Assignment 4- Creating a DEM

Introduction

Assignment 4 involves planning, creating, and measuring the beginnings of a Digital Elevation Model (DEM).  In order to do this, students created a landscape within a sandbox.  The sandbox was overlayed with a grid using strings.  Each box that the strings created was measured and recorded in order to create the DEM later.  There are different techniques used to measure these boxes overlaying the landscape, which is called sampling.  There are three main types of sampling.

• Random
• This is the least biased method.
• Can be used with large amounts of data.
• May lead to poor representation
• Systematic
• Samples are evenly distributed and more straight-forward.
• Can be regularly numbered.
• This method is biased, can lead to underrepresentation.
• Stratified
• Can be used along with the other two methods.
• Flexible and applicable to many different areas.
• Proportions of the areas must be accurate in order to work properly.

The main objectives of this lab were to create a landscape, overlay it with a grid, take measurements and input those into an excel sheet.  This sheet will be used in the next assignment in order to create the DEM.

Methods

In this assignment, the systematic sampling method was used.  It was decided that this is the most reliable method and best for this situation.  This is similar to the stratified method, but it was ultimately decided that systematic would work better in this scenario.  In this assignment, each group got to create their own landscape in the sandbox.  The only requirements were to have five specific features which were,

  A. Ridge

  B. Hill

  C. Depression

  D. Valley

  E. Plain

These five landscapes are displayed in Figure 1 below with the corresponding letters.

Figure 1: The five landscapes displayed with corresponding letters.

The materials used for this assignment were the sandbox, sand, our hands, string, thumb tacks, and yard sticks.  The sandbox was measured at 114 cm on each side.  The group decided to divide each side into 6 cm.  Below in Figure 2 the thumb tacks on each side can be easily seen.  6 cm was decided to be the best length because it would create small squares.  Small squares means that the DEM will turn out more accurate because more measurements were taken.  Also in Figure 2 in the left hand corner is where the first measurement was taken.  From there, it followed upwards along the Y axis to take measurements. Once that row was done, the next column upwards along the X axis was measured.  This was followed in the last measurement was taken in the upper right hand corner of the sandbox.

Figure 2: Sandbox with tacks and a partial grid showing the X, Y layout.

Before taking measurements, sea level needed to be decided.  It was decided that the wooden edge of the box was sea level (0).  In order to take down these measurements, a grid was created on a piece of paper with a box on the paper representing a box on the actual grid.  Then, one group member held the yard stick in each box, another member read the yard stick measurement out loud, and the final group member wrote down the measurements on the piece of paper.  This method was chosen after some debate between the group members.  Figure 3 below displays how the grid kept going across more of the sandbox.

Figure 3: The grid overlaying the sandbox.

Results/Discussion

There were a total of 18 squares in each row and column.  19x19 = 361.  Some boxes had to be measured on the left and right side rather than just in the middle, because of greater relief which needed to be measured.  This resulted in 434 total measurements.  Below in Figure 4, the measurements for the sandbox are displayed.  Figure 4 shows which columns are X, Y, and Z.

Figure 4: Measurements in the excel sheet.

The mean, maximum, minimum, and standard deviation were found from these statistics in the Z column.  They are displayed below in Figure 5.  

Figure 5: Values for the measurements in the Z column.

The sampling technique stayed the same during the survey because the group found that it was working quite well.  The only problems that occurred during this assignment were the the group had to split up square into halved in order to make measurements more accurate.  This may be difficult to transfer into a DEM, but these difficulties will be outlined in the next post.

Conclusion

The sampling technique used was systematic which was best for this situation.  It gave an accurate result which will transfer well into a DEM.  Sampling is defined as "a shortcut method for investigating a whole population".  Systematic sampling related to this definition in that it doesn't measure each area equally.  Certain measurements are taken across an equally distributed area and then recorded.  Using sampling in a spatial situation makes thing simpler especially when the area to measure is very large.  Sampling gives a good overall view of an area without having to measure each individual square inch.  The survey performed on the sandbox was adequate and produced accurate results.