Ex 8 Raster Modeling

Raster Modeling

Goal and Objectives

     The goal of this final exercise is to use various geoprocessing tools, such as euchlidean distance, reclassify, raster calculator, filter, topo to raster interpolation, and viewshed, to create a suitability model and a risk model for sand mine locations. The suitability model will include factos such as geology, landcover, slope, distance to rail terminals, and depth of water table. The risk model includes factors such as distance from streams, farmland quality, distance from populated areas, distance from schools, and distance from designated wildlife areas. After the two models are created I will combine them using raster calculator and remove any areas that fall in exclusionary zones, such as residential areas. Finally, I will calculate a viewshed of cemeteries in the study area using a DEM to determine ares that are visible from cemeteries.

Methods

     The process of locating the best location for sand mines is broken down into two main steps. First I will create a suitability model which determines the best location from the perspective of sand mining companies. Suitability is classified as 3 for high suitablity, 2 for medium suitability and 1 for low suitability. Then I will create a risk model to determine locations that pose a significant threat to the general public and the environment. Risk is classified as 1 for high risk, 2 for medium risk, and 3 for low risk.

Suitability Modeling

Figure 1  Data flow model of the process used to determine suitable locations from the perspective of a sand mining company.

Figure 2 Factors that were used to create the suitability model and the ranks that were used to classify each criteria.

     Geology is an extremely important factor in determine where sand mines should be located (Figure 3). Since fracking operations need a very specific grain size it is important to place a mines in those areas. In this case the most desirable geologic formations in the Wonecon Formation and Jordan Formation (part of the Trempealeau Group). Therefore, these two geologic formations are classified as a 3 (high suitability). Considering the location of active mines I determined the Eau Claire Formation, Ancell Group, and Prairie de Chien Group also contain sands suitable for fracking operations, therefore they were classified as 2 (medium suitability). Finally, I noticed the Tunnel City Group only contained 1 mine in the area. Based on that information I classified this geologic formation as a 1 (low suitability). 

Figure 3 Map showing the suitability of geologic formations. 

     Landcover is also an important factor in determine suitability of sand mine locations (Figure 4). Areas that are heavily forested will require a significant amount of money to remove the vegetation while shrubland and barren lands will require little to no extra cost to remove vegetation. Therefore, I classified barren land, shrubland, and herbaceous land as having the highest suitability (rank of 3), forested land as medium suitability (rank of 2), and wetlands, developed areas, and open water is low suitability (rank of 1).

Figure 4 Suitability of landcover in Trempealeau County, WI based on the cost require to remove vegetation.

      Distance to rail terminals is also an important factor in assessing sand mine suitability (Figure 5). It will cost the sand mining company a lot less to export the sand if the mine is located closer to rail terminals. Using the Euclidean distance tool and natural breaks I determined the most suitable location was located within 8000 meters of a rail terminal (suitability of 3). Medium suitability (2) was located from 8000-16000 meters from rail terminals. Any location farther than 16000 meters from a rail terminal has low suitability (1).

Figure 5 Suitability ranks based on distance from rail terminals.

Slope is also an important factor in suitability (Figure 6). Areas that have more gradual slopes are easier and cheaper to work on. I calculated slope using the slope tool and a digital elevation model. I then ran a 3 cell by 3 cell low pass filter and reclassified areas with less than 9.8% slope as having high suitability (3), areas between 9.8% and 23.5% slope as having medium suitability (2), and areas that have greater than 23.5% slope as low suitability (1).

Figure 6 Areas with high slope are less desirable than areas with gradual slope.

     The last factor that will be taken into account in determining the most suitable areas for sand mining companies to open new mining operations is depth to water table (Figure 7). Since sand mines require a large amount of water in the processing of frac sands it is imprortant to be located in areas that have shallow water tables to reduce to cost of pumping. I began by downloading water table elevation data from the Wisconsin Geologic Survey website. This data had to be converted into raster format using the topo to raster tool. Then I had to convert the elevation from feet to meters. Finally, I subtracted the water table elevation raster from the digital elevation model to get the actual depth of the water table. I then reclassified the water table depth into 3 classes. Anywhere with a water table less than 10 meters has high suitability, between 10 and 30 meters is medium suitability, and anywhere with a water table depth of over 30 meters has low suitability.

