Assignment Task

Task 1

Structure-from-Motion Photogrammetry

You should compile all of your answers to Portfolio Tasks set throughout the module in this single document prior to submission. Make sure you clearly specify the title for each Portfolio Task above your answers for that task in the document. For this first task, you should use no more than three pages and address the following:

1) Visualise a 3D point cloud of the dam created using the nadir-only (max 1 page) Using the Save current view tool under the Tools menu in VisualSFM or cropped print screen images from CloudCompare, visualise the dense, scaled and orientated point cloud you have produced using the nadir-only images. Provide a caption for your image that describes its content and the software used to generate the visualisation.

2) Produce a map showing spatial variation in elevation differences between the digital elevation model you produced using nadir-only imagery and one created using the full multi-angle image data set from the UAV survey (i.e. dem_multi)

Spatial differences in elevation between the digital elevation models produced using nadir-only and multi-angle imagery can be visualised by subtracting one surface from the other in ArcMap. To do this, ensure both rasters are visible in the Table of Contents window in ArcMap (add them using the Add Data button if not). Find and click the Raster Calculator (Spatial Analyst) tool using the Search window. Double click on the first of your two rasters to add it to the map algebra expression box. Then click the minus (“-“) button and, finally, double click the second of your two rasters to finalise the expression. Specify a sensible name and output location for the new “elevation difference” raster and click OK to produce it. You should apply digital cartographic principles to produce a professional map (landscape format) of the spatial variation in elevation differences between the two surfaces. You should give particular consideration to the map elements you include and symbology you employ to ensure that the map is clear and informative. Note that because your newly created DEMs do not have a specified coordinate system, you may need to specify their spatial units. This can be set by clicking View > Data Frame Properties > General tab and setting the Map and Display units to metres.

3) Write a short evaluation of the relative accuracy of the two elevation models (nadir-only and multi-angle) and reflect on the importance of photo coverage when creating 3D models using Structure-from-Motion photogrammetry.

The accuracy of the elevation models created using nadir-only and multi-angle imagery can be assessed with reference to the elevation of independent ground control points (ICPs) surveyed at the site. These are known, marked points on the surface of the dam that have not been used in the creation of the 3D point clouds. These have been provided for you as a shapefile called ICPs in the GIS_data subfolder in the Practical_3 folder. Add them to the Table of Contents window in ArcMap using the Add Data button and move the layer so that it is above the two elevation rasters in the list. Remove any other layers you have in the Table of Contents (right click, remove). To query the elevation value of each of the ICPs, and the raster surface elevations at that location, click on the Identify button (white “i” in a blue circle). Click anywhere in the map window and change the Identify from: value to . Now click on each of the ICPs in turn and note their elevation values (visible in the bottom window of the main Identify window), together with those of the cells in each of the elevation rasters at the same location (visible next to the relevant raster layer in the top window of the main Identify window). You can use these values to help inform your evaluation of the relative accuracy of the two elevation rasters. To inform your reflection on the importance of photo coverage when creating 3D models using Structure-from-Motion photogrammetry, you should make reference to both the map you created in Part 2 of this Portfolio Task and relevant academic literature.

Task 2

Land cover change detection using satellite imagery

1) Evaluate and explain the effectiveness of different band combinations for identifying earth surface features.

Present a maximum of 2 images showing your best band combinations and write a short commentary detailing the bands used, the types of features/phenomena they facilitate identification of, and (using appropriate academic literature) explain why they are effective.

2) Produce a map illustrating vegetation change in the study area between 2003 and 2011.

Using the vegetation raster layers you have created, produce a map that illustrates how vegetation has changed in the study area between 2003 and 2011. Think carefully about how you order the various layers and use symbology (colour/no colour/transparency) to highlight vegetation creation/removal, along with areas of vegetation and gravel/water that remained unchanged. You should apply digital cartographic principles to produce a professional map in landscape format. Carefully consider the map elements you include and symbology you employ to ensure that the map is clear and informative. Quantitative data on vegetation coverage change should also be included.

3) Produce a map illustrating river channel change in the study area between 2003 and 2011 

Using the surface water raster layers you have created, produce a map that illustrates how river channel morphology has changed in the study area between 2003 and 2011. Think carefully about how you order the various layers to highlight areas of erosion/deposition, along with sections of channels that have remained unchanged. You should apply digital cartographic principles to produce a professional map in landscape format. Carefully consider the map elements you include and the symbology you employ to ensure that the map is clear and informative.

Applied raster analysis

Aims and Objectives

In this practical you will be continuing to work on raster analysis techniques, employing DEMs and other raster datasets. We will be building on the material in the lecture, looking at how we can create viewshed maps, and perform Multi Criteria Evaluation (MCE).

By the end of this class you should be able to:

1. Calculate viewsheds for a range of observer points

2. Create and interpret viewshed maps

3. Apply boolean MCE techniques to real data

4. Apply factor MCE techniques to real data

5. Interpret the results of MCE analyses