Sandbox Survey (part 1): Creating a Digital Elevation Surface Model

Introduction:

Sampling is an effective time and resource saving tool that is often utilized in the process of data collection. Many times, a study interest may realistically be too large to undertake a thorough and detailed collection of all the existing data that is beneficial to the study. In these instances, sampling allows the data collector focus on a small-scale representation of their study interest in order to make generalizations about the larger picture as a whole. For geographers, this is an ever-familiar skill set that has been frequently exercised-- identifying key themes and patterns at local scales, and analyzing them in relation to other themes and patterns found elsewhere, in an overall attempt to make better sense of our world. 

The process of sampling can be executed in a variety of ways. The three main types of sampling include random, systematic, and stratified. Somewhat self-explanatory, the process of random sampling involves the spontaneous selection of data, where all available data has an equal chance at being selected. This method is useful because it is the least bias of all three sampling techniques and can be applied to large sample populations. Systemic sampling is done according to a predetermined strategy or system. Data collected is done in even intervals along the study area. This strategy is useful since it allows for thorough coverage of the area in study. Lastly, the third type of sampling, stratified sampling, is used when the area in study is composed of smaller areas of a standard size. These smaller areas are each individually a smaller representation of the larger area, and should therefore be reflective of that. One example of this might be blocks within a given neighborhood. The neighborhood is the overarching study area, but each block may be miniature representations of what the neighborhood looks like as a whole.

For this lab, students were placed into groups of three and asked to construct an elevation surface of terrain using a sandbox approximately one square meter in size. The terrain needed to include a ridge, hill, depression, valley, and plain. Students were given tape, string, thumb tacks, and a few meter sticks in order to construct grid system and survey the terrain to be digitized in ArcMap during lab in the week to follow.

Methods

In beginning the project, the group determined that the best method to use for sampling the sandbox's terrain was the systemic line sampling method. This method uses the intersections on a standardized grid with uniform intervals as points for data collection across the sandbox terrain. The group felt that this was the best choice in order to have a good coverage of the sandbox terrain overall that would clearly outline the terrain features required to be used in its construction (a ridge, hill, depression, valley and plain).

Figure 1: Group Surveying and Recording Data

Of the two sandboxes located across Roosevelt Street from Phillips Hall, the group chose to begin constructing the terrain in the furthermost sandbox. Once each terrain feature had been constructed, the group focused on constructing the grid, pinning string to the sandbox's wooden frame at equal intervals measured by a meter stick. The intervals used were predetermined as a result of the sandbox frame size. The approximate square meter sandbox could be split into a 20x20 grid system with each square approximately 2x2 inches in size, allowing for a total of 400 data points to be collected, a sizable amount that could be conducted in reasonable time. One person in the group was responsible for recording the data while the other two alternated between rows in measuring the elevation level of the sand at the southwest corner of each grid line intersection.

Figure 2: Taking Measurements from Terrain Grid

In total, the operation took just over an hour and a half to complete. The string line was determined to be the surface (or sea level) of the terrain model, so most of the data collected was then negative in value. Due to the cold weather, the group decided first to transcribe these data points in a notebook to be transferred into an Excel file later on. After transferring it to the Excel file, a color scheme was added to the data plotted in the table so that the group could get a glimpse at what the terrain would look like once digitized in ArcMap. The group then also charted the data in a format that would be useful for transferring these points into ArcMap during the lab next week. Fragments of the resulting Excel data tables can be found below:

 

Chart 1: Excel Data and Normalization

Results/Discussion

Overall, the group managed to collect a total of 400 data points within the 20x20 grid.  The minimum value in our collection was -8 inches while the maximum value was a +4 inches. Since the string line was established as sea level, most of the data points fell below the line at negative values, the most commonly occurring value being -2 inches. Given the cold temperatures and tedious measurement requirements of the lab, this sampling method proved to be really useful and seemingly effective. 

Some issues did arise during the process of completing the lab. To begin, the freezing temperatures had managed to freeze much of the sand within the sandbox, and made digging the terrain for surveying difficult as it limited the areas which were soft enough to be molded. Secondly, as the group went on collecting data points, the string began to slack some in certain areas, It may have been more beneficial to double tack alternating string lines to improve its security. A third improvement could have been made in being more specific in collecting measurements. It seems the group did a lot of rounding to quarter inch markers in looking at the final data set. A better method would have been to conduct the measurements in centimeters to promote more accurate readings across the survey sample.

Conclusion:

Sampling is an effective tool to utilize in spatial settings as are commonly found in the field of Geography because it allows for larger scale analysis to be done on smaller scale levels, conserving both time and resources in the process. This activity relates well to the system used by the Public Land Survey System, which also maps out land plots into squares, but at a much larger scale. Overall, the survey system utilized was a decent system given the constraints of the weather, but of course, more data is always better. Perhaps creating more rows and columns within the grid system would have benefited the group, as it would have also resulted in the collection of more data points. Also, as noted previously, the group would have done better conducting the data point measurements in units of centimeters rather than in inches for better accuracy in numbers. 

Sources

http://www.rgs.org/OurWork/Schools/Fieldwork+and+local+learning/Fieldwork+techniques/Sampling+techniques.htm