-------------------------------------------------------------------------------- -------------------------------------------------------------------------------- 6. In ArcMap, open Redlands.mxd in your PSEArcGIS\basicsgis\Lesson01 folder. Use the Select By Attributes dialog to select all streets whose CLASS = 3. How many records are selected? 203 is correct! Refer to Use ArcGIS to explore geographic data, Step 8 7. In ArcMap, open Redlands.mxd in your PSEArcGIS\basicsgis\Lesson01 folder. What is the map tip for the donut shop located furthest to the east on the map? Winchell's Donut House is incorrect. Refer to Use ArcGIS to explore geographic data, Step 6 9. In ArcMap, open Redlands.mxd in your PSEArcGIS\basicsgis\Lesson01 folder). What is the class description of New York Street? Major street is incorrect. Refer to Use ArcGIS to explore geographic data, Step 6 7. Your PSEArcGIS\basicsgis\Lesson02 folder contains a geodatabase named National. Using ArcCatalog, determine which of the following coordinates describes the western bounding coordinate for the Counties feature class. -180.000000 is incorrect. Refer to Explore ArcMap and ArcCatalog Hint: Use the Metadata tab. 10. Your PSEArcGIS\basicsgis\Lesson02 folder contains a table in the National geodatabase called state_names. Using ArcCatalog, explore state_names. What is the data type of the STATE_FIPS field? String is incorrect. Refer to Explore ArcMap and ArcCatalog Hint: Use the Metadata tab. 9. In ArcMap, open Redlands.mxd in your PSEArcGIS\basicsgis\Lesson01 folder. What is the land use abbreviation for the parcel on which ESRI is located? TNS is incorrect. Refer to Use ArcGIS to explore geographic data, Step 4 10. Your PSEArcGIS\basicsgis\Lesson02 folder contains a table in the National geodatabase called state_names. Using ArcCatalog, explore state_names. What is the data type of the STATE_FIPS field? Double is incorrect. Refer to Explore ArcMap and ArcCatalog Hint: Use the Metadata tab. 11. What kind of analysis is used to integrate soils data with slope, vegetation and tax assessment data? Overlay analysis is correct! Refer to Analyzing data 12. What kind of analysis is used to determine whether an apartment building is within 1 mile of an earthquake fault? Proximity analysis is correct! Refer to Analyzing data 13. In ArcMap, open Redlands.mxd in your PSEArcGIS\basicsgis\Lesson01 folder. What is the land use abbreviation for the parcel located furthest east on the map? VAC is correct! Refer to Use ArcGIS to explore geographic data, Step 4 14. Using ArcCatalog, navigate to your PSEArcGIS\basicsgis\Lesson02 folder. In the Rhode_Island workspace, there is a coverage named zip, with a point feature class. The table for the point feature class contains all the columns listed below, except one. Which one? CODE is correct! Refer to Explore ArcMap and ArcCatalog 15. Based on your work in the project exercise "Use ArcGIS to explore aquaculture in Zambia," how many road segments are in the new, clipped layer? 12 is incorrect. Refer to Use ArcGIS to explore aquaculture in Zambia, Step 12 16. Based on your work in the project exercise "Use ArcGIS to explore aquaculture in Zambia," how many populated place features have no name? 349 is incorrect. Refer to Use ArcGIS to explore aquaculture in Zambia, Step 6 17. Based on your work in the project exercise "Use ArcGIS to explore aquaculture in Zambia," which field name is set for the map tips in the soils layer? SOILS_ID is incorrect. Refer to Use ArcGIS to explore aquaculture in Zambia, Step 4 18. Based on your work in the project exercise "Use ArcGIS to explore aquaculture in Zambia," how many populated places are within areas that have the Alfisols soil type? 18 is incorrect. Refer to Use ArcGIS to explore aquaculture in Zambia, Step 7 20. Based on your work in the project exercise "Use ArcGIS to explore aquaculture in Zambia," how many countries had less than a 2000 (kcal/day) DES in the years 1996 to 1998? 8 is correct! Refer to Use ArcGIS to explore aquaculture in Zambia, Step 3 11. Which ArcMap tool would you use to click on a building and see its square footage? Select is incorrect. Refer to Querying data 6. When you want to show discrete boundaries of features, you should use the raster data model to store geographic data. True is incorrect. Refer to Storing data 13. In ArcMap, open Redlands.mxd in your PSEArcGIS\basicsgis\Lesson01 folder. What is the land use abbreviation for the parcel located furthest east on the map? RES is incorrect. Refer to Use ArcGIS to explore geographic data, Step 4 5. All of the following are options for customizing ArcMap and ArcCatalog, except one. Which one? Adding new macros and commands is incorrect. Refer to Customizing ArcCatalog and ArcMap 5. Under the Metadata tab in ArcCatalog there are three additional tabs. Which of the below is not one of them? Description is incorrect. Refer to ArcCatalog 19. Based on your work in the project exercise "Use ArcGIS to explore aquaculture in Zambia," how many road segments are in the new, clipped layer? 71 is incorrect. Refer to Use ArcGIS to explore aquaculture in Zambia, Step 12 8. Using ArcCatalog, navigate to your PSEArcGIS\basicsgis\Lesson02 folder. In the Rhode_Island workspace, there is a coverage named zip, with a point feature class. The table for the point feature class contains all the columns listed below, except one. Which one? STATE is incorrect. Refer to Explore ArcMap and ArcCatalog 10. Your PSEArcGIS\basicsgis\Lesson02 folder contains a geodatabase named National. Using ArcCatalog, determine which of the following coordinates describes the western bounding coordinate for the Counties feature class. -71.368211 is incorrect. Refer to Explore ArcMap and ArcCatalog Hint: Use the Metadata tab. 1. There are several ways to preview data in ArcCatalog. Which of the following is not an ArcCatalog preview method? Thumbnail is incorrect. Refer to ArcCatalog 3. With ArcView you can do all of the following tasks, except one. Which one? Create and manage annotation is incorrect. Refer to ArcView 9. Your PSEArcGIS\basicsgis\Lesson02 folder contains a geodatabase named National. Using ArcCatalog, determine which of the following coordinates describes the western bounding coordinate for the Counties feature class. -21.073394 is incorrect. Refer to Explore ArcMap and ArcCatalog Hint: Use the Metadata tab. 3. Under the Metadata tab in ArcCatalog there are three additional tabs. Which of the below is not one of them? Spatial is incorrect. Refer to ArcCatalog 10. Using ArcCatalog, navigate to your PSEArcGIS\basicsgis\Lesson02 folder. In the Rhode_Island workspace, there is a coverage named zip, with a point feature class. The table for the point feature class contains all the columns listed below, except one. Which one? PONAME is incorrect. Refer to Explore ArcMap and ArcCatalog 7. What is the most important component of a GIS? Data is incorrect. Refer to Components of a GIS 8. Which of the following best describes the types of functions for which you would use ArcMap? Data management, viewing and editing metadata is incorrect. Refer to ArcMap 2. Under the Metadata tab in ArcCatalog there are three additional tabs. Which of the below is not one of them? Attributes is incorrect. Refer to ArcCatalog

