Introduction to GIS Modeling Week 3 Reclassifying and

Introduction to GIS Modeling Week 3 — Reclassifying and Overlaying Maps GEOG 3110 –University of Denver Presented by Joseph K. Berry W. M. Keck Scholar, Department of Geography, University of Denver Suitability Modeling Logic Reclassifying Maps (position, initial value, size, shape and contiguity) Overlaying Maps (point-by-point, region-wide and map-wide)

Evaluating Habitat Suitability <Homework Question #4> Assumptions – Hugags like… gentle slopes, southerly aspects, and lower elevations Generating maps of animal habitat… (digital slide show Hugag) Manual Map Overlay (Binary) (See Beyond Mapping III online book, “Topic 23” for more information) Ranking Overlay (Binary Sum) Rating Overlay (Rating Average) (Berry)

Conveying Suitability Model Logic gentle slopes Elevation Slope Rows Model Criteria Columns Analysis Slope Reclassify Preference Bad 1 to 9 Good Levels (Times 1) southerly aspects Elevation Aspect Reclassify Preference Habitat Rating (1) Overlay Bad 1 to 9 Good lower elevations Elevation Reclassify Preference Elevation Bad 1 to 9 Good Algorithm Base Maps Calibrate Derived Maps Interpreted Maps Fact Covertype Bad 1 to 9 Good (1) Lines Processing Steps (Commands) 0= No, 1 to 9 Good Weight Solution Map Judgment Reclassify Water Mask 0= No, 1= Yes (See Beyond Mapping III online book, “Topic 22” for more information) Overlay Habitat Rating Constraint Map …while Reclassify and Overlay commands are not very exciting, they are some of the most frequently used operations (Berry)

Extending Model Criteria gentle slopes Elevation Slope Preference Bad 1 to 9 Good Rows Model Criteria Columns Analysis Levels (Times 10) southerly aspects Elevation Aspect Preference (1) Bad 1 to 9 Good lower elevations Elevation Preference Elevation (1) Bad 1 to 9 Good forests Forest Proximity Forest Preference (10) Bad 1 to 9 Good water Water Proximity Water Preference Bad 1 to 9 Good Habitat Rating (10) Lines Processing Steps (Commands) Additional criteria can be added… —Hugags would prefer to be in/near forested areas —Hugags would prefer to be near water —Hugags are 10 times more concerned with slope, forest and water criteria than aspect and elevation (Berry)

Grid-Based Map Analysis Spatial analysis investigates the “contextual” relationships in mapped data… ü Reclassify— reassigning map values (position; value; size, shape; contiguity) ü Overlay— map overlay (point-by-point; region-wide; map-wide) ü Distance— proximity and connectivity (movement; optimal paths; visibility) ü Neighbors— ”roving windows” (slope/aspect; diversity; anomaly) Weeks 3, 4, 5, 6 Spatial statistics Surface modeling maps the spatial distribution and pattern of point data… ü Map Generalization— characterizes spatial trends (e. g. , titled plane) ü Spatial Interpolation— deriving spatial distributions (e. g. , IDW, Krig) ü Other— roving window/facets (e. g. , density surface; tessellation) Data mining investigates the “numerical” relationships in mapped data… ü Descriptive— aggregate statistics (e. g. , average/stdev, similarity, clustering) ü Predictive— relationships among maps (e. g. , regression) ü Prescriptive— appropriate actions (e. g. , optimization) Weeks 7, 8, 9 (Berry)

Classes of Spatial Analysis Operators …all spatial analysis involves changing values (numbers) on a map as a mathematical or statistical function of the values on that map (See Example Applications, “Cross-Reference” for a cross reference of Map. Calc operations and those of other systems) (Berry)

An Analytic Framework for GIS Modeling …recall Map Analysis organization and evolution discussion from Week 1 class presentation/reading (GIS Modeling Framework paper) Reclassify operations involve reassigning map values to reflect new information about existing map features (Berry)

Reclassifying Maps (Berry)

Reclassifying Maps <Homework Question #2> Contiguity Shape Initial Value Size Initial Value CLUMP -- Assigns new values to contiguous groups of cells within each map category. CONFIGURE -- Assigns new values characterizing the shape of the area associated with each category. The most frequently used map analysis operation …and the most dangerous!!! RENUMBER -- Assigns new values to the categories of a map. SIZE -- Assigns new values according to the size of the area associated with each map category. SLICE -- Assigns new values by dividing the range of values on a map into specified intervals (contouring). (Berry)

Renumber Operation Renumber— assigns new values to the categories of a map. …any real number from -3. 4 E 38 to + 3. 4 E 38 can be assigned to any existing value on a map …often integer values are assigned based on user reasoning (as in this example) Note: PMAP_NULL is a special value that can be assigned indicating “no data” and the grid location will be ignored in processing and display RENUMBER Elevation ASSIGNING 1 TO 500 THRU 1800 ASSIGNING 0 TO 1800 THRU 2500 FOR E_Pref …context Help provides information on use of operations <more info in the Map. Calc Manual> (Berry)

