Spatial Analyst is an extension in ArcGIS specially designed for working with raster data. 1
Do you remember the difference between vector and raster data in GIS? 2
In Lesson 2 you learned about the difference between vector and raster data. Vector data representations are points, lines and polygons, while raster data is composed of pixels (cells) of a certain size. So far in this class we have worked with the vector data tools, however today s lesson and exercise will cover raster data tools. It is important to be aware of the fact different analysis tools in ArcGIS are used for vector and raster data. This slide shows three raster data layers. 1) an ArcInfo GRID displaying the landcover types of Latah county, Idaho. The brown pixels represent agricultural lands while the green and yellow pixels represent forested lands. 2) a black&white aerial photograph if you zoom in far enough on any photograph you will see that it is made up of pixels. 3) a digital elevation model (DEM) for Latah county, Idaho where each pixel value represents the elevation at that particular location in this DEM the light colors represent high elevation and the dark areas represent low elevation. 3
Satellite imagery is raster data. These are clips from Landsat 7 imagery at 30 meter pixel size. 4
As you zoom in the image looks more and more fuzzy. You are beginning to see the individual building blocks of the image the pixels, or picture elements. 5
The pixels are clearly visible here. Pixels are sometimes called grid cells, cells, etc. Pixel stands for picture element. In a GIS you can measure the size of the pixels using the measuring tool. 6
Each pixel in a raster dataset is represented by a number, often ranging from 0 to 255. In B&W images darker pixels are represented by lower values. 7
Color images are composed of three bands, often Red, Green, Blue. Multi-spectral images have many more bands recorded along the electromagnetic spectrum, e.g. infrared and far infrared bands in addition to red, green, blue Hyperspectral images may have over 100 bands. 8
In the raster data model, each location is represented as a cell. The matrix of cells, organized into rows and columns, is called a grid. Each row contains a group of cells with values representing a geographic phenomenon. Cell values are numbers, which represent nominal data such as land use classes, measures of light intensity or relative 9
All raster datasets (or grids) are rectangular. If they don t look rectangular it is because the NODATA is displayed as transparent. 10
This is the DEM with the study area boundary overlayed in ArcGIS 11
Make a MASK from the study area outline. This is done in Spatial Analyst we will learn how to do this next time. 12
This is the masked out elevation within the study area. 13
This slide shows a raster dataset (GRID) of the vegetation types on Craig Mountain along with the associated tabular data. Raster (grid) tables are different than vector tables. Grid tables always contain at least two fields VALUE and COUNT. The VALUE represent the different grid features. In this example grid VALUE 2001 represents the Dry-land crop types on Craig Mountain. The COUNT column tells you that there are 845 pixels (cells) of this VALUE-type. The area of Dry-land crop types (value = 2001) can be calculated l if the pixel size is known. The pixel size for the Craig Mountain vegetation grid is 30 m and the area of Dry-land crop types is therefore 30x30x845 square meters = 760500 square meters (= 76.05 hectares). Notice how the raster table is similar in structure to a vector SUMMARY table. 14
In raster data, a feature is a group of pixels that have the same value. The area of a feature is the pixel size times the number of pixels of that feature type. 15
How much area is YELLOW? What would the attribute table look like? 16
You are looking at a digital elevation model where each pixel value represents the elevation at a particular location. In this image red is low elevation and dark blue is high elevation. The histogram for the raster dataset shows the pixel count along the range of pixel values, in this case the elevation values. Approximately what elevation value has the highest number of pixels? What is the mean elevation value? What is the min and max elevation values? 17
Raster data can be continuous or classified. In this example the same dataset (elevation in the Owyhee mountains) is shown as a continuous dataset ranging from 651 m to 2559 m and as a classified dataset with five classes. The classified data may simply be the continuous data displayed as classified or it can be reclassified to only have five classes (1,2,3,4,5). 18
Continuous data does not have an attribute table but you can display the histogram The classified data has can be displayed in classes (using Properties Symbology Classified) or it can be Reclassified into 5 classes. If the data is reclassified the raster will have a table with 5 values and the pixel count for each value. The range for each class is not preserved in the reclassification you have to remember what class breaks you selected! 19
In example 2 continuous data is reclassified from values 1-30 into 3 new values (1,2,3). All cells that were in the value range 1-10 in the input grid will be 1 in the output grid. All cells that were in the value range 10-20 in the input grid will be 2 in the output grid. All cells that were in the value range 20-30 in the input grid will be 3 in the output grid (there are no data in this class). 20
When performing analysis, make sure you are asking appropriate questions of the cell size. For example, it is unlikely you will study mouse movement when the cell size is five kilometers. Five kilometer cells may be more applicable when studying the effects of global warming over the earth. 21
There are three resampling methods in ArcGIS: Nearest neighbor Bilinear interpolation Cubic convolution Nearest is used for categorical data (such as vegetation or soil types) while bilinear or cubic is used for continuous data such as elevation 22
Nearest neighbor assignment is the resampling technique of choice for discrete (categorical) data since it does not alter the value of the input cells. Once the location of the cell's center on the output raster dataset is located on the input raster, nearest neighbor assignment will determine the location of the closest cell center on the input raster and assign the value of that cell to the cell on the output raster. The nearest neighbor assignment does not change any of the values of cells from the input raster dataset. The value 2 in the input raster will always be the value 2 in the output raster it will never be 2.2 or 2.3. Since the output cell values remain the same, nearest neighbor assignment should be used for nominal or ordinal data, where each value represents a class, member, or classification (categorical data, such as a landuse, soil, or forest type). Bilinear interpolation uses the value of the four nearest input cell centers to determine the value on the output raster. The new value for the output cell is a weighted average of these four values, adjusted to account for their distance from the center of the output cell. This interpolation method results in a smoother looking surface than can be obtained using nearest neighbor. Since the values for the output cells are calculated according to the relative position and the value of the input cells, bilinear interpolation is preferred for data where the location from a known point or phenomenon determines the value assigned to the cell (that is, continuous surfaces). Elevation, slope, intensity of noise from an airport, and salinity of the groundwater near an estuary are all phenomena represented as continuous surfaces and are most appropriately resampled using bilinear interpolation. 23
If you need to project raster data it is important to select the correct resampling method. Select BILINEAR or CUBIC for continuous data and NEAREST for categorical (classified) data. 24
The Raster calculator in Spatial Analyst can be used to perform grid algebra, i.e. add, subtract, multiply or divide grids on a cell by cell basis. This example shows how the two input grids are added and then divided by 2 to create an output grid. 25
The mathematical functions in Spatial Analyst are not limited to arithmetic operators but also include trigonometric functions, power functions, logarithmic functions and exponentials. This concludes Lesson 7. In Exercise 7 you will practice using some of the tools we have been talking about in this presentation and in Lesson 8 you will learn more about how to use Spatial Analyst for raster modeling using the raster calculator in applications to natural resource problems. 26
Remember that the tools and techniques that you have learned to use for vector data, for example CLIP and INTERSECT, do not work the same way for raster data. You have to use different tools and techniques for vector and raster data! 27