In this lesson, you will learn: what the five components of a GIS are what geographic data consists of how real-world objects are represented in a GIS what attributes are how spatial relationships and topology affect geographic analysis how geographic data is organized in a GIS what six operations a GIS should be able to perform What is a GIS? GIS is used by many industries, including utilities, commercial businesses, law enforcement, transportation, health care, agriculture, and local, state, and federal governments. Industry is using GIS for things like natural resource management, land use planning, demographic research, emergency vehicle dispatch, fleet management, environmental assessment and planning, and much more. The number of GIS applications on the Internet is growing rapidly. OK, GIS is a popular technology, but what exactly is it? What does it do? Basically, a geographic information system (GIS) is a computer-based tool for solving problems. A GIS integrates information in a way that helps us understand and find solutions to problems. Data about real-world objects is stored in a database and dynamically linked to an onscreen map, which displays graphics representing real-world objects. When the data in the database changes, the map updates to reflect the changes. In general, people use a GIS for four main purposes: data creation, data display, analysis, and output. You can display objects according to the data in your database (this is a powerful feature that you'll appreciate later). GIS analysis tools allow you to do things like find out how far your best customers travel to visit your store, which land parcels are within a flood zone, and which soil type is best for growing a particular crop. Output options include cartographic-quality maps as well as reports, lists, and graphs. You'll learn more about how a GIS works in the concepts that follow A GIS has five components. They are: people, data, hardware, software, and procedures. True to its name, a GIS is in fact a system. All five components are required to produce the desired results. People—people are the most important component of a GIS. People must develop the procedures and define the tasks the GIS will perform. People can often overcome shortfalls in other components of the GIS, but the opposite is not true. The best software and computers in the world cannot compensate for incompetence. Data—the availability and accuracy of data affect the results of queries and analysis. Hardware—hardware capabilities affect processing speed, ease of use, and the types of available output. Software—this includes not only GIS software, but also various database, drawing, statistical, imaging, and other software programs. Procedures—GIS analysis requires well-defined, consistent methods to produce correct and reproducible results. There are three main components to geographic data: Geometry represents the geographic features associated with real-world locations. Geographic features are abstracted into (drawn as) points, lines, or polygons (areas). Attributes are descriptive characteristics of the geographic features. Behavior means that geographic features can be made to allow certain types of editing, display, or analysis, depending on circumstances that the user defines. Feature behavior is most easily implemented in the geodatabase. Here, city streets are being represented in a GIS. Their geometry is lines, because they are stored and drawn as line features in the GIS. Each street can be described by its name, which is an attribute. You can also specify rules for the streets. For example, you could add behavior that says the streets cannot have more than four lanes. Feature spatial relationships On a map, feature spatial relationships, or where they are located in space relative to one another, communicate important information. The spatial relationships implicit on a map determine what the map conveys to the reader. For example, connectivity and adjacency are two types of spatial relationships. Interstate 80 connects San Francisco with New York City. San Francisco is adjacent to the Pacific Ocean, while New York City is adjacent to the Atlantic Ocean. Other feature relationships can be determined from this map. For example, if you look at California, you can see that it is contained within the United States, it's intersected by I-80, and it's adjacent to the Pacific Ocean. Topology is a mathematical procedure used to determine feature spatial relationships and properties, including: Connectivity of lines Direction of lines Length of lines Adjacency of areas Area definition Topology makes most types of geographic analysis possible because it allows us to answer questions having to do with feature spatial relationships. For example, for the geographic features in the graphic below, topology can answer questions like: What is adjacent to the ESRI Campus? What street intersects State Street? How large is the ESRI parking lot? What is the quickest path from the Warehouse to the Post Office? Topology tells us that the parcels containing the Post Office and the ESRI Campus are adjacent because they have a common borderline. How a GIS organizes geographic data A GIS organizes and stores information about the world as a collection of thematic layers that can be linked by geography. Each layer contains features having similar attributes, like streets or cities, that are located within the same geographic extent. This simple but extremely powerful and versatile concept has proven invaluable for solving many real-world problems—from tracking delivery vehicles to recording details of planning applications to modeling global atmospheric circulation What can a GIS do? Any geographic information system should be capable of six fundamental operations in order to be useful for finding solutions to real-world problems. A GIS should be able to: Capture data Store data Query data Analyze data Display data Output data Capturing data Data describing geographic features is contained in a geographic database. The geographic database is an expensive and long-lived component of a GIS, thus data entry is an important consideration. How are you going to get data that exists only on paper into the database? What about data that is in a digital, but unusable, format? A GIS must provide methods for entering geographic (coordinate) and tabular (attribute) data. The more input methods available, the more versatile the GIS. You can bring data from many different sources into a GIS. There are two basic models used for geographic data storage: vector and raster. A GIS should be able to store both types of geographic data. How features appear in a GIS depends on the data format in which they're stored. In the vector data model, geographic features are represented similarly to the way they are on maps, with points, lines, and polygons. A graticule or an x,y(Cartesian) coordinate system is used to reference the real-world locations of these features. The raster data model represents features using a grid of cells. Values are assigned to cells that cover the features' locations. The amount of detail you can show for a particular feature depends on the size of the cells in the grid. Raster format is well-suited to spatial analysis and is also appropriate for the