat rows indicating a crop field of some kind
• Vegetation with no distinct patterns north of the road

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?

Patterns that can be noted on the DSM include the following:

• A dark grey line running east to west towards the top of the image with thinner, darker lines appearing above and below the thicker line
• A rectangular shape in the middle of the model with fairly uniform coloring
• A rectangular shape in the west and south part of the model with fairly uniform coloring, where the western section has some darker shades of grey
• Very light grey to white colored blotches with very rough edges appearing in the southwest corner of the middle rectangular shape and to the east and extending south of the same shape

It should be noted that medium to dark grey colors make up a large portion of the model. Black and dark grey colors represent low elevation areas while very light grey and white areas represent high elevation. Relating back to the statistics that were calculated for this DSM, it makes sense that the mean value was 289 when the minimum was 283 and the maximum 310. The mean value is closer to the minimum value (dark grey) rather than the maximum value (light grey/white) which indicates the image should contain more colors that are dark than light and it does.

The light grey/white blotches on the DSM match up directly with the trees in the orthomosaic. The dark grey line running east to west matches up with the road in the orthomosaic and the thinner, darker lines above and below the thicker line are probably ditches. The rectangular area in the middle of the model with uniform color matches up with the grassy, rectangular area in the mosaic. And what appears to be a crop field in the mosaic aligns directly with the rectangular shape with uniform color west and south of the middle area. 

Where is the data quality the best? Where is poor data quality noted? How might this related to the application?

The data quality appears to be the best in the middle area of the image while the poor data quality can be noted along the edges of the mosaic and the DSM, particularly on the lower right side of the DSM and on the right side of the mosaic where the road is cut off. This is probably related to the application in that the area of interest was the grassy, rectangular area in the middle of these rasters, therefore the quality is greatest within that area and decreases going away from that area.


Conclusion

As a tool, UAS data is useful to a cartographer and GIS user because it can be analyzed in so many and has numerous real life applications. Combining it with cartographic elements can help communicate the geographic information that is associated with it to people in a way that they will be able to understand what taking place on the map. All UAS have limitations which can include, but not limited to, sensor quality and capabilities, pilot experience, GPS quality if data is collected using GPS, weather conditions, and UAV quality. When it comes to working with collected data, the user should know the platform and sensor that was used, who collected the data, and the conditions of the flight, among other things. To make UAS data more useful, it could perhaps be combined with satellite imagery for certain types of analyses.

Sources

Wetland monitoring example:  https://books.google.com/books?id=BHwZBwAAQBAJ&pg=PA213&lpg=PA213&dq=what+can+spatial+patterns+of+data+tell+someone+about+UAS+data&source=bl&ots=vP1FpsAjHo&sig=YdB2SA_a3MPNPD4GiWMY9S9yeY8&hl=en&sa=X&ved=0ahUKEwijo6CJ1JrSAhUK5YMKHfqjAnsQ6AEIJTAC#v=onepage&q=what%20can%20spatial%20patterns%20of%20data%20tell%20someone%20about%20UAS%20data&f=false 

Farming example: http://news.nationalgeographic.com/news/2013/12/131202-drone-uav-uas-amazon-octocopter-bezos-science-aircraft-unmanned-robot/

DEM and DSM information: http://gisgeography.com/dem-dsm-dtm-differences/

Georectification definition: http://support.esri.com/other-resources/gis-dictionary/term/georectification

Orthorectification definition: http://support.esri.com/other-resources/gis-dictionary/term/orthorectification