Figure 7 Depth of water table is important in determining where sand mines should be located.

      Using the raster calculator tool we can add up the rankings of the previous 5 factors to give the overall suitability of sand mine locations from the perspective of a sand mine company (Figure 8). Notice that areas with higher depth to water table also have higher slope, more forested landcover, and less suitable geology.

Figure 8 Suitability of sand mine locations from the perspective of a sand mining company.

Risk Modeling

     The goal of the second part of this exercise is to create a risk model based on factors that will be affected by sand mines from the perspective of the community/environment. The most important factors that will be used in determining the risk model include distance from streams, quality of farmland, distance from populated areas, distance from schools, and distance from designated wildlife areas. Areas that have a high risk of being affected by sand mines are given a ranking of 1 and low risk areas are given a ranking of 3.



Figure 9 Data flow showing steps used to create the risk model

Figure 10 Most important risk factors and how they are ranked.

     A very important factor in determining the location of sand mines is the risk imposed on streams. Sand mines could lead to an increase in sediment load in streams which could be detrimental to stream health. Therefore, it is imperative to keep sand mines at a safe distance from streams. Risk ranks were determined using natural breaks of Euclidean distance from streams. The highest risk (1) is located from 0-637 meters from streams. Medium risk (2) is located between 637-1928 meters from the stream. Finally, the lowest risk (3) is located 1928+ meters from streams.

Figure 11 Distance from streams plays an important factor in determining possible sand mine locations that have the least risk to the environment.

Another important factor in determining sand mine suitability is the risk to prime farmland. Since Wisconsin uses a significant amount of cropland to provide feed for livestock and for many other uses it is important to protect prime farmland. Therefore, areas that contain prime farmland have the highest risk factor, while areas that are not prime farmland have the lowest risk. Areas that could possibly become prime farmland if drained have a medium risk value (Figure 12).

Figure 12 Risk associated with quality of farmland.

      Distance from residential areas is another important factor. Since there is a lot of noise and dust caused by sand mining operations it is important to keep sand mines at a safe distance. Therefore, less than 640 meters from any populated area has a high risk factor, between 640 and 3049 meters is considered medium risk, and anywhere over 3049 meters away from populated areas is considered low risk (Figure 13).

Figure 13 Risk associated with distance from populated areas.

      Similarly to populated areas, dust and noise can impose significant risk to school zones. Therfore, it is important to keep sand mines away from schools. Using the sites feature class I selected schools and exported them into the geodatabase. Then, using Euclidean distance to set up a buffer around each school I determined areas that pose the least risk to schools. Areas within 1343 meters of schools have a high risk, areas between 1343 and 2329 meters from schools have a medium risk, and areas over 2329 meters from schools have low risk to school children (Figure 14).

Figure 14 Risk of sand mines to schools

     Lastly, I find it importnt to keep sand mines away from designated wildlife areas to reduce disturbances to important wildlife. Therefore, I used Euclidean distance to create bufferes around designated wildlife areas and reclassified the buffers based on risk to the wildlife. Any area within 1500 meters poses a high risk to the wildlife, areas between 1500 and 3000 meters away pose medium risk, and areas over 3000 meters pose low risk (Figure 15).

Figure 15 Risk to designated wildlife areas

     Using raster calculator we add up the 5 factors above to determine overall risk (Figure 16). Areas with higher risk are less desirable for sand mine locations while areas will low risk are better locations for sand mines.

Figure 16 Overall risk on people and the environment from sand mining

Overlay

     Finally, in order to determine the overall suitability of sand mine we use raster calculator to add the suitability model with the risk model (Figure 17). 

Figure 17 Best locations for sand mines based on suitability and risk.