Slice Operation Slice— assigns new values by dividing the range of values on a map into specified intervals (“Equal Ranges” contouring) …Map Range = Max_Value – Min_Value = 2500 – 500= 2000 feet …Contour Interval = Map Range / # ranges = 2000 / 20 = 100 feet SLICE Elevation into 20 FOR Relief_100 ft …a user can specify the minimum and maximum values of the range– SLICE Elevation INTO 20 FROM 1000 THRU 1200 Zero. Fill FOR Relief_10 ft …Slice <map. Name> is often used to collapse a map with a large set of map values to just a few intervals for a quick view of the pattern of the spatial data distribution (Berry)

Size Operation Size— assigns new values according to the size of the area associated with each map category. …the Size operation assigns the number of cells comprising each map region (category) …in this instance there are three map regions (Open Water=1, Meadow=2, Forest=3)– note that Water occurs in two different places. SIZE Covertype FOR Coverype_size …to calculate the actual area of each map region multiply the size map times the area per grid cell– 10, 000 m 2 or 1 ha in this case …to calculate the size/area of each occurrence you must first Clump the map “regions”, then use the Size command (Berry)

Clump Operation Clump— assigns new values to contiguous groups of cells within each map category. …a map “category” identifies all locations with the same characteristic or condition– e. g. , Open Water, Meadow, Forest …a “clump” is a contiguous group (individual spatial instance of a map “category”)–e. g. , five cover type clumps with two instances of Open Water CLUMP Covertype AT 1 DIAGONALLY FOR Coverype_clumps “At” identifies how far to reach in defining clumps “Orthogonally” reaches horizontally and vertically only; “Diagonally” includes off angles (Berry)

Configure Operation Configure— assigns new values characterizing the shape and integrity of the area associated with each map category. Boundary Configuration Convexity is the ratio of the Edge to the Area (Size) and normalized to that of a circle of the same area CONFIGURE Covertype Edges FOR Covertype_edges Edge cells Spatial Integrity Euler = (# Holes) – (1 -#Fragments) (Berry)

Some Reclassifying Things to Keep in Mind The Covertype map contains Nominal data that is Discretely distributed in geographic space. As such, it is best displayed in 2 D using cells (Grid) and with layer mesh on as the stored values do not form gradients in either numerical or geographic space. The Size command assigns new values according to the size of the area (# of cells) associated with each map category. In this instance the input map is Covertype (Nominal/Discrete data) and the output map is Covertype_size (Ratio/Discrete data). The size algorithm “counts” the number of cells for each map category (stored map value). Size works on Nominal data (Categorical) but usually is not appropriate for ratio data as far to many values (decimal places); results in most of the map being assigned the cell size value of 1 because elevation values with decimal points rarely are the same. For example, sizing the Elevation map… …identifies that “one cell in size” elevation values occur in 64% of the map area (384/1= 384 times; each value is unique). “Three cells in size” areas occur in 2. 88% of the analysis window (18/3= 6 times; six sets of the same value). (Berry)

An Analytic Framework for GIS Modeling …Map Analysis organization and evolution (GIS Modeling Framework paper) Overlay operations involve characterizing the spatial coincidence of mapped data (Berry)

Map Accuracy (Error Propagation– simple overlay) Topological Overlay Intersect Polygons Spatial Table Attribute Table Soil Type Forest Type Join Data Coincidence Search 100% certain everywhere in the derived polygon? Believable? Vector Raster Joint Probability — likelihood of two conditions occurring together… (digital slide show Honest) P(x, y) = P(x) * P(y) Joint Probability …evaluated cell-by-cell (See Beyond Mapping II, “Topic 4, “Toward and Honest GIS”) (Berry)

Overlaying Maps (Berry)

Overlaying Maps <Homework Question #3> Region wide COMPOSITE -- Creates a map summarizing values from one map which coincide with the categories of another. Point-by. Point CALCULATE and COMPUTE -- Creates a map as the mathematical or Point-by. Point COVER -- Creates a new map where nonzero values of the top map Point-by. Point CROSSTAB -- Generates a spatial coincidence table of two maps. Point-by. Point statistical function of two or more maps. …true “map-ematics” replace the values on the previous (bottom) map, or stack of maps. INTERSECT -- Creates a map that assigns new values to pair-wise combinations of values on two maps. …Map-wide overlay involves spatial statistics (spatial data mining) (Berry)

Compute/Calculate Operation Compute/Calculate— creates a new map as the mathematical function of two or more maps. All of the basic mathematical operations on a typical pocket calculator can be performed on grid maps… …including Add, Subtract, Multiply, Divide, Exponentiation, Square Root, Max, Min, And, Or, & Trig functions …other math functions? COMPUTE Covertype TIMES 10 Plus Water FOR Coverype_Water …in this example one map is multiplied by 10 then added to another map, thereby creating a 2 -digit code indicating the first map’s categories as the tens digit followed by the second map’s categories as the one’s digit (Berry)

