Lab 4: Creating a Digital Elevation Surface Model using critical thinking skills and improvised survey techniques

Lab 4

Introduction:

The purpose of this lab is to construct a elevation surface model of terrain that our group constructed in a square meter "sandbox". In order to gather points, we needed to figure out what type of sampling we wanted to conduct. There are many different ways to sample points in any given model. The three main types of sampling are random, systematic, and stratified. Random sampling has the least amount of bias, but it could lead to a poor representation of the overall area in question. Systematic sampling samples the majority of the study area using set intervals, but is more biased and can lead to over or under representation of the area. Stratified sampling can generate accurate results that represent the study area as a whole, and it's flexible when it comes to data correlations and comparisons. The one major disadvantage of stratified sampling is that is that the proportions of the areas in question must be known and accurate. Random sampling was not a good choice, because we needed a structured sampling system. Stratified sampling was not an ideal choice either because we wanted to measure the entire box, not just small portions of the box. We wanted to make sure we measured the entire box to make sure we captured all of the elevation change. We used systematic sampling by measuring out even plots for the entire "sandbox". All of the plots we of equal size. It is important that we chose the correct sampling method because we want to know elevation changes which can occur rapidly in many areas, so choosing a sampling method that covers the entire area is critical.

Methods:

My group and myself chose the systematic sampling method. We chose this method because it seemed to make the most sense since we were working with primarily elevation. We wanted to make a grid to make sure we collected a data point from every spot on the model. We created a grid system that allowed us to record a data point ever 6 cm. Figure 1 below shows the 114 cm by 114 cm "sandbox" that was the study area for this lab.

Figure 1: The study area terrain model

As seen in figure 1, there are pushpins outlining the study area. We placed a pushpin on the perimeter of the sandbox every 6 cm. We chose 6 cm because 6 cm x 19 cm = 114 cm. This means there were 19 columns and 19 rows resulting in 361 6 cm by 6 cm plots. The string was then wrapped around the pushpins creating a grid system. Figure 2 below depicts the grid system nearly complete.

Figure 2: The grid system that was used to record data points

The study area/sandbox is located in a backyard near Philips Hall. The backyard is across the road from the Philips Hall garage/shed. In order to create our sampling system, we needed measurement tools. In order to start our lab, we had to create a terrain model with the following landforms: ridge, hill, depression, valley, and plain. We used meter sticks to measure out and place pushpins every 6 cm. We then used string to create our grid system. In order to create the elevation model, collecting the "z coordinate" was critical. We chose to have the top of the wood of the sandbox be sea-level. This meant everything below the wooden sandbox was below sea-level, and the data points that were above the wood were above sea-level. In some areas where elevation relief was steep, we took two points which, in a sense, split the plot in half. This will allow for a more accurate DEM. In order to keep some level of standardization, we had one person hold the meter stick and place it in each plot in the same general area. A different person read aloud the measurement taking into account sea-level. The last person was the scribe and wrote all of the data points in a notebook.

Results/Discussion:
The sampling method (systematic) that we chose worked very well. If I were to recreate this lab, I would use the exact same sampling method. As we were setting up the sandbox and grid system, it seemed like overkill, but now I am glad we chose to take a lot of data points. Gathering the data only took one time, so I would say this lab has been a success in that sense. The total number of sample points we recorded was 433. 433 data points for a 114 cm area. This was definitely overkill, but it created a very accurate terrain model. The highest point was at 10 cm above sea-level whereas the lowest point was at -13 cm below sea-level. The average elevation for all of the points was -2.16 below sea-level. The standard deviation was 4.15 meaning that close to 68% of all the points were between -6.3 cm and 2 cm.
o Did your sampling technique change over the survey, or did your group stick to the original plan. How does this relate to your resulting data set?
o What problems were encountered during the sampling, and how were those problems overcome.

Conclusion:
o How does your sampling relate to the definition of sampling and the sampling methods out there.
o Why use sampling in spatial situation?
o How does this activity relate to sampling spatial data over larger areas
o Using the numbers you gathered, did your survey perform an adequate job of sampling the area you were tasked to sample? How might you refine your survey to accommodate the sampling density desired.

Sources:

"Sa      "Sampling Techniques." Sampling Techniques. N.p., n.d. Web. 09 Oct. 2016. http://www.rgs.org/OurWork/Schools/Fieldwork+and+local+learning/Fieldwork+techniques/Sampling+techniques.htm