Lab 4: Normalizing Geodatabases

• Introduction: 
• In this week's lab we learned how to create our own geodatabases on ArcCatalog and how to normalize data entry for feature classes to make recording attribute data in the field with GPS devices much easier plus more accurate and efficient. Collecting attribute data along with GPS points can give a much more in depth view of how smaller scale components such as dark pavement on a road or a specific building have on a wider area being studied. This can be pertinent to ones research in providing a detailed description of an exact location in a collection of points.
• Study Area: 
• The area where the points of data collection will be taken is the lower campus of U-W Eau Claire, Wisconsin, U.S. A campus aerial photo with zones is also added to the Geodatabase for visual ground references. On Tuesday, February 23rd of 2016 at around 5:30 pm the atmospheric conditions were  foggy and cloudy with light rain and a temperature readings around 45 degrees Fahrenheit.
• Methods:
• In ArcCatalog a new geodatabase was created to take microclimate recordings of UWEC's lower campus. A point feature class was added with fields including temperature, humidity, wind direction, speed and several others. These fields were then normalized to fit the data they would soon contain. Normalizing fields helps with data entry errors which can throw off an entire study. Some fields get normalized by an acceptable range of numbers while others can be created to have a selection of text to choose from. It is important to have recognizable field names for ease of data entry and to have the data type pertain to the specific field. This way data entry errors can be cut down. For example if a temperature field is being created and the it is known that the temperatures in the area being record won't reach or exceed certain numbers, a range can be set. Therefore if a temperature outside of the range is incorrectly inputted then it will be rejected. This can be done in the geodatabases feature class properties window. (shown below in figure 1.) 

Figure 1. a view of the fields in the feature class in ArcCatalog

•  After normalizing the fields the geodatabase is opened up in ArcMap where it is paired with an aerial photo and deployed through ArcPad to the GPS device, in this case the Trimble Juno.
• Once the geodatabase was transported to the GPS we were able to pair into groups to collect sample readings around campus. After a couple readings were taken the GPS was connected to ArcMap once again where the new points and attribute table could be seen.
• Results/Discussion: 
• As this lab was in preparation for the next only two data points were recorded. While the readings show little variance on a large scale, It can be noticed that proximity to buildings and elevation play a role and truly have an affect on a surrounding area.  Being able to take multiple data points in an area and combining that information can show an overall glimpse of an area and can show variances in specific spots as well.
• Conclusions:
•  What makes data collection with mapping GPS different than just gathering a point is the fact that you truly pinpoint a specific location tied to the Geographic Coordinate System and have the data associated with that point linked together. Proper database setup and data normalization ease the process of data collection and can help prevent errors.
• My first try creating the geodatabase I realized I had actually created a polygon feature class instead of a point feature class and soon found out I wouldn't be able to record any data that way. I was then able to quickly remake a new geodatabase correctly and catch up before I fell behind. The second time creating the geodatabase I also included a notes field which was most helpful once in the field and something I won't forget in the future.

Lab 3: Navigation Maps

Intro: 

The goal of this lab was to create two maps that will be used in a later lab for navigation purposes

Methods:

After given access to a geodatabase filled with data on the specific area of Eau Claire Wisconsin selectivity was used to decide which data was helpful and which was unnecessary to create a map easiest for navigation in the field. Two maps needed to be created, the first with a UTM grid and 50 meter spacing and the other with the area's Geographic Coordinate System Decimals Degrees. A pace count was also recorded for additional assistance in the field, every right step was counted over a 100 meter distance.

Results:

In this first map the main focus is the 2 foot contour lines that show the surface elevation of the area. A aerial photo was placed in the background for visual ground references. This map also includes the UTM grid, pace count and a 50 meter grid.

The second map uses the deep purple to identify the low areas in the terrain and a defined aerial photo for navigational references. This map also includes GCS decimal degrees. Both maps are in the North American  Datum 1983 UTM zone 15 north coordinate system and projected in a Universal Transverse Mercator. The UTM coordinate system that complies 60 6 degree longitudinal zones with little distortion around the earth. Eau Claire Wisconsin happens to be in the 15 north section.

Lab 2: Survey of Terrain Surface - Revisted

•   Introduction

• After completing the previous lab where a terrain was created and surveyed, This lab goes through the process of taking those survey points, putting them into GIS software and creating 3D maps using the 5 main interpolation methods. Then after comparing each method one is chosen by its closeness to the actual created terrain and additional sample points are taken and added through a stratified sampling method to create better likeness to the real terrain. 