    

     We must also note that developed areas and open water must be excluded from the suitability model. In order to do this we reclassify landcover based on a pass/fail grade. Areas that are developed or are open water fail and are given a 0 for suitability since sand mines cannot possibly be located there. Any other area is given a 1 since sand mines can be located there. We use raster calculator to multiply the Final Suitability (Figure 17) to the Excluded Landcover raster (Figure 18) to give the Overall Suitability (Figure 19).



Figure 18 Any areas where mines cannot possibly be located, such as cities or lakes, are given a failing grade of 0 while areas were sand mines could be located are given a passing grade of 1.



Results and Discussion

     As you can see from Figure 19 there are a lot of suitable areas for sand mine locations. The northwester portion of the study area has the largest area of suitable land available.

     

Figure 19 Overall best locations for sand mines. Red areas are the least suitable for sand mine locations.

      Other factors could also play a role in determining sand mine locations. For instance, it would not be a good idea to have sand mines located where they are visible to cemeteries or other ceremonial sites. To exclude these areas we develop a viewshed using the location of cemeteries and the Digital Elevation Model of the area. 

Figure 20 Viewshed of cemeteries in the study area.

Conclusion 

     As you can see from Figure 19, there are plenty of viable locations for sand mines to be located. In order to determine the best possible location one must consider several different factors and determine which of those factors are the most important. On the other hand, some factors are going to be ignored when determining the location of sand mines. For example, nobody wants sand mines to be visible from cemeteries but there are more important factors such as stream health or risk to schools and other populated areas. 

 

 



Exercise 7: Network Analysis

Exercise 7: Network Analysis

Goal/Objective

The most controversial aspect of sand mining in West Central Wisconsin is the effects large trucks will have on local roads. The cost of road construction falls directly on the taxpayers in each county and not on the sand mining companies. Therefore, it is important to calculate to total cost of road maintenance that each county will have to pay for road upkeep. This exercise will help become familiar with some of the network analyst tools ArcMap has to offer and to give a rough estimate of the total cost each county will incur due to the transportation of sand.

Methods

The first step in the process was to write a python script to select mines that meet certain criteria. In this case we want mines that are active and are more than 1.5 km away from any railroad. After these mines were selected they were exported into the geodatabase. Next, we ran the network analyst tool to calculate routes, using ESRI Street Maps USA, from the sand mine to the closest rail terminal facility using the "New Closest Facility" function. After the routes were calculated we exported them into the geodatabase. Exporting the routes created a shapelength field which will be used to calculate distance. Next, we intersected the new routes feature class with county boundary feature class to get the total distance traveled per county. Finally, we used the predefined criteria of 100 trips per mine at 2.2 centers per mile to calculate the total cost each county would incur. Figure 1 below is the data flow model that describes the steps to calculate total cost per county.

Figure 1 Data flow model that shows the steps used to calculate total distance traveled and total cost per county.

Results

Figure 2 shows the routes that were created using the "New Closest Facility" function in the network analyst toolbar. Each mine was routed to the closest rail terminal.

Figure 2 Routes used to transport sand from mines to the closest rail terminal

Table 1 and Figure 3 show the total cost each county would incur if there were 50 truckloads of sand to the rail terminal at 2.2 centers per mile traveled. As you can see Chippewa County has the largest cost at 452.59 dollars. It is interesting to note that although Eau Claire County does not contain any sand mines they will still have a cost of 205.06 dollars.

Table 1 The total distance traveled by trucks in each and the total cost associated.

 

Figure 3 Map showing the total cost each county will incur due to sand transportation.

 

Conclusion

In conclusion, the total amount each county would have to spend to maintain roads is significantly lower than I originally thought it would be. However, this is a rough estimate because not every mine will be using the routes as specified. Some mines will ship their sand to other processing facilities in the area. Another factor that could lead to error is that not every mine will only ship 50 truckloads of sand. Some mines are significantly larger than other and will need more trucks to transport all the sand to the rail terminals. Also, the 2.2 cents per mile traveled may not be accurate. If this price per mile traveled is determined to be higher the cost per county will increase dramatically. Therefore, this lab was interesting to learn how to use different tools ArcMap has to offer, but the results may not necessarily be very accurate.

