Increasingly, slope analysis has become a more common request for projects not only for the planning of fieldwork and report production, but also for job proposals. If areas of a high percent rise in slope can be identified, then such areas can be eliminated for prospective subsurface testing thus bring potential cost estimates down and the likelihood of winning a job up. The value of such analysis is appreciating greatly from project to project.
To perform an analysis of slope using ArcGIS, the Spatial Analyst license is required. The Spatial Analyst toolset is included with the license and contains several tool kits and tools, a few of which, are used in this process. Before the use of such tools, the elevation data need to be acquired for the area of interest. Elevation data comes in many forms- LAS point cloud, TIN (Triangular Irregular Network), DTM (Digital Terrain Model) and DEM (Digital Elevation Model). For our purposes, we will stick with the Digital Elevation Model as it is now widely available and usually comes in the form of a georeferenced raster tile that it easily inserted into a project .mxd and is compatible with the toolset we will use. There are many resources online to find elevation data for your area of interest (AOI). Many websites now offer a data viewer or web map that allows you navigate to and draw your AOI and returns lists of available datasets. Below are a few links to such sites. http://gis5.oit.ohio.gov/geodatadownload/ https://coast.noaa.gov/dataviewer/#/lidar/search/ https://viewer.nationalmap.gov/advanced-viewer/ http://maps.psiee.psu.edu/ImageryNavigator/ http://maps.psiee.psu.edu/preview/map.ashx?layer=3188 In other cases, a tile index is provided in the form a shapefile and the DEM is downloaded from a list of files that match the index tile. It is often difficult to find a DEM tile that completely covers the AOI. In this case, it is necessary to download multiple tiles. Performing analysis on multiple tiles is more time consuming and cumbersome, so it is best practice to merge the tiles into one raster. This can be done by using the Mosaic To New Raster tool found in the Raster Dataset toolkit in the Data Management toolbox. Upon opening the tool window, drag the group of downloaded DEM tiles from the .mxd Table of Contents or ArcCatalog to the Input Rasters field. Set the output location to a Raster folder and name the aster dataset accordingly using a .tif extension (TIFF) as that is likely the same file type as the input rasters. Change the pixel type to 32_BIT_UNSIGNED and set the number of bands to 1. Everything else can be left to its default setting. Once the tool has run, you will have a rectangular raster with cells outside of the original tiles but within the extent of the group set to a value of 0. Now is a good time to clip this raster to a smaller size with boundaries closer to the area of interest. A buffer of you area of interest or project area at a desired distance would be a good place to start. To clip a raster you need to return to the Data Management toolbox and to the Raster Processing toolkit. There you will find a Clip tool for raster clipping. Drag the Mosaicked raster to the input raster field and the AOI buffer (or alternative) to the output extent field. Check the box that reads “Use Input Features for Clipping Geometry” and assign the output raster dataset to the raster folder of your project. The result is a raster shaped to the clipping feature, much smaller in file size and quicker to run the following processes on. Now that the elevation data has been minimized, the Spatial Analyst toolset can be used. In the Spatial Analyst toolbox open the Slope tool in the Surface toolkit. Drag the clipped DEM raster to the input raster field and set the output raster accordingly. *Important Note: When running any raster processes from the Spatial Analyst toolbox and error occurs when the output is set to save on the company server. Whenever running a raster process from this toolbox save the output to a place your desktop. It is best practice in this case to keep a “Raster Processes” folder on your desktop for such occasions. Most of these processes are interim and will likely be discarded after the final process. Set the output measurement to “PERCENT_RISE,” method to “PLANAR” and keep the z-factor as 1. It is possible that the DEM being used has different vertical measurements than horizontal, but highly unlikely. In any case, it is always a good idea to check the original metadata from the downloaded tiles to look at the unit specifications. If the DEMs vertical units do not match the horizontal units read the tools instructions for adjusting the z-factor value. The output of this process is a raster dataset that is a replica of the input raster except the cell values now represent a percent rise in slope from its neighboring cells rather than an elevation value. Do not be alarmed when you see the highest value of the raster exceed 100. One hundred percent slope would account for a 45-degree angle, as it is a one to one ratio in elevation increase. Percent rise can be measured to infinity depending on distance and increase in elevation. Next, the values within this raster need to be extracted and put into a tangible vector dataset. Before that can happen, the values must be converted into integers to minimize the possible number of values represented in the resulting vector. This is accomplished by using the Int tool that can be found in the Math toolkit of the Spatial Analyst toolbox. Simply input the slope raster and name the output by referring to the tool or process used to create it such as “IntSlope”. This, also being a raster process, must be saved locally to a computer and not on a network server. The integer output can now be converted to a vector polygon by using the Raster To Polygon tool found in the From Raster toolkit of the Conversion Tools toolbox. Depending on the size of the inputted integer based raster, this process may take a while. Set field to VALUE and keep “Simplify Polygons” checked. The output of this process is a vector polygon that can be saved to the network server in a shapefiles folder. After the slope polygon has been created, the layer can be turned off in the Table of Contents in the ArcMap document. It will take a long time for it to draw on the page and is not necessary for the next step. Open the attribute table of the resulting polygon layer and under the Table Options menu choose Select by Attributes. Create a new selection by choosing “gridcode” as the WHERE clause. The gridcode field represents the slope value for each raster cell produced by the Int and Slope tools and is now grouped by each mutipart of the polygon. Depending on the requirements for the project, the slope percentage cut off may be between 12 and 15 percent. For example, if the cut off is 15% the WHERE clause should read: “gridcode” <=15. This would select all the features within the layer that have a gridcode of 15 or less. Export the selected features as a new shapefile of areas that are of 15 percent or less slopes. Perform a similar selection by attribute with the WHERE clause as: “gridcode”>15. This will select all the features of greater than 15 percent slope. Export these selected features as a new shapefile of areas of greater than 15% slopes. The final step is to dissolve all the rows in each new polygon shapefile so that it each one is a single feature. Add an integer field to each shapefile called “dissolve”. Using the field calculator set the value for every row in this field to 1. Open the Dissolve tool found in the Generalization toolkit of the Data Management Tools toolbox or in the Geoprocessing dropdown menu on the ArcMap task bar. Input the shapefile and set he output accordingly. Choose the dissolve field you added as the dissolve field and leave everything else to default. Repeat this process with the opposite shapefile and the result is to single feature shapefiles of testable and untestable areas. It is a matter of personal preference to merge the two together as a single shapefile.
0 Comments
Leave a Reply. |