storage of data collected in grid format. Raster data is inappropriate for applications like parcel management, where discrete feature boundaries must be known. A GIS must provide tools for finding specific features based on their location or attributes. Queries, which are often created as logical statements or expressions, are used to select features on the map and their records in the database. A common GIS query is determining what exists at a particular location. In this type of query, the user knows where the features of interest are, but wants to know what characteristics are associated with them. This can be accomplished with a GIS because geographic features on the map display are linked to their attributes (descriptive characteristics) stored in the database. In a GIS, you can click on a map feature to see the attributes associated with the feature in the database. Another type of GIS query is to determine which location or locations satisfy certain conditions. In this case, the user knows what characteristics are important and wants to find out where the features are that have these characteristics. Suppose you wanted to find landlocked countries with a population greater than 20 million. You would create a query expression with those criteria. When the GIS finds features that meet the query's criteria, it highlights them on the map. Analyzing data Geographic analysis usually involves more than one geographic dataset and requires working through a series of steps to reach a final result. A GIS must be able to analyze the spatial relationships among multiple datasets to answer questions and solve problems. There are many types of geographic analysis. While this course cannot cover all of them, two common types of geographic analysis are described below. Proximity analysis Proximity analysis uses the distance between features to answer questions like: How many houses lie within 100 meters of this water main? What is the total number of customers within 10 kilometers of this store? What proportion of the alfalfa crop is within 500 meters of the well? GIS technology often uses a process called buffering to determine the proximity relationship between features. A 50-foot buffer is created on either side of the road to find those parcels within the 50-foot distance. Overlay analysis The integration of different data layers involves a process called overlay. At its simplest, this could be a visual operation, but analytical operations require one or more data layers to be joined physically (i.e., combined into one layer in the database). Overlay analysis could be used to integrate data on soils, slope, and vegetation or land ownership data with tax assessment data. Overlay analysis results can help determine appropriate land uses. Analyzing data The integration of different data layers involves a process called overlay. At its simplest, this could be a visual operation, but analytical operations require one or more data layers to be joined physically (i.e., combined into one layer in the database). Overlay analysis could be used to integrate data on soils, slope, and vegetation or land ownership data with tax assessment data. Overlay analysis results can help determine appropriate land uses. Outputting data Sharing the results of your geographic labor is one of the primary justifications for spending resources on a GIS. Taking displays created through a GIS (maps, graphs, and reports) and outputting them into a distributable format is a great way to do this. The more output options a GIS can offer, the greater the potential for reaching the right audience with the right information. GIS displays, like this map, can be output and distributed in a variety of ways. What is a GIS database? A geographic database is the core of a GIS. Its completeness and accuracy affects all the applications it supports. The database is a collection of the spatial and descriptive attributes of real-world objects. For best results, the database should be organized to efficiently serve one or more applications and be maintained by a set of well-documented and well-administered procedures. In a GIS database, real-world objects are abstracted into the digital world. How these digital features are organized and stored depends on several factors, including the intended application of the data, the software you plan to use, and the nature of the features themselves. Attributes describing the geographic features are usually linked with the features in some way. In this topic, these data issues will be presented in more detail. Abstracting real-world entities GIS users must somehow abstract real-world entities into a geometric representation of those entities. Real-world entities can be abstracted into three primary shapes: points, lines, and polygons (areas). These shapes are often called geometric objects, geometric features, feature types, or feature classes. Each feature class has its own unique set of characteristics. A point is composed of one coordinate, or x,y location. A line is a sequence of coordinates and has an intrinsic length value. A polygon is composed of one or more lines whose starting and ending coordinates have the same value. Polygons have two intrinsic values, perimeter and area. How you represent geometric objects affects the way you can display and analyze information. Our interaction with objects in the world is diverse, and you can represent them in many ways. Consider one example: rivers. Rivers are natural features that are used for transportation, delimit political or administrative areas, and are an important feature in the shape of a surface. Below you'll read about a few of the many ways you can think about representing rivers in a GIS. A set of lines that forms a network Each section of line has flow direction, volume, and other attributes of a river. You can apply a linear network model to analyze hydrographic flow or ship traffic. Rivers represented as lines. A border between two areas A river can delimit political areas, such as provinces and counties, or it can be a barrier for natural regions, like wildlife habitats. A river can form a boundary. An areal feature A river can be represented as an areal feature with an accurate representation of its banks, braids, and navigable channels. A river has area. A trough in a surface model A river can be a sinuous line forming a trough in a surface model. From the river's path through a surface, you can calculate its profile and rate of descent, the watershed it drains, and its flooding potential for a prescribed rainfall. A river is a trough in a surface based on elevation values. Linking features and attributes Each individual feature is assigned a unique numerical identifier and is characterized by a unique location in space and corresponding record in a feature attribute table. While the exact name of the numerical identifier may differ according to the data format, it is important to understand this one-to-one relationship between feature, identifier, and attribute record. Each feature has a record in the table. A uinique identifier links a feature with its attributes Storing abstracted objects There are two basic models for storing geographic data: the vector and raster data models. Vector storage requires that point, line, and