Tuesday, October 18, 2016

Assignment 4

Introduction:

Sampling is a key method across all forms of science. It involves the process of gathering precise data/measurements of an often times large study area. Sampling is quite important and it is most important to be precise in sampling geography because maps and data will directly affect peoples lives. Sampling allows for a project to be completed more efficiently, quickly, and can save resources in the process of creating a sample of a study site. Random, Systematic, and Stratified are the 3 types of sampling. Random sampling uses completely random chance to select points, it is entirely unbiased. Anything can be selected, but it can also be a bad choice because it can leave large areas un-sampled. Systematic sampling uses a regular interval to collect samples. The most common would be a grid. Systematic sampling is reliable and straight forward, however it can be biased. Stratified sampling is final method, it uses known groups of a study and collects data points relative to the size of the groups. This method can generate accurate data, but if the size of the groups are unkown the data can become skewed.

The lab's objective is to accurately construct an elevation surface of a terrain. The study will be completed using a 1 m x 1 m sandbox that students will construct a landscape within using sand to create a unique terrain. This terrain will then be sampled using the sampling method of choice by the students and entered into a spreadsheet to be used for creating a DEM in ArcGIS.

Methods:

Our group chose to use a a stratified sampling technique with a systematic sampling grid to create an easier sample collection. This was most useful when faced with time constraints. The method is similar to a systematic sampling method, but the shape of the terrain acted as "groups" within the total area. Areas of higher and lower elevation had higher more points. The sample plot is located near UWEC's Science building L.E. Phillips Hall, Figure A below.


Figure A, the completed sandbox terrain with a forest in the east, prairie lands to the west, and other random terrain in the center.
Materials used:


  • String
  • Meter Sticks
  • Data Collection Notebook
  • Thumb Tacks
  • Pencil
  • Samsung Galaxy s6 Edge
The sampling scheme used a grid with 10 cm spaces between each reference point on the XY axis. Sea level (0 cm) was the top of the box. Strings were then set up as a spatial reference point at the x and y points to lay over the grid. This allowed us to create a stratified system of uniform groups in the overall model. See Figures B and C below.




Figure B, the grid lines were drawn in to take advantage of the soft sand.

Figure C, shows the tacks displaying areas of higher relief.

After the grid lines and tacks were in place the our group used a meter stick to measure areas of relief. The values were entered directly into our drawn sampling grid. This created a system that was quick and easy to execute when we were short on time.


Results/Discussion:

178 points were collected in the 1m X 1m sandbox using the stratified method; ranging from very low at the deepest depression to quite high at the highest peak. Throughout most of the prairie lands in the west, the elevation stayed roughly the same. The sampling method was adequate for what was needed to be done, and it became even better when the grid depression was made using the strings.

Minimum: -16 cm
Maximum: 12 cm
Mean Elevation: -4.71 cm
Standard Deviation: 4.64 cm


Conclusion:

The stratified system employed used a solid stratified system that allowed for a systematic approach to make data collection easier and more reliable. One of the most important things to consider when sampling is time and resources available, because as seen in this lab, there were little resources and little time, probably just like the real world. This sampling technique is reliable enough to be used again and could produce good data over a large sample area. In the next assignment a DEM will be created from our data on ArcGIS and it will prove if our sampling method was accurate and successful. 

No comments:

Post a Comment