 

Exercise 6: Data Normalization, Geocoding, and Error Assessment

Exercise 6: Data Normalization, Geocoding, and Error Assessment

 

Goals and Objectives

     The goal of Exercise 6 is to become familiar with the process of table normalization the different methods of geocoding addresses. To accomplish this goal we will be examining a table containing the most up-to-date sand mine locations in West Central Wisconsin provided by wisconsinwatch.org.

Methods

     The first step in any geocoding task is table normalization. When downloading tables from online sources the tables frequently contain errors in the address field which prevents it from being geocoded properly. Table 1 shows an example of the sand mine table that was downloaded from wisconsinwatch.com. Notice several of the address fields contain street addresses and Public Land Survey System coordinates, while others only contain PLSS coordinates. Since the geocoding tool will not be able to process addresses based on PLSS we need to normalize the table. To do this we create a new address field, city field, state field, and zip field in the table to fill in the information correctly. Entries that only contain the PLSS information will be left blank at this point and the location will be manually entered later. An example of the normalized sand mine table is found in Table 2.

 

Table 1 Table of sand mine locations that was downloaded from wisconsinwatch.org. Notice how the address field contains a large variety of address types from street addresses to PLSS.

 

 

Table 2 Table of sand mine locations after normalization. Notice how the address fields have been split up into address, city, state, and zip fields.

 

 

     After the table was normalized to include fields for Address, City, State, and Zip we were able to input the data into the geocoding tool in ArcMap10.2.2. This tool selects the address that best matches the input address and places the point in that location. Mines that only contain PLSS information are manually inputted. 

 

Results

      Figure 1 shows the result of my geocoded mine locations against comparison mines that were geocoded by my classmates. As you can see only a small fraction of the total geocoded mines were coincident with the comparison mines. Mine locations differ anywhere from 177 feet to 11 miles (Table 3)depending on the geocoding method used.  

Figure 1 Map showing the location of my geocoded mines compared to the same mines geocoded by classmates.

Figure 2 A closer look at the error associated with geocoding. In theory, the comparison mines should be coincident to the geocoded mines.


 

Table 3 The distance between mines with the same unique ID. Only 5 of the mines were coincident with the comparison mines, while the rest have a large range of distances between them.

 

Discussion

     With any GIS project there is going to be inherent error and operational error. In this case the inherent error comes from projecting complex real world features onto a 2-dimensional surface. Another important source of inherent error is caused by the geocoding software itself. The geocoding tool breaks down street segments and divides the segment by however many parcels are on that street. This gives every parcel along that segment the same size, which is rarely the case. The geocoding tool places points based on the average of the parcel size. Therefore, addresses may be geocoded to neighboring parcels.

 

     The largest source of error in this exercise comes from operational error. Operational error consists of incorrect data entry into the table or incorrectly digitizing the location for mines that only had PLSS or a range of addresses available. Incorrect digitization led to large differences in the location of geocoded mines and comparison mines. Many of the mine locations were not visible on basemaps so the actual location was unknown.

 

Conclusion

     In conclusion, the process of downloading data from online sources can be very frustrating if the data is not entered correctly. Tables must be normalized before address data can be geocoded. Even after normalization there are going to be some points that must be entered manually. Even if the geocoding tool can automatically select an address it may not select the correct address. Therefore, it is important to check you data after geocoding to see whether each point is located in the correct spot.

Sources

Lo, C.P., Yeung, A.K.W. (2003) Concepts and Techniques in Geographic Information Systems. Pearson Prentice Hall.



Exercise 5: Data Downloading, Interoperability, and Working with Projections in Python

 

 

Overview

     Downloading data from different sources is a necessary skill to have when it comes to geospatial analysis. In this exercise we will be downloading data from several different sources, reprojecting the data to reduce distortion, cropping the data the Trempealeau County, WI, and importing it into a geodatabase using the Python scripting program.