polygon features be explicitly referenced with positional coordinates such as an x,y Cartesian coordinate system. Each point has a single x,y coordinate pair referencing its position. Lines can be stored either as a series of x,y coordinates defining the points between straight line segments, or as two end-point coordinates and a curve formula. Polygons are composed of lines that close to form the polygon boundaries. The vector data model stores positional coordinates for each shape. Raster data storage uses a grid of equally-sized square cells to model points, lines, and polygons. Instead of explicitly defining the shapes with x,y coordinates, georeferenced raster formats store a single x,y coordinate pair (usually the lower-left corner of the data layer), the number of rows and columns, and the cell size. With these four pieces of information, features within a raster dataset are assigned their position in space. The raster format uses a grid of square cells to represent real-world entities. There are slight differences between vector and raster formats in how features and attributes are linked. The feature and attribute table model described in the previous concept primarily applies to vector data. Introducing metadata The definition typically given for metadata is "data about data." While this definition is not incorrect, it is also not very informative. More precisely, metadata is supporting descriptive information about data. Common examples of "everyday" metadata include notes written on the back of a photograph telling the photograph's date and subject and nutrition labels on food containers. In ArcGIS, there are three main categories of metadata: Description, Spatial, and Attributes. Description information contains basic information about the dataset, including source, organization, date, uses, and restrictions. Spatial information contains the coordinate system information and geographic extent of the dataset. Attributes information includes fields, attribute domains, and related information. Metadata for spatial data includes descriptive information such as date, creator, geographic extent, coordinate system, and attribute domains. Metadata gives spatial data credibility and, in many situations, spatial data may be impossible to interpret or use without metadata. It is important to maintain standards for metadata so that it can be easily distributed and interpreted. As you will see, with ArcGIS software you can create, edit, import, and export metadata for any given dataset. Metadata is stored in a data format called Extensible Markup Language, or XML. This means that you can use the metadata with other software that can read XML documents, including Web browsers. Metadata standards Summary A geographic information system (GIS) integrates information about the real world to help people find solutions to problems. Data about real-world objects is stored in a database and dynamically linked to an onscreen map, which displays the real-world objects. A GIS is used for three main purposes: data display, analysis, and output. There are five components to a GIS. They are: people, data, hardware, software, and procedures. Of these, people are the most important component. A GIS organizes and stores information about the world as a collection of thematic layers that can be linked by geography. Each layer contains features having similar attributes that are located within the same geographic extent. Geographic data is what you work with in a GIS. Geographic phenomena can be represented using raster or vector models. The raster data model represents geographic features using cells, while the vector data model epresents features using points, lines, and polygons. Geographic vector data has three components: geometry, attributes, and behavior. Geometry (points, lines, or polygons) represents the geographic features associated with real-world locations. Attributes are descriptive information about features. Behavior means that geographic features can be made to allow certain types of editing, display, or analysis, depending on circumstances that the user defines. Feature spatial relationships, or where they are located in space relative to one another, communicate important information. Topology is a mathematical procedure used to determine feature spatial relationships and properties. Topology makes most types of geographic analysis possible. Any GIS should be able to perform six fundamental operations to be useful for finding solutions to real-world problems. A GIS should be able to capture, store, query, analyze, display, and output data. Penn State Edition: Basics of ArcGIS You Passed the Exam! Congratulations, Luis Cesar — you correctly answered 7 of the 10 questions. These questions may be asked again on the exam. You may want to review the questions below or continue to the next lesson. -------------------------------------------------------------------------------- ArcView 8 works with all core ArcGIS software products, including ArcSDE and ArcIMS®. It also works with all new ArcGIS extension products, including: Spatial Analyst ArcView 3D Analyst™ ArcPress™ Geostatistical Analyst StreetMap™ With ArcView 8 you can: explore, display, and query spatial data access a comprehensive suite of editing tools for shapefiles and perform simple feature editing on personal geodatabases perform geoprocessing operations create presentation-quality maps with a comprehensive suite of cartographic tools and wizards create reports and sophisticated two- and three-dimensional graphs perform layer-level projection (including imagery) create and manage annotation view maps from the Internet import data from and export data to a variety of common formats Each component of ArcGIS (ArcView, ArcEditor, and ArcInfo) consists of three separate applications that represent the fundamental methods people use to interact with a GIS. These methods—data, maps, and tools—form the ArcGIS user environment. Users will typically have two or all three of these applications open at the same time. The applications are: ArcCatalog—used to browse geographic data sources and create and update metadata ArcMap—used to display and query geographic data on maps and to edit and output data ArcToolbox—contains powerful tools for performing geographic analysis and data conversion Summary ArcGIS consists of three desktop components: ArcView, ArcEditor, and ArcInfo. Depending on your organization's requirements, you may employ one or a combination of the ArcGIS applications across your network. All three ArcGIS components are built using Component Object Model (COM) technology, which makes them highly customizable and extensible. Each ArcGIS component consists of three separate applications: ArcCatalog, ArcMap, and ArcToolbox. Together, these three applications represent the fundamental methods people use to interact with a GIS—data, maps, and tools. ArcCatalog is used to browse geographic data sources and create and update metadata. ArcMap allows you to display, edit, and query geographic data on maps and to output map documents. ArcToolbox is the powerful engine behind ArcGIS geoprocessing and spatial analysis funtionality. It contains many tools and wizards to help you perform geographic analyses and data conversion.