Crosstab Operation Crosstab— generates a spatial coincidence table of two maps. Map 1 = Districts …the maps are compared “cell-by-cell” and the number of joint occurrences between the map categories are summarized in a table Optional Question 3 -2 Map 2 = Covertype CROSSTAB Districts WITH Covertype Simply TO Newtextfile. txt In this example there are… • 58 cells classified as District 1 on the Districts map • 82 cells classified as Open Water on the Covertype map • 58 cells identified as having the joint condition of District 1 and Open Water representing 9. 28 percent of the entire map area. Note: all of the District 1 cells are in Open Water (Berry)

Intersect Operation Intersect— creates a map that assigns new values to pairwise combinations of values on two maps. Map 1 = Districts The maps are compared “cell-by-cell” and a user specified number is assigned to designated category combinations… …“zerofill” assigns 0 to all combinations that are not specified Map 2 = Covertype … “oldfill” retains 1 st map values for non-specified combinations INTERSECT Districts WITH Covertype ASSIGN 1 to 1, 1 Zerofill FOR Districts 1_Cover 1 …if “completely” is specified all combinations are automatically identified using unique sequential numbering for map values Note: Intersect is similar to geo-query operations in desktop mapping packages by identifying all locations having specified category (map value) combinations (Berry)

Cover Operation Cover— creates a new map where the non-zero values of the top map replace the values on the previous (bottom) map, or stack of maps. Water 4 4 0 1 Cover 1 3 Covertype …the maps are compared “cell-by-cell” and the value in the top-most cell replaces the previous values unless that value is zero, then the top most non-zero value is retained (0, 3 3) (4, 1 4) “Zero” is treated as transparent as maps are staked; non-zero values treated as opaque COVER Covertype WITH Water IGNORE 0 FOR Districts 1_Cover 1 …in the example coincidence 4, 1 4 because 4 is non-zero and replaces what is beneath it … 0, 3 3 because zero is ignored and does not replace the previous value in the map stack (Berry)

Composite Operation Composite— creates a map summarizing values from one map that coincide with the categories (termed regions) identified on another map Template Map “Cookie-cutter” …the regions identified by category values on one map serve as cookie-cutter shapes (Template map) for summarizing data contained on another map (Data map) Data Map COMPOSITE Districts WITH Slope Average IGNORE PMAP_NULL FOR Districts_avg. Slope …data summary procedures include Average, Standard deviation, Coefficient of variation, Total, Maximum, Minimum, Median Majority, Minority, Diversity, Deviation and Proportion (Berry)

Generalizing Continuous Mapped Data Districts Map (2 D Discrete) 3 D Plot of Elevation (3 D Continuous) (Template Map) …region-wide overlay Districts Draped over Elevation (3 D Continuous) (Data Map) Elevation Contours (2 D Discrete) (Berry)

Thematic Mapping (Average elevation by district) Worst “Thematic Mapping” Average Elevation of Districts 500 1539 (0) 653 (29) (39) 2176 (9) 1099 (21) 1779 1080 (24) (9) Best …include Coffvar in Thematic Mapping results (Berry)

Creating a Flowchart (using Power. Point) Enter map title Enter yet another map title 2 Slope Enter map title 1 Under the Home tab, 1) use the Rectangle drawing tool to create and size a box to represent a map in the flowchart and 2) use the Text box drawing tool to enter the map name in the Rectangle. Select both and use Format Group to group the two objects. Enter other map title Copy and Paste the Rectangle to form other maps. Use the Line drawing tool to connect the boxes. Use the Text box drawing tool to enter the command, rotate and place over the Line. Repeat the process to additional maps (boxes) and processing steps (lines) to complete the flowchart containing the logic of the GIS model. There a lot of other things you can do to make the graphic a bit more unique, such as borders, transparency, shadowing and animation (viewed as a slide show). Also, keep in mind that Format Align can be used to align multiple graphic objects if things get out of whack. (Berry)

Compound Graphic (Campground model results) Campground Suitability Use Snag. It to capture one of the graphic elements, such as the S_Pref map then Paste and Size at the appropriate location on the “canvas” (white background shape). Repeat for all of the other graphic elements. Use the Text Box tool to embed text as appropriate. Group the figure elements in logical groupings and then use the Custom Animation tool to control their sequencing for display if you intend to present as a Power. Point slide deck. (Berry)

Optional Questions Flowchart of Binary Habitat Model “Simply” and “Completely” Crosstab Tables Average Suitability Rating for each Covertype parcel (Berry)

Optional Questions (continued) Average Fuel Loading index for each District Average and Coefficient of Variation maps for the 200 foot contour polygons of Elevation (Berry)