For this project the class was given two bins that were filled full of sand, these bins represented our study area. In this activity, the first step was to create a variety of different types of terrain such as; a ridge, hill, depression, valley and plain. In order to map out these types of terrain a sampling method must be set up, but first what is sampling especially in a spatial sense? Sampling is taking a number of points/samples to determine the big picture, in a way its a shortcut method to investigate the variable(s) at question. Each sample point contains data about that specific location and that will allow a map to be created from the area that the sample points are collected from. There are a few different ways to go about sampling this landscape, the first is random sampling, basically this is where points would be randomly placed around the landscape and the data would be taken from each of those points and recorded. Hopefully the random points would be on the terrain features but since the sampling technique is random, there are no guarantees. Systematic sampling is another technique that could be used for this activity, this technique is where you start with a random point then move on from that point at a periodic interval that doesn't change throughout the experiment. This way there will likely be a grid pattern created which will make all of the sample points organized and easy to collect data from. The final sampling method is stratified sampling, this is where the samples are divided up into separate groups of where the groups are based on similar features or data points. The lab objective is to create a landscape in a bin that is provided full of sand, create a grid system to map out the landscape, log the data from the grid system into an excel spreadsheet, and finally use a computer program that will create a digital image of the landscape created.
Methods:
The sampling method used for this activity was stratified sampling because it makes the most sense with the landscape that was created, and it would be the least time consuming method. Another method that would be similar to this one would be the systematic sampling method. But since the stratified method was used, there were fewer points taken, and the overall accuracy will be much higher because of how similar points are being combined together. The materials that are used in this lab includes a sandbox with a wooden frame, copious amounts of beach sand, a meter stick for measuring, push pins to mark important locations for the grid system, and string to create a grid system for the landscape created. The sampling scheme was 6 cm x 6 cm, this is because it's small enough to collect enough data for the activity, but large enough to the point where the amount of data points isn't overwhelming and hard to organize. This sampling scheme was completed by a meter stick being put on the wooden frame of the sandbox, this is where the push pins were put 6 cm apart from one another on every side of the wooden frame. The string was then used to connect the pins that are directly across from each other so then the grid system can be created. Below is an image of the grid system that was created using push pins, lines of string, and a meter stick.
 |
| Figure 1: Grid System for sampling method and Topography that was created |
Since the data collection started at one of the corners of the sandbox a traditional (x,y) coordinate system was set up, where each point correlates with two points, but in this project there is also a z value for the elevation of each. For the zero elevation level, the top of the wooden frame was used as a reference point for zero elevation. To record all of the data, a table was drawn out in a notebook with the exact same grid layout as the figure above so then the elevation data for each sample point can be correctly recorded and organized. The data was entered in this fashion starting at the first point; 1-1-Z, 2-1-Z, 3-1-Z, 4-1-Z, and so on. this made it especially easy to transfer the data from a notebook to a excel spreadsheet.
Results/Discussion:
The final number of sample points collected is 361 points. The minimum number that was collected was -14, maximum number was 4, the mean was -4.7, and the standard deviation was 3. The sampling method that was chosen worked perfectly, as discussed by the group each grid unit was one sample point, and this made it much easier to organize all of the data that was collected. One of the major problems that occurred during this activity was that the lines were producing some slack which made it difficult to take the elevation points because the lines were at the same height of the wooden frame which was designated as the zero point for elevation. This was overcome by tightening up the lines to the required height for the zero point for elevation. Also each grid unit was hard to measure not based on the height of the line but the terrain itself because the entire area within each grid unit isn't flat making it hard to average out the elevation for each individual sample point.
Conclusion:
The sampling method that was used in this activity relates to most other sampling methods and the definition of sampling itself. But in this case it especially relates with the systematic sampling method because of how a grid system was created to acquire the points and not just placing random points out there and taking data from them. It is crucial when doing spatial analysis to have a sampling method, a sampling method allows you to generalize some data of a large area into a smaller sample. There is no possible way to collect every single point of data in a large area, but through sampling every point doesn't need to be collected and the same results are achieved. In relation to sampling spatial data for a larger area this is essentially the same concept but at a smaller scale. Once all of the numbers were analyzed the sampling method chosen did a decent job, each of the sampling points came out with an elevation level, all of the land types were accounted for, there are just some sampling points that could be more accurate. To refine the survey to increase the sampling density, a smaller sampling scheme should be used so than even more accurate data can be collected from each grid unit.
No comments:
Post a Comment