1. A GIS must provide methods for entering geographic and tabular data. True is correct! Refer to Capturing data 2. In ArcCatalog, you copy, delete, and rename shapefiles. True is correct! Refer to ArcCatalog 3. What is the most important component of a GIS? People is correct! Refer to Components of a GIS 4. Which procedure is used to determine feature spatial relationships? Topology is correct! Refer to Feature spatial relationships 7. What is a collection of features having similar attributes and located in the same geographic extent called? Layer is correct! Refer to How a GIS organizes geographic data 8. All of the following are tabs in ArcCatalog for exploring data, except one. Which one? Geography is correct! Refer to ArcCatalog 9. Your PSEArcGIS\basicsgis\Lesson02 folder contains a geodatabase called National. From the National geodatabase, USAContainer, add Capitals and States to a map in ArcMap. Turn on map tips for the Capitals layer. Zoom in to the state of California and hold your mouse pointer over the capital city. What displays as the map tip? Sacramento is correct! Refer to Explore ArcMap and ArcCatalog, Step 8 10. What kind of analysis is used to determine whether an apartment building is within 1 mile of an earthquake fault? Proximity analysis is correct! Refer to Analyzing data 12. What kind of analysis is used to integrate soils data with slope, vegetation and tax assessment data? Overlay analysis is correct! Refer to Analyzing data 14. Using ArcCatalog, navigate to your PSEArcGIS\basicsgis\Lesson02 folder. How many points does the zip coverage in the Rhode_Island workspace contain? 3 is correct! Refer to Explore ArcMap and ArcCatalog 15. Based on your work in the project exercise "Use ArcGIS to explore aquaculture in Zambia," what is the population per square mile in Zambia? 34 is correct! Refer to Use ArcGIS to explore aquaculture in Zambia, Step 2 16. Based on your work in the project exercise "Use ArcGIS to explore aquaculture in Zambia," how many countries had less than a 2000 (kcal/day) DES in the years 1996 to 1998? 8 is correct! Refer to Use ArcGIS to explore aquaculture in Zambia, Step 3 17. Based on your work in the project exercise "Use ArcGIS to explore aquaculture in Zambia," how many populated place features have no name? 502 is correct! Refer to Use ArcGIS to explore aquaculture in Zambia, Step 6 18. Based on your work in the project exercise "Use ArcGIS to explore aquaculture in Zambia," which field name is set for the map tips in the soils layer? SOIL_TYPE is correct! Refer to Use ArcGIS to explore aquaculture in Zambia, Step 4 20. Based on your work in the project exercise "Use ArcGIS to explore aquaculture in Zambia," how many populated places are within areas that have the Alfisols soil type? 139 is correct! Refer to Use ArcGIS to explore aquaculture in Zambia, Step 7 2. You can browse and retrieve geographic data available on the Internet using ArcGIS software. True is correct! Refer to ArcView 4. ArcEditor has all the capabilities of ArcView plus tools for editing ArcInfo coverages and geodatabases. True is correct! Refer to ArcGIS components 5. ArcInfo Workstation includes all of the following components, except one. Which one? Avenue is correct! Refer to ArcGIS components 6. Using ArcCatalog, navigate to your PSEArcGIS\basicsgis\Lesson02 folder. How many points does the zip coverage in the Rhode_Island workspace contain? 3 is correct! Refer to Explore ArcMap and ArcCatalog 7. Using either ArcCatalog or ArcMap, explore the USAContainer feature dataset in the National geodatabase in your PSEArcGIS\basicsgis\Lesson02 folder. Which of the following feature classes is not included in USAContainer? Time zones is correct! Refer to Explore ArcMap and ArcCatalog, Step 1 8. With ArcInfo you can edit all of the following data types, except one. Which one? Image files is correct! Refer to ArcInfo 1. Which of the following best describes the types of functions for which you would use ArcCatalog? Data management, viewing and editing metadata is correct! Refer to ArcCatalog 2. Which of the following best describes the types of functions for which you would use ArcToolbox? Overlay processing, buffer creation, map transformation is correct! Refer to ArcToolbox 4. By default, the Table of Contents is located on the left side of the ArcMap application window. True is correct! Refer to ArcMap 6. With ArcInfo you can edit all of the following data types, except one. Which one? Image files is correct! Refer to ArcInfo 7. Your PSEArcGIS\basicsgis\Lesson02 folder contains a geodatabase called National. From the National geodatabase, USAContainer, add Capitals and States to a map in ArcMap. Turn on map tips for the Capitals layer. Zoom in to the state of California and hold your mouse pointer over the capital city. What displays as the map tip? Sacramento is correct! Refer to Explore ArcMap and ArcCatalog, Step 8 -------------------------------------------------------------------------------- 1. Which of the following best describes the types of functions for which you would use ArcCatalog? Data management, viewing and editing metadata is correct! Refer to ArcCatalog 3. There are several ways to preview data in ArcCatalog. Which of the following is not an ArcCatalog preview method? Text is correct! Refer to ArcCatalog 4. You can browse and retrieve geographic data available on the Internet using ArcGIS software. True is correct! Refer to ArcView 5. All of the following are tabs in ArcCatalog for exploring data, except one. Which one? Geography is correct! Refer to ArcCatalog 6. Using ArcCatalog, navigate to your PSEArcGIS\basicsgis\Lesson02 folder. In the Rhode_Island workspace, there is a coverage named zip, with a point feature class. The table for the point feature class contains all the columns listed below, except one. Which one? CODE is correct! Refer to Explore ArcMap and ArcCatalog 19. Based on your work in the project exercise "Use ArcGIS to explore aquaculture in Zambia," what is the population per square mile in Zambia? 34 is correct! Refer to Use ArcGIS to explore aquaculture in Zambia, Step 2 8. With ArcInfo you can edit all of the following data types, except one. Which one? Image files is correct! Refer to ArcInfo 9. Using ArcCatalog, navigate to your PSEArcGIS\basicsgis\Lesson02 folder. How many points does the zip coverage in the Rhode_Island workspace contain? 3 is correct! Refer to Explore ArcMap and ArcCatalog 1. When represented on a map, how can geographic features be drawn? Points, lines, or polygons is correct! Refer to Components of geographic data 2. You can browse and retrieve geographic data available on the Internet using ArcGIS software. True is correct! Refer to ArcView 3. What are the five components of a GIS? People, data, hardware, software, procedures is correct! Refer to Components of a GIS 4. All of the following are tabs in ArcCatalog for exploring data, except one. Which one? Geography is correct! Refer to ArcCatalog 6. ArcEditor has all the capabilities of ArcView plus tools for editing ArcInfo coverages and geodatabases. True is correct! Refer to ArcGIS components