 

Data Collection

     Data was collected from a variety of sources. We downloaded railroad data from the United States Department of Transportation using their National Transportation Atlas Database. Next, the National Elevation Data and Landcover data was downloaded using the National Map Viewer. Then, Cropland data was downloaded from the United States Department of Agriculture National Resources Conservation Service geospatial data gateway. Next, we downloaded the Trempealeau County geodatabase from the Trempealeau County Land Records Department. Finally, we used the USDA NRCS to download the Web Soil Survey for Trempealeau County, WI.

 

Geoprocessing

     Now that all the data is downloaded and unzipped into the appropriate folder we use Python to project each of the data sets into a projection that will reduce distortion for Trempealeau County, extract the data that is only found within the boundary of Trempealeau County, and import the data into the Trempealeau County geodatabase. Below are maps showing a reference map of the location of Trempealeau County within Wisconsin (Figure 1), the landcover of Trempealeau County (Figure 2), the digital elevation of Trempealeau County (Figure 3), the cropland of Trempealeau County (Figure 4), and the soil drainage index of Trempealeau County (Figure 5).

 

Figure 1. Reference map showing the location of Trempealeau County in Wisconsin.

Figure 2. Trempealeau County crop land

 

 

Figure 3. Trempealeau County Digital Elevation

Figure 4. Trempealeau County Land Cover.



Figure 5. Trempealeau County soil drainage index.

 

 

 

Data Quality

    

 

 

     As you can see from the table above, the metadata for the data we downloaded earlier is very incomplete. The majority of data does not include any minimum mapping units, planometric accuracy, or attribute accuracy. This makes it difficult to know for sure whether data is good enough to be used for a project. Therefore, there is always going to be some sources of uncertainty and error in any data that is downloaded from the internet, even when it comes from reliable sources like the USGS or USDA. 

 

 

 Sources

 

Trempealeau County Land Records

http://www.tremplocounty.com/landrecords/

US Department of Transportation

http://www.rita.dot.gov/bts/sites/rita.dot.gov.bts/files/publications/national_transportation_atlas_database/index.html 

USDA Geospatial Data Gateway

http://datagateway.nrcs.usda.gov/GDGOrder.aspx 

USDA Web Soil Survey

http://websoilsurvey.sc.egov.usda.gov/App/WebSoilSurvey.aspx

USGS National Map Viewer

http://viewer.nationalmap.gov/viewer/ 

 

 

 

 

Python Script

 Python is a coding program that allows us to do geoprocessing outside of the ArcMap program. Below is the Python script that was used to reproject land cover, elevation, and cropland rasters to a projection that minimizes distortion for Trempealeau County. Next, the script extracts the 3 rasters by the Trempealeau County boundary. Finally, the script imports the 3 rasters into the Trempealeau County geodatabase.

 

 

The goal of the script below is to select mines that will be using roads to transport sand. To do this we select all of the entries that are active. Out of all the active entries we select all of those that contain the word "mine" in the facility type field. Then we remove selections that contain the word "Rail" in the facility type field. Finally, we remove selections that are within 1.5km of any railroad assuming that these mines will have rail spurs built to the railroad, and therefore will not use roads. This final product is all the active mines that will transport sand on roads.

 

 

 

 

 

     The final python script is designed to create a risk model with weighted variables. In this scenerio I decided that distance from populated areas was the most important risk factor. Therefore, I gave the risk associated with populated areas a a weight of 1.5. This weight gives distance from populated areas a large influence on the overall suitability model. After the weighted raster was created I added up all the risk variables to create a risk model (Figure 1).

 

 

 

 

 

 

 

Figure 1 Weighted risk model based created using Python.

 

 

Overview of Sand Mining in Western WI

 

 

Overview of Sand Mining

 

Introduction

          Frac sand mining is the process of excavating large quantities of sand which will be used in the extraction of natural gas and oil from rock formations. As the technology associated with hydrofracking has increased the need for frac sands has also increased significantly. Therefore, we will look at the requirements sand needs to fall in to be considered frac sands, where frac sand mining is occurring in Wisconsin, the processes associated with a typical frac sand mining operation, common issues that are associated with frac sand mining, and how GIS can be used to explore some of these issues.

