Sampling is a technique that is used to get a general idea of something by only taking data from certain areas rather than the whole. There are three main ways to sample an area, random, systematic and stratified. Random sampling is just as it sounds, each data value has a equal chance of being selected, a efficient way to do this is with a number generator. Systematic sampling is done by taking a sample at a set interval, for example every two inches, or maybe every 5th house. The third and final sampling method to be discussed is stratified. There is both systematic and random stratified sampling, and each has there own applications. The goal of the week one activity was to encourage students to work together in order to grid out and survey a student created terrain. From here, the data was collected first on paper and later entered into Microsoft Excel. The size of the sand box was approximately one square meter, each group was tasked with creating a unique digital elevation surface. In the beginning of the lab there were no official guidelines on how to begin, the sandbox was shown to the class and the tools were laid out. The ultimate objective of this lab is to collect sample data, enter it in to Excel and in later weeks the data will be analyzed in ArcMap.
Methods
Our study was completed on Monday, January 30th from 3:00 pm until 4:30 pm. The square meter sand boxes were located south of the steam near the garden and across the road to the east from the Philips garage. The sampling technique that my group decided would be the most accurate was the systematic method. Due to the fact we had a nearly perfect square meter to work with, the systematic method of sampling works really nicely because there is no discrepancies when measuring the grid on the box. To assist in data collection and documentation a variety of tools were used including tape, string, thumb tacks, notebooks, meter sticks and a cell phone. The group decided that zero elevation, which is sea level was going to be flush with the top of the wood on the sandbox which is where the string was attached. Negative values were referred to as below sea level and positive values accounted for above sea level. The sandbox had a grid that was 20 x 20, this equates to 400 data points collected. The data was collected from left to right, starting at the bottom left of the sandbox which was coordinate 0,0. Using a coordinate system helped to keep the data organized.
Results/ Discussion
As stated above, the sandbox had a grid that was 20 x 20 and there were 400 data points collected from the sandbox in order to ensure enough data points that the changes in elevation would be noticeable on the map, but not so many points that this exercise became overwhelming. The data points collected were approximately 5 cm or 1.96 inches apart. The terrain elevation data that was collected was done in centimeters, the max value above 0, or sea level was 4 an the minimum value, or below sea level was -8.
There were a few issues that did occur during the lab. For starters some of the sand was frozen solid which dictated partially where our changes in elevation had to be placed due to the fact that we did not have a shovel. Next as data collection was in full swing, the sting used to grid out the sandbox would sometimes loosen up. A better way to do this would have been with three inch nails that would not pull out rather than flimsy thumb tacks, and another improvement would have been to use string that does not have any pliability, that way as the surveyor there is no concern that the string will stretch and become slack. Also looking back, additional sand would have helped to make the exercise run smoothly. The frozen ground, coupled with a half full sand box accounted for much of our values being below 0.
Conclusion
Sampling is effective in a spatial setting because with enough sample points an accurate map of elevation can be created. In a way the method used to grid out this sandbox is similar to how the Public Land Survey System (PLSS) maps out land into 40 acre squares, it is time efficient and accurate. When looking at the numbers that were gathered throughout this exercise, it certainly would not have been a bad idea to add a little more data to our sampling. What that means is making the grid 25 x 25, which would be 625 data points. The smaller distances between each elevation taken would allow for better visualization of the changes in elevation. That being said, the data collected was certainly adequate, as it did show the changes in elevation, it just could have been more detailed had the smaller grid been implemented.
Sources
- http://www.rgs.org/OurWork/Schools/Fieldwork+and+local+learning/Fieldwork+techniques/Sampling+techniques.htm
No comments:
Post a Comment