1. ArcMap automatically produces a balanced layout suited to the intended map use. True is incorrect. 2. Which two file formats are suitable for viewing maps on the Web? JPEG and PDF is correct! Refer to Distributing maps on the Web 3. What is visual hierarchy? Visual ordering, from elements that stand out to less prominent background elements is correct! Refer to Map layout using visual hierarchy 4. Which characteristic is typical of a map designed to be viewed on a computer screen (typically 72 dots per inch resolution)? Large type is correct! Refer to Designing for resolution and viewing distance 5. Which constraint determines the visual hierarchy that a map designer chooses? The purpose of the map is correct! Refer to Map layout using visual hierarchy 6. JPEG files have a large size because all data is retained for the chosen resolution. False is correct! Refer to Distributing maps on the Web 7. Storing data files and ArcMap .mxd files on different disk drives eases the task of moving map files to another computer. False is correct! Refer to Preparing to move ArcMap files 8. Before moving an ArcMap map file to a different computer, you should set the option to store relative path names for its data sources. True is correct! Refer to Preparing to move ArcMap files 9. What can you do to maintain the stability of a map layout? Turn off the "scale map elements proportionally to changes in page size" option is correct! Refer to Controlling layout stability 10. How does a mapmaker decide on the detail suitable for supporting elements, such as the scale bar and orientation indicators? The importance of the supporting element for the intended map use determines the detail of its design. is correct! Refer to Using decorative design elements 11. When laying out a map, what needs to be balanced? The arrangement of empty spaces between map elements is correct! Refer to Balancing empty spaces 12. Why might you use the Data Frame toolbar? All of the above is incorrect. Refer to Review of toolbars used in design 13. What is a primary difficulty with using boxes around map elements, such as titles, legends, and inset maps? Boxes may produce awkward gaps and spaces around elements. is correct! Refer to Balancing empty spaces 14. Which ArcMap tool is used to precisely position map elements on the layout page? Guides is correct! Refer to Using rulers and guides 15. Which export option setting can be increased to improve the detail encoded in an exported raster file? Dots per inch is correct! Refer to Raster export formats 16. What is the general type of projection suitable for maps showing data distributions? Equal area is correct! Refer to Choosing map projections 17. In ArcMap, the ruler's units and divisions determine the precision of the layout page you will be able to structure using guides. True is correct! Refer to Using rulers and guides 18. When would you refine a map layout by adjusting the position of geographic boundary lines within the map frame? When frame lines are linked to a location on another map is incorrect. Refer to Refining a layout 19. Which list places these four media in order from highest to lowest resolution? Magazine print (high), computer monitor, laser print, TV (low) is incorrect. Refer to Designing for resolution and viewing distance 20. Which statement below describes the most suitable strategy for refining a map layout in which selected map elements are not aligned? The map elements should be centered. is incorrect. Refer to Refining a layout

1. If you're planning to share a map with others, which ArcMap option should you specify? Scale map elements proportionally to changes in page size is incorrect. Refer to Controlling layout stability 2. Before moving an ArcMap map file to a different computer, you should set the option to store relative path names for its data sources. True is correct! Refer to Preparing to move ArcMap files 3. Which two file formats are suitable for viewing maps on the Web? JPEG and PDF is correct! Refer to Distributing maps on the Web 4. Multiple designs can be produced with the same map elements. True is correct! Refer to Designing for the intended audience 5. Which characteristic is typical of a map designed to be viewed on a computer screen (typically 72 dots per inch resolution)? Thin lines is incorrect. Refer to Designing for resolution and viewing distance 6. Which list of items below consists of only map elements? Scale, inset map, legend is correct! Refer to Map layout using visual hierarchy 7. JPEG files have a large size because all data is retained for the chosen resolution. True is incorrect. Refer to Distributing maps on the Web 8. What is visual hierarchy? Visual ordering, from elements that stand out to less prominent background elements is correct! Refer to Map layout using visual hierarchy 9. How does a mapmaker decide on the detail suitable for supporting elements, such as the scale bar and orientation indicators? Supporting map elements should always be as simple as possible. is incorrect. Refer to Using decorative design elements 10. What is the difference between the zoom tools on the Tools toolbar and the tools on the Layout toolbar? The zoom tools on the Tools toolbar change the view of the layout page, while the zoom tools on the Layout view change the spatial extent of data in a selected data frame. is incorrect. Refer to Review of toolbars used in design 11. When laying out a map, what needs to be balanced? The selection of complex and