In the previous lab we created a sandbox that was 114cmx114cm filled with brown sand. We then created a 6x6cm grid made into 19 parts. This was to create a sample area that we could measure for height, which was based on the top of the sand box, to use as data for an excel table to be brought into ArcMap. The term data normalization means to normalize or the processes of organizing data fields. This was important in this lab because to create the 3D model three numbers were needed to be recorded which included an X, Y, and Z. Data normalization is also used to improve data integrity because when it is normalized it is set at a standard, therefore making it easier to find and fix mistakes created in the data. Data normalization also relates to this lab because when taking data from field based surveys there needs to be some kind of normalization to create accurate, smart, and usable data. The data points collected for this lab are used to show the difference in relief from other data points. The best way to go about creating a landscape from this data is to use interpolation which is a tool found in Arcmap. The interpolation creates high values in a raster based on the data points around it. The different types of interpolation include IDW, Natural Neighbors, Kriging, spline, and TIN.
Methods
To start this lab off the first thing to be done was to create a folder specifically for this project named 'sandbox,' for obvious reasons. Next, is to create a geodatabase within the sandbox folder. The next step is to upload the numeric excel file containing x,y, and z data to be added into Arcmap. Next is using the add XY data to create a new feature class. The final step was using the different interpolation tools to create a continuous surface map. The Inverse distance weighted (IDW) (Figure 1) interpolation determines the raster cell value using a linearly weighted combination of sample points. The next tool used is called natural neighbor (Figure 3). This find the closest subset of input samples to a to a query point. It then adds weights to them based on proportions. This method is also known as the "area-stealing" interpolation. The next interpolation tool used is called the Kriging formula. This is based off of surrounding measured values to derive a prediction for an unmeasured location (Figure 2). Spline is another type of interpolation tools. This tool can be thought of as a sheet of rubber that passes through the input points while minimizing the curvature of the surface (Figure 4) The last tool used is called a TIN or triangular irregular network. This method of interpolation connects points together in a triangle shaped fashion (Figure 5). To use the image from ArcScene in Arcmap The only way to do this is to save it as a layer file, and then save the image as a picture. Then open the picture up and put it over the top of the layer file which gives the legend information about the map created. The orientation used is the one most resembling the data originally taken in the survey with the X on the bottom y on the side and the Z as height. Scale is reflected by____. It is important because without scale there is no way to tell how much relief there is present.
Data/Discussion #1
In Figure 1, the IDW method, shows a decent example of what the sandbox looked like. It does actually give the ridge in the bottom right corner more of a multi-point mountain look, and gives the hill in the center a weird middle line extending into the picture. It did not display the depressions well in the top-right of the map either. Each one turned out smaller than it should have been. It also gave the long valley on the left side of the map not nearly as much depth as it looks on the other interpolations. The next method was kriging, (Figure 2) which was actually the worst at showing the heights of the ridge and hill and lows of the depressions and valley. It makes it look almost like it is in a 2D look. The colors make it show where the different structures are, but there just is not really any relief shown. It also really does not show the ridge in the bottom right corner. This interpolation method also is not showing any of the depressions at all in the top right corner. This method is also showing that the Hill is the highest point when the ridge and hill are actually the same high values, therefore this is one of the more inaccurate interpolation methods for this survey. In figure 3 the method of interpolation used is natural neighbors. This interpolation method did a fair job at showing the relief of the sandbox created. This one shows the ridge in the bottom right corner. It shows the hill in the center. The depressions in the top right are clearly visible and are the correct size. The valley on the left side is actually really well detailed in this method and shows what it looked like in our sand accurately. Overall, this is the second best method for this survey. In figure 4 is the spline interpolation method. This one is the overall best one for displaying the survey of the sandbox. It shows all of the features listed above very well. It gives a good display of the relief throughout the model, and it overall looks just like what was created. This is the most accurate method for the landscape. The last method was Triangular Irregular Network. This is not really an interpolation method, but it can be used as one for displaying elevation. This model does a fair job at showing the relief from the hill to the plains to the valley. Although, it does not do as well of a job as the spline does.
 |
| Figure 1: IDW Interpolation |
 |
| Figure 2: Kriging Interpolation |
 |
| Figure 3: Natural Grid Interpolation |
 |
| Figure 4: Spline Interpolation |
 |
| Figure 5: TIN |
Revisit Survey
On the remake of the survey Group 2 went out to conduct a more detailed survey of the upper right corner where the depression are located. To do this we went from every 6x6 cm to 3x6 cm in that area which gave a better recreation of the relief in the top right corner for the depressions along with the ridge being a little more well developed in the bottom right corner. Due to having more data points this is going to make the accuracy of this data more valuable to use for the model. It will be more precise. In figure 6 it shows the ridge in the bottom left and even shows where the sand for the ridge came from when the group built it. This was a little harder to see in figure 1-5, but this one is more detailed and accurate of the survey.
 |
| Figure 6: Spline Revised |
Conclusion
This survey relates to other field based surveys because it teaches how important it is to normalize the data. Without having normalized data it makes it a lot more time consuming than it needs to be. It also relates because the z value does not need to be height. It could be population or number of something else. This makes this lab extremely valuable to be able to comprehend and use for more than one type of survey. It is different because it will not be giving an elevation model, but a different kind of model. It is not always realistic to preform such a detailed grid especially when the area starts to get into the acres of sizes. This can be extremely difficult to go and take a data point when they are each so far apart, therefore it is important to understand the different kinds of surveying, which was discussed in the last blog post. Interpolatin can e used for many different types of data. It could be used to interpolate temperatures for example. It can take the temperature from data points and use the interpolation tools to create a continuous surface of temperatures, therefore almost any continuous surface map can be used with interpolation.
No comments:
Post a Comment