GEOG 586
Geographic Information Analysis

Project 6: Interpolation Methods

PrintPrint

Introduction

This week and next, we'll work on data from Central Pennsylvania, where Penn State's University Park campus is located. This week, we'll be working with elevation data showing the complex topography of the region. Next week, we'll see how this ancient topography affects the contemporary problem of determining the best location for a new high school.

The aim of this week's project is to give you some practical experience with interpolation methods, so that you can develop a feel for the characteristics of the surfaces produced by different methods.

To enhance the educational value of this project, we will be working in a rather unrealistic way, because you will know at all times the correct interpolated surface, namely the elevation values for this part of central Pennsylvania. This means that it is possible to compare the interpolated surfaces you create with the 'right' answer, and to start to understand how some methods produce more useful results than others. In real-world applications, you don't have the luxury of knowing the 'right answer' in this way, but it is a useful way of getting to know the properties of different interpolation methods.

In particular, we will be looking at how the ability to incorporate information about the spatial structure of a set of control points into kriging using the semivariogram can significantly improve the accuracy of the estimates produced by interpolation.

Note: To further enhance your learning experience, this week I would particularly encourage you to contribute to the project Discussion Forum. There are a lot of options in the settings you can use for any given interpolation method, and there is much to be learned from asking others what they have been doing, suggesting options for others to try, and generally exchanging ideas about what's going on. I will contribute to the discussion when it seems appropriate. Remember that a component of the grade for this course is based on participation, so, if you've been quiet so far, this is an invitation to speak up!

Project Resources

The data files you need for the Lesson 6 Project are available in Canvas in a zip archive file. If you have any difficulty downloading this file, please contact me.

  • Geog586_Les6_Project.zip (zip file available in Canvas) for ArcGIS 

Once you have downloaded the file, double-click on the Geog586_Les6_Project.zip file to launch WinZip, PKZip, 7-Zip, or another file compression utility. Follow your software's prompts to decompress the file. Unzipping this archive, you should get a file geodatabase directory (centralPA_gdb.gdb) and an ArcGIS Pro package or ArcMap .mxd. 

  • pacounties - the counties of Pennsylvania
  • centreCounty - Centre County, Pennsylvania, home to Penn State
  • pa_topo - a DEM at 500 meter resolution showing elevations across Pennsylvania
  • majorRoads - major routes in Centre County
  • localRoads - local roads, which allow you to see the major settlements in Centre County, particularly State College in the south, and Bellefonte, the county seat, in the center of the county
  • allSpotHeights - this is a point layer of all the spot heights derived from the statewide DEM

Summary of Project 6 Deliverables

For Project 6, the minimum items you are required to submit are as follows:

  • Make an interpolated map using the inverse distance weighted method. Insert this map into your write-up, along with your commentary on the advantages and disadvantages of this method, and a discussion of why you chose the settings that you did.
  • Make a layer showing the error at each location in the interpolated map. You may present this as a contour map over the actual or interpolated data if you prefer. Insert this map into your write-up, along with your commentary describing the spatial patterns of error in this case.
  • Make two maps using simple kriging, one with an isotropic semivariogram, the other with an anisotropic semivariogram. Insert these into your write-up, along with your commentary on what you learned from this process. How does an anisotropic semivariogram improve your results?

Questions?

Please use the 'Discussion - Lesson 6' forum to ask for clarification on any of these concepts and ideas. Hopefully, some of your classmates will be able to help with answering your questions, and I will also provide further commentary there where appropriate.