In this first lab we will explore the graphical user interface (GUI) and the command structure of Fragstats, and also explore what impact pixel size (resolution or grain) has on the results of our analyses (an example of how changing the grain changes our view of the landscape is presented below--Figure 2.2 from the text Landscape Ecology). We will do this by considering how the landscape around Saskatoon, Saskatchewan changed from 1966 through to 1976, using Fragstats to calculate some landscape metrics. We will then use Excel to plot the results. (ArcMap etiquette)

For this lab we will be using data from the Canada Land Use Monitoring Program (CLUMP), data which is available from the Geogratis web site. We will be working with the data from Saskatoon, Saskatchewan , collected from 1966 through to 1986 on a 5-year basis (although for the purposes of this lab we'll only be using the 1966 and 1976 data).

You will need to download the data from the Geogratis web site:

1. Go to the Geogratis web site.
2. In the menu boxes on the left side of the page, look for , and then, beneath that text, click on 'Search by keywords'
3. In the search window, enter CLUMP.
4. Select Canada Land Use Monitoring Program (CLUMP) Urban Land Use 1966 to 1986.
5. Select your dataset -- Saskatoon, Saskatchewan (you will be assigned a year in class).
6. Click on the Download button below the dataset you selected.
7. Click on the green arrow and Save the file.
8. UNZIP the two files to C:\Data.

Open the Readme file in order to identify the datum and coordinate system associated with the data, and scroll down until you find the legend information. It would be helpful if you cut / pasted the legend text (Valid CLUMP land Use Codes) into a new Notepad txt file (e.g., save it as legend.txt).

Start ArcMap. Within ArcToolbox, import the interchange file (SASSA##.e00) (Conversion Tools / To Coverage / Import from E00). In the Import window, save the file in C:\data as Sask## [where ## is the year of your CLUMP file] . It may appear that the import process fails, but if you click on the Add file button [+] you should be able to find the coverage file in C:\data.

Although we won't be exploring this aspect of the data in this class, note that the land use classification used in the first set of maps (from 1966 through to 1976) was changed for the later set of maps (from 1981 to 1986). This change in the attribute resolution would affect any temporal analysis we could make of the data (an example of how changes in the classification scheme changes our view of a landscape is presented below--Figure 5.5 from the text Landscape Ecology).

Some of the important concepts you should become familiar with before using the program--some landscape ecology terms--can be found here.

Fragstats--the version we are working with--doesn't work with coverages or shapefiles, it only works with raster files. Therefore, we need to convert the vector coverage into a raster grid. However, before converting any vector file into a raster, you should always view the vector file to ensure that it is correct. Use ArcMap to create a map showing the land use / land cover for Saskatoon. Then, convert the vector file into a raster grid using ArcToolbox / Conversion Tools / To Raster / Polygon to Raster tool. The Input features will be the Sask## polygon file, the Value field should be set to Use. Set the Cellsize to either 100 or 250 m [as assigned in class]. Save the file to C:\data as Sas## (not to a geodatabase). (You may need to turn the Spatial Analyst Extension on: Customize / Extensions.)

BEFORE converting the polygon file to a raster file it will help to set the GeoProcessing / Environments... / Processing Extent in order to exclude the large unmapped areas to the east of the main urban area. Use the values given below:

Once you have created a raster file start Fragstats. In class we will go over the process of running Fragstats and of setting the run parameters. Here is a link to the new FragStats version 4 help file.

For the purposes of this lab we will only be be examining a few of the many metrics that are available: Class metrics (Area: Total Area, Percentage of Landscape, Edge: Total Edge, and Core Area and Core Area Percent of Landscape) and Landscape metrics (Aggregation: Number of Patches, Patch Density; Diversity: Shannon's Diversity Index and Shannon's Evenness Index).

The basic steps are:

1. Start Fragstats and click on New:
   
   
   
   
2. Click on Add layer and select ESRI grid:
   
   


Note: Currently that option (selecting an ESRI grid file) doesn't work--we are working on fixing the problem. So, for now, to get around the problem you need to export your raster images to a GeoTIFF format. To do this, go to ArcTooobox / Conversion Tools / To Raster and select Raster to Other Format (Multiple). Select your input raster, select C:\Data as the Output Workspace, and keep TIFF as the Raster Format (Optional). Click on OK to export the file. Now, in FragStats you would select GDAL GeoTIFF grid as the Data input dataset type.



3. Select the GeoTIFF file (Dataset name) you just created in C:\data
   
   
4. In order to make it easier to interpret the results you should create a Features Class Descriptors (fcd) file. You will need to create an fcd file for each layer since there is no guarantee that the ESRI grid values (Value) are consistently associated with a specific land use (Use). To create an fcd file, you first need to examine the Attribute file of your raster data, and note down the Value associated with each Use, and then add the Values to the legend.txt file you saved earlier (note that you have to remove all of the commas from the legend entries and only have them present in the fcd file where required). You can then Save AS the file as an fcd file.
   
   
5. You will need to set some parameters in Common Tables. The 'Edge depth' parameter is used to identify how far into an area the edge effects can be felt. For example--the edges of forests often contain species that are more representative of the adjacent areas than they are of the forest interior and, thus, in determining how much 'true' forest lands remain, we should exclude those forested areas that are within 100 m of an edge.
   
   


6. You will need to set some Analysis parameters:
   
   
   
   
7. Select Class Metrics and select the metrics mentioned above (note the acronyms associated with each metric).
   
   
8. Select Landscape Metrics and select the metrics mentioned above.
   
   
9. Select Run
   
   
10. View the Results (Class and Land) in order to ensure that the program ran okay.

Using Excel we can take the output files from Fragstats (the 'Automatically save results' files) and produce a number of plots (the output files [e.g., *.class] are comma delimited files that Excel can easily import). We will then compare the results of our analyses and see how the landscape around Saskatoon has changed over time, and how changes in spatial resolution can affect the results. Your graphs should have the bars labelled appropriately (when making the chart, select Series; Category (X) axis---select the Land Cover values).