Introduction
This is technically part two of the previous exercise, so we were with the same group members as in Assignment One. We are now looking at the data we have collected from our "field work" in a digital medium. With the data we collected, we can run different types of interpolation methods in ArcGIS. Based on how the models turn out, the groups assessed whether they needed to revisit any areas of their terrain to acquire more accurate data.
Methods
Our Excel spreadsheet has three separate columns for x, y, and z coordinates which can easily be imported into ArcGIS. We were required to run at least five different kinds of interpolation methods. Every method has a different way of acquiring values for each cell and each methods can have a different purpose. It all depends on what the analysis is about and what you are trying to look at.
IDW (Inverse Distance Weighted)
This kind of interpolation method estimates the value for each cell by averaging the values of points of data in neighborhood of each processing cell. If a point is closer to the center of the cell, the greater the influence it has on the overall average.
 |
| 2-D terrain using IDW interpolation method in ArcMap. |
 |
| 3-D terrain using the IDW interpolation method in ArcScene. |
Natural Neighbors
The natural neighbor interpolation method finds the closest sequential values to the point and applies a weight to it based on proportionate areas. It can also be called Sibson or "area stealing" interpolation (Sibson, 1981) (ArcGIS Resources, 2012).
 |
| 2-D terrain using the natural neighbors interpolation method in ArcMap. |
 |
| 3-D terrain using natural neighbors interpolation method in ArcScene. |
Kriging
A geostatistical method known as kriging estimates the surface from a set of scattered points that contain z values. With any interpolation method, one should take care to do thorough attention towards the z values and their spatial behavior.
 |
| 2-D terrain using the kriging interpolation method in ArcMap. |
 |
| 3-D terrain using kriging interpolation method in ArcScene |
Spline
The interpolation method that uses spline estimates values using a mathematical function that essentially lowers the amount of curvature the overall surface has while going through all the input points, and gives a smoother surface as a result.
 |
| 2-D terrain using the spline interpolation method in ArcMap |
 |
| 3-D terrain using spline interpolation method in ArcScene. |
TIN (Triangular Irregular Surface)
This method creates triangles by connecting three of the points of data. The plane of that triangle takes into account the x, y, and z coordinates, and therefore adjusts itself at the proper angle for modelling the elevation.
 |
| 2-D terrain using TIN interpolation method in ArcMap. |
 |
| 3-D terrain using TIN interpolation method in ArcScene |
Discussion
I think our results turned out pretty well from when we did measurements the first time around. The very first time our group went outside to assess our area and our terrain, we had decided to be very accurate with our coordinate system. Because we had marked off 5 cm increments for both the longitude and latitude of our plant box, we were able to acquire precise data the first time
Natural neighbor seemed to be the best representation of our terrain. It is not choppy, yet it does not oversimplify things by smoothing everything over. You can clearly see where we tried to build sharp features compared to gradual ones; other methods do not show this as well. It was insightful to see all five methods and compare them to each other.
Conclusion
The geography part of my education was a positive after though as I was working through my comprehensive geology degree. One important lesson they teach their students when doing field work is to do a good job the first time, because there is no time to go back to an area to acquire more data later. If you are in a professional situation, you cannot afford to be just ok when you do field work. When people depend on you, they will not settle for data that is not accurate enough. Besides that, usually deadlines are restricting your time frame for data collecting.
No comments:
Post a Comment