Figure 1. data points in orange with new points in purple

•   Methods 
• Using ArcScene, the excel files of the X Y and Z coordinates that were surveyed in the previous lab were imported in and converted to points on a grid by making a point feature class and then brought to life by these 5 different interpolation methods  by creating a continuous surface along the points. This is done by simply inputting the point feature class into the interpolation methods found in the ArcToolbox. Below are the 5 interpolation methods used:
• IDW - (inverse distance weighted) This method follows the 1st rule of geography which states that things closer together are more alike than things farther away. It makes its prediction by measuring the values of the points closest to it. Seen in figure 2.
• Kriging -  Creates it's estimated surface by looking at a larger picture of the sample points and finds a more generalized value by taking in the other high and low values. Seen in figure 3.
• Natural Neighbor - Puts weighted values on the nearby sample points based on their proportionate areas and uses those subsets to create its surface. Seen in figure 4.
• Spline - generates a continuous surface by passing through all the data points  while remaining above a calculated minimum curvature. Seen in figure 5.
• TIN - (Triangular Irregular Networks) which are vector based and create a surface by triangulating the vertices of the sample points. Seen in figure 6.
• Maps were then created of each method using ArcMap.  Areas that needed extra attention and additional sampling had more data recorded by recreating the previous landscape and new points 4cm from the points to the side of it.

•   Discussion 

Figure 2. IDW Interpolation of the surface

Figure 3. Kriging Interpolation of the surface

Figure 4. Natural Neighbors Interpolation of terrain surface

Figure 5. Spline Interpolation of terrain surface

Figure 6. TIN Interpolation of terrain surface

After adding the additional points a final map was created using Spline Interpolation.

•   Conclusion 
• Over the past 2 labs a terrain including a ridge, hill, depression and valley were created out of snow in a planter box in the courtyard of Phillips science building at the University of Wisconsin, Eau Claire campus. After the creation of the landscape the systematic sampling method was was chosen to record points over a grid created with strings and pins over the landscape. The sample points were transported into Excel, ArcScene and ArcMap to create a 3D visual representation of the landscape created in order to gain a better understanding of different methods used in geospatial fieldwork and techniques, how to use and work with them in addition to knowing how and when to use the proper method or technique. Spline Interpolation seemed to fit the real terrain the best and the additional points added to the overall likeness.

Lab 1 : Survey of Terrain Surface

• Introduction
• In this first lab the class was broken down into groups that would create a terrain surface in the snow to be surveyed through a sampling technique to gather points that could later be transformed into a 3d map and to gain a basic understanding of common sampling techniques over a landscape. Sampling is a way to gather information in a fair manner on a smaller amount of a whole area being studied to answer questions. There are three main types of sampling methods:

• Random - sample points generated in an unbiased manner
• Systematic - sample points selected by measurements
• Stratified - selected areas with either random or systematic techniques applied
• Methods
• A systematic sampling technique was chosen to maximize the ultimate number of fairly distributed sample points. This method was chosen by the group because it was thought to best way to take the most accurate samples in particular lab. The location of the created landscape was in a planter box in the  back courtyard of Phillips science hall at the University of Wisconsin Eau Claire, USA on January 29^th at 2:30 pm. The materials used in this lab  consisted of strings, pins, measuring tape, yard sticks, a clip board, pen, paper and Microsoft Excel. The spacing determined for the sampling was done every 8 centimeters on a y and x axis. The plot the created landscape was in was 112 cm wide by 88 cm tall. A grid was made using the strings and  pins after being measured out as accurately as possible.(see figure 1.)

Figure 1. The created terrain and string measurements

• The group also decided to have the height of the planter box to be the zero elevation level. And simple method was chosen for recording the data. Using X Y and Z columns with each sample points measurement on the grid combined with its height, Z and was entered into Excel.

Figure 2. Group members taking measurements of landscape

• Results/Discussion


• The total number of samples recorded was 181. Out of those 181 samples, the minimum z value was -10 while the highest value was 19.5 with a mean of -.5. While we used a systematic sampling method, the other considered method was stratified-systematic but the group found that systematic was the best option to complete the objective to cover the whole terrain equally. Originally the plan was to survey an extended length as a 112 by 112 cm block but cold weather and an approaching sunset lead to a group decision in cutting back 3 sample rows at the edge of the landscape since they were thought to contribute little to the overall project since it was the drop off edge of the ridge. ONly running into mainly two issues, the first problem the group ran into was in the ability to construct or modify our landscape. The weather in the day prior melted the majority of the snow and with a cold night it was frozen and un-formable. Therefore the group had to use a previously formed landscape by another group. The second was how to as accurately as possible measure the frozen landscape features above the zero elevation. After digging through the box of available materials provided for this lab and trying several different methods. The chosen was to measure the height of the feature from its base and subtract the difference between that and the distance from the zero elevation.

Figure 3. Group members measuring out 8cm for systematic sampling method

• Conclusion


• The sampling done in this lab worked out well and represented the landscape as a whole. Sampling in spatial areas can be a good way to visualize boundaries, groupings and can show how dynamic an area is in a whole. While this was on a very small can it can display how sampling larger areas can work and issues that can arise with it and how to select a good sampling method based on your objective. Overall the survey performed did a good job of sampling the area.