simple map elements is incorrect. Refer to Balancing empty spaces 12. What is the recommended way to check the suitability of selected map colors? Check their appearance in black and white is incorrect. Refer to Designing for final color quality 13. Which ArcMap tool is used to precisely position map elements on the layout page? Guides is correct! Refer to Using rulers and guides 14. Why might you use the Data Frame toolbar? To align features with the data frame is incorrect. Refer to Review of toolbars used in design 15. What is the general type of projection suitable for maps showing data distributions? Equal area is correct! Refer to Choosing map projections 16. In ArcMap, the ruler's units and divisions determine the precision of the layout page you will be able to structure using guides. True is correct! Refer to Using rulers and guides 17. Which statement about raster file resolution is true? A raster file has higher resolution than a vector file. is incorrect. Refer to Raster export formats 18. Which list places these four media in order from highest to lowest resolution? Magazine print (high), laser print, computer monitor, TV (low) is correct! Refer to Designing for resolution and viewing distance 19. Which export option setting can be increased to improve the detail encoded in an exported raster file? Dots per inch is correct! Refer to Raster export formats 20. Which statement below describes the most suitable strategy for refining a map layout in which selected map elements are not aligned? Frames around the map elements should be wider. is incorrect. Refer to Refining a layout

Raster export formats There are three basic raster export formats available in ArcMap: BMP, TIFF, and JPEG. Bitmap (.bmp) produces a pixel-by-pixel rendition of the map. TIFF (.tif) files are a raster format common in print publishing environments. If you want to send a publisher a high quality file that requires no changes, a high resolution TIFF file is often the preferred choice. JPEG (.jpg) is a raster format that is commonly used for Web publishing; it will be described in more detail in a concept coming up. The resolution settings that you can specify for export files vary among the formats. For a BMP file, you set the height and width in pixels. When the map of Joshua Tree National Park was exported to a bitmap, the default resolution choice (1056 x 816 pixels) produced a file with coarse resolution but a relatively small size (2525 KB). An enlarged section of the bitmap file is shown below. Portion of an exported map of Joshua Tree National Park (enlarged to 300 percent). You can clearly see the pixelation that the bitmap format produces. Exporting the map at three times the resolution (3168 x 2448) produced a higher quality image. Notice in the portion of the higher resolution version shown below that the type edges are not jagged and that the lines are much smoother. The pixels comprising this image are one-third the size of the coarse resulution version, so smaller features can be recorded in this bitmap file. Another version of the Joshua Tree map. This one was exported as a bitmap with a high resolution setting (3168 x 2448). The improved quality comes at a price, however—the high resolution version is much larger than the coarse version (over 22 MB). To put this size difference in context, you could store 40 coarse resolution files on a 100 MB Zip disk but only four of the higher resolution files on the same disk. The color depth setting you choose also has an effect on file size, with 24-bit files producing the largest exported files. With the TIFF (Tagged Image File Format) format, you specify resolution by choosing the number of dots per inch (dpi) in the exported file. The default is 96 dpi, which results in a coarse image. The TIFF example shown below was exported at 296 dpi, which produced a high quality image suitable for print publication. Again, this quality came at the price of large file size (over 24 MB). All raster file formats (BMP, TIFF, and JPEG) have the common characteristic that the file is built pixel by pixel. Labels, line widths, and colors are difficult to edit in raster files. Therefore, these export formats should be used for maps that you want to show or print "as is." For example, they are suitable for displaying maps in a projected presentation using software such as PowerPoint. Vector export formats There are three basic vector, or object, export formats available in ArcMap: EMF, EPS, and AI. CGM is also a vector format, but it is rarely used in print design and won't be discussed here. For the simple map of Joshua Tree, the sizes of the vector files shown below are much smaller than the sizes of the raster files you saw in the previous concept. The EMF file has a size of 42 KB, the EPS is 129 KB, and the AI file is 162 KB. Recall that raster files of comparable quality topped out at over 24 MB. The simplicity of the source map is key. An elaborate map with many small features and numerous labels could easily produce a larger vector export. EMF (Enhanced Metafile) is a multi-purpose vector format native to the Windows operating system. When exported to an EMF file, the example map did not fare well. Notice below that the type was shifted relative to the lines. The EMF file also opened in Adobe Illustrator (a vector-based graphics software program) at a very large size, which would require the extra step of resizing the work. This format does have the advantage that entire text strings remain complete, so they can be edited if necessary. Notice that "Park Boundary" can be selected as a complete phrase (as indicated by the blue line below the letters). The type is not broken into individual letters or groups of letters, as with some of the other vector formats. EPS (Encapsulated PostScript) is a common and fairly generic high quality vector format. Unfortunately, with EPS, type is exported in segments rather than complete text strings. Notice below that the "rk" in Park can be selected; the label has been broken into four segments. If you needed to edit or restyle the type, the segments would overrun each other or gaps would appear between them. The EPS export format is therefore not a useful option for maps in which type quality is important for final publication. Exporting to the AI (Adobe