 

 

Sand Mine Locations

          Sand must meet certain requirements for use in hydrofracking operations. First off, the sand must be pure silicon dioxide (SiO2), very well rounded, have a compressive strength between 6,000 psi and 14,000 psi, and must fall within a certain size range (Wisconsin Geological and Natural History Survey, 2012). Sand with these qualities is commonly found in Cambrian and Ordovician sandstones that were poorly cemented, such as the Mt Simon Formation and the St. Peter Formation located in Western Wisconsin (Wisconsin Department of Natural Resources, 2012). Estimates from the Wisconsin DNR show that Wisconsin has approximately 63 active mining operations, 45 processing plants (Wisconsin Department of Natural Resources, 2014), and an average of 12 million tons of sand are being excavated yearly (Wisconsin Department of Natural Resources, 2012).  Figure 1 shows the extend of sandstone formations in Wisconsin and the location of existing frac sand mines and processing plants as of December 2011.

 



Figure 1. Sandstone formations in Wisconsin and the location of frac sand mines as of December 2011 (Wisconsin Geological and Natural History Survey, 2012).



         

 

Sand Mine Operations

          Next, we will discuss the process that occurs at a typical sand mine location. First, any unnecessary topsoil is removed from the sand formations and piled along the outskirts of the operation. Next, the actual sand itself is excavated from the site, followed by blasting any tightly cemented sandstone. After any tightly cemented sandstone is blasted into large chunks it needs to be crushed into smaller particles that fit the size requirement. The sand is then sent through a processing plant which washes any fine grained particles off the sand, sorts the grains based on size and shape, and dries the sand. The sand is then transported via large trucks to rail spurs where it will be shipped to its final destination or other facilities for further processing. The final step in the mining process is reclamation of the land (Wisconsin Department of Natural Resources, 2012).





Impacts of Sand Mining

          As with any mining operation there are going to be environmental impacts. Internal combustion engines in heavy machinery and fine grained sand particulates will cause an increase in air pollution, although not significant (Syverson, 2012). A large amount of water is needed to wash the material, which will lower the groundwater table and could have potential impacts on wetlands or other surface water. Water being used to wash the sand will also dissolve certain minerals that will penetrate into the groundwater supply causing contamination. It is also important to look at the socio-economic impacts of sand mining. Heavy machinery will degrade roads and property value surrounding sand mines will decrease dramatically. There are also a lot of economic benefits associated with sand mining. Sand mining creates a lot jobs for electricians, engineers, truck drivers,

accountants, and welders. These jobs will have a large economic benefit associated with them (Syverson, 2012).

Benefits of Using GIS

          GIS is a very important tool in frac sand mining. We are able to analyze areas that would be good locations for sand mines based on several factors including slope, water table height, total amount of sand available, amount of topsoil that will need to be removed, etc. We can also analyze transportation routes that will provide the most effective way to transport sand to areas that need it.

Sources

Syverson, K. (2012, January 15). Kent M. Syverson: Benefits of sand mining outweigh any negatives. LaCrosse Tribune. Retrieved October 1, 2014, from http://lacrossetribune.com/news/opinion/kent-m-syverson-benefits-of-sand-mining-outweigh-any-negatives/article_a7607fc6-3e36-11e1-9401-001871e3ce6c.html

Wisconsin Department of Natural Resources. (2012, January). Silica sand mining in Wisconsin.         Retrieved from http://dnr.wi.gov/topic/Mines/documents/SilicaSandMiningFinal.pdf

 

Wisconsin Department of Natural Resources. (2014, July). Locations of industrial sand mines and processing plants in Wisconsin. Retrieved from http://dnr.wi.gov/topic/Mines/ISMMap.html

 

Wisconsin Geological and Natural History Survey. (2012). Frac sand in Wisconsin. Retrieved from http://wcwrpc.org/frac-sand-factsheet.pdf