Lab 2: Surveying a Miniature Landscape and Digital Represention of Data

 Intro:

For this project, we were asked to create a miniature landscape in one of the sandboxes located in the courtyard of Phillips lecture hall, complete with geographic features such as plains, plateaus, valleys, depressions, mountains, and/or canyons. When completed, the features were to be surveyed, using a grid, Cartesian, coordinate system. With the measurements of each grid block carefully recorded, X, Y, and Z coordinates were input into a Microsoft Excel spreadsheet, to be them imported into ArcMap for further data manipulation. An appropriate interpolation was then selected and a TIN, or Triangulated Irregular Network, was created through the survey points that were collected. The TIN was then imported into ArcScene, in order to visualize the created landscape of the sandbox digitally within the computer program. However, due to a general lack of experience in survey methods, we were allowed an opportunity to make necessary changes in our landscape and survey methods in order to make a more pleasing image in both ArcMap and Arc Scene. After the second survey, the same process was followed, and the changes made in data collection and methods produced a superior digital representation of the sandbox landscape.

Methods:

• Creation of the Landscape



Figure 1: Zach creating the river bed in the sandbox landscape

 For the landscape that we had decided on creating, a mountain, river valley, plateau, depression, canyon/river bed, and several plains areas were created. To do this, it was decided that using the snow that was already plentiful inside of the sandbox would be simpler and easier to form than using the sand that was underneath the snow. In Figure 1, Zach can be seen creating the depression in the river bed/ canyon feature. Figure 2 shows myself flattening down snow to create plains. After the landforms were created, the snow was moist enough to stick together and remain that way, preserving the features that we had built.

Figure 2: Drew packing down snow to create a region of plains

• The Grid System and Recording of Data



Figure 3: Laying out the string into grid sections to survey the features








For our coordinate system, our group decided that a Cartesian system of grids would be the most appropriate for accurate data collection. The first time we conducted this exercise, grids of 10cm by 10cm were thought to be the most accurate. To get these grids, string was laid in sections across the sandbox after the landforms were created. The entire box was not partitioned out, however. The measuring tape was used as a guide, and marks were made in ten centimeter increments, with measurements being at the middle of every single increment.

Figure 4: Hannah recording survey points of the landscape in Cartesian format

Unfortunately, after the first phase of the surveying process, it was found out that 10cm by 10cm grids was not fine enough to accurately represent the landscape, so it was decided to move to 5cm by 5cm grids, for a total of 1,056 survey points, as illustrated by Figure 5.This provided much more detailed coordinates, giving our group the data necessary to make an accurate representation of the landscape that we had created.

Figure 5: The data points collected after deciding on a 5x5 cm grid system

 

•  Data Importation, Interpolation Methods, and TIN Creation

After diligently recording the points onto paper, we migrated the points into a Microsoft Excel spreadsheet, and then moved to import those coordinates into ArcMap for the first step of data manipulation. Importation was simple due to the program's ability to integrate with Excel files. From this point, interpolation was done on the survey points, with the ultimate goal being to find a method that would produce the best looking and most visually representative model for the landscape that we had created in the sandbox. Surprisingly, our new data set was significantly better than the first one, and there was little to no difference between the interpolation methods of IDW, Kringing, Natural Neighbor, and Spline, illustrated in Figures 6, 7, 8, and 9, respectively.

Figure 6: Survey points illustrating IDW method of interpolation

Figure 7: Survey points illustrating Kringing method of interpolation

Figure 8: Survey points illustrating Natural Neighbor method of interpolation, the best looking method

Figure 9: Survey points illustrating Spline method of interpolation

 After deciding on a method of interpolation that looked the best visually, it was then time to create a TIN, or Triangulated Irregular Network of the survey points. A TIN uses the Z values of elevation to assist in the creation of a three dimensional image of an otherwise two dimensional figure in ArcMap. The TIN can be then imported into ArcScene, and then a 3D model of the survey points can be created.

• ArcScene and the Creation of a 3D Model of the Landscape

 Finally, using the TIN that was created in the previous section, a digital, three dimensional model of the landscape created in the sandbox and recorded through the survey points can be created in ArcScene. The updated survey method of creating 5x5cm grids created a very accurate representation of the landscape that we created in the sandbox. Figure 10 demonstrates the digital landscape created in ArcScene, and when compared to Figure 11, the picture of the actual landscape created in the sandbox, it is clear that the survey method was effective, and our group's hard work was successful.

Figure 10: 3D Representation of the landscape created in the sandbox

Figure 11: Physical landscape created in the sandbox

Discussion: 

The process of finishing this project from start to finish was the most time consuming part. Taking 1,056 individual survey points was extremely time consuming, taking roughly 2.5 hours to complete fully. But this was undoubtedly the most effective way to produce enough points to get an accurate representation of the landscape that we had created in the sandbox. Moving from 10x10 cm to 5x5cm grids is the sole reason for smooth looking TIN image, which also translates into an image in ArcScene that closely resembles the real landscape produced within the sandbox.

There were some difficulties in the data collection and the importation of the survey points into ArcMap. The first was the recording of the points themselves. There were so many points to record that it was inevitable that there would be some points that were recorded wrongly. In some instances, negative signs were omitted or added, and that threw the landscape off in the interpolation and TIN creation. Thankfully, this issue was easily remedied by a simple edit session within ArcMap. Another issue occurred when deciding to create a false sea level. It seemed like a good idea, considering that most of the landscape was under the wooden barriers of the sandbox, and therefore most of the values for elevation would be negative and certainly not representative of a real environment. However, when the points with the false elevation were imported into ArcScene, a grossly unrepresentative image was produced, as illustrated by the image in Figure 12. It is obvious that this was not an accurate image, so the original elevation coordinates were used instead, again demonstrated in Figure 10.

Figure 12: ArcScene image of survey points with false elevation

Conclusion:

This lab was a very useful tool in teaching about surveying a landscape and then taking the data retrieved and importing it to create a digital copy of it. Working effectively in a team was absolutely essential in producing an accurate representation of the data, and communication was vital in doing this. But simple recording of points was the easy part, and critical thinking skills were necessary to address the numerous issues that came up over the course of the project. Editing individual survey points, deciding whether or not to use a false elevation, and selecting an appropriate interpolation method were all issues that had arisen throughout the course of the project. But trial and error created a favorable result that not only completed the goals of the project, but also taught valuable skills in surveying and added expertise in both ArcMap and ArcScene.