Illustrator) format results in complete text strings and high quality lines. "Park Boundary" can be selected as a complete string, then edited or restyled. This is important because ArcMap works only with a category of font called TrueType, but publishers often need to use a PostScript font before doing prepress processing on the map for printing. This is not usually something you need to do yourself, but your maps will be easier to publish if you construct them to make those types of changes painless. You want the process of replacing all fonts to be easy; that is, to not require tedious repairs or manual replacement of individual labels. Some special type effects were lost with the AI export, perhaps because ArcMap exports to an older version of Adobe Illustrator (3.x). Illustrator is currently at version 9, but using 3 extends compatibility to earlier Illustrator versions and other drawing packages that import AI files. In the example above, the letter spacing used for "MOJAVE" did not export, so the type is not registered with its halos. Despite the problems with halos and letter spacing, the AI format will be the most trouble-free for publication. You should be aware that some custom type effects do not export well; therefore, you should always test your choices before relying on them for a design that will need to move beyond ArcMap. The example below reflects a variety of design changes that were made in Illustrator to the AI file. Line widths, colors, and textures (dashing) were changed. The background color was lightened. The type was changed to a different font and some labels were resized and colored. Letter spacing was reapplied, and note that the green lines break before intersecting with the letters. Additionally, halos around the type were added using fonts rather than drawings (for a smaller file size). The halos around MOJAVE are light in color so you can see them, and the halo is the same color as the background around the large "J" on the right. The matched halo color is the more usual and more subtle application of this design effect. In the three graphics above, the halos for the MOJAVE label are line drawings, so the black letters do not register with their halos. This characteristic of halos is common to all the vector formats described here (EMF, EPS, and AI). When you export to EMF or EPS, you can specify the resolution. The examples above were exported at 300 dpi. The notion of dots per inch resolution is more abstract for a vector file, since the file is not made up of pixels for which you are setting the size. The dpi setting for vector formats controls how carefully lines are drawn. You will need to experiment with dpi settings to see how sensitive your particular map lines are to the dpi setting for export. With ArcMap's Adobe Illustrator (AI) export choice, you can't set the resolution, but this format does produce high quality line drawings with a dpi of 720. Big Picture Design Lesson 2 Concept Distributing maps on the Web ArcMap has two export formats suitable for displaying maps on the Web. JPEG (.jpg) is a raster format and PDF (.pdf) is a vector format. JPEG (Joint Photographic Experts Group) refers to a sophisticated compression algorithm used to make high quality bitmap files smaller. You can see the difference this compression makes by comparing the size of the example below (823 KB) to the high resolution BMP and TIFF files shown earlier, which exported at over 22 MB (over 25 times larger). JPEG is a lossy compression algorithm, which means that data is lost when a map is exported as a JPEG. In contrast, a TIFF file has a larger size because all data is retained for the chosen resolution. When exporting to JPEG, you have two controls on image quality, resolution (in dpi) and quality (ranging from low to max). The map shown above was exported to a JPEG file at 300 dpi and a setting of maximum quality. The result is a high quality image with a reasonable file size (823 KB). This size is a bit large for Web display, but the quality is good enough that it could be used for some printed contexts. The second JPEG, shown below, was saved at 300 dpi with a medium quality. The savings in file size is good (249 KB) but notice the speckled artifacts around lines and type where compression losses are visible. The reduced file size comes at the cost of a poorer quality image. This file would be fine for many Web applications (and for a PowerPoint presentation), but the quality is not good enough for print publication. JPEG files are raster files, so you cannot edit lines and type as objects using illustration software. You should use JPEGs to show your finished work when no further changes will be needed. PDF (Portable Document Format) is the vector format available for Web display. PDF files can be viewed with Acrobat Reader software, which is available free from Adobe. PDF works well for displaying a small (the file shown below is 31 KB), high quality vector image that can be panned and zoomed. You should use PDF for maps with a completely finished design and when you do not care about details of type positioning. Notice that the font has changed in the PDF file below (compare it to the JPEG shown above). Also, note that a single letter, the "k" in Park, is selected. ArcMap exports type in PDF files letter by letter. This characteristic limits the usefulness of this format. Although PDF is a vector format, type is very difficult to edit and adjusting font characteristics (such as changing a label to bold) will produce a garbled image. It is therefore not a good export format for print publication. Question Which file has the smallest size? exer6_8bit.bmp Correct. exer6_24bit.bmp exer6_72dpi.tif exer6_300dpi.tif Question Which file has the best quality? exer6_8bit.bmp exer6_24bit.bmp exer6_72dpi.tif exer6_300dpi.tif Correct. Question At a resolution of 300 dots per inch, which format provides the smallest file size? TIFF EMF Correct. EPS BMP Which format would you choose if you needed to distribute this map over the Web? When choosing the file format, consider the needs of map readers Answer to question PDF is probably your best bet here. The difference in file size is minimal between the highest quality JPEG file and the highest quality PDF file that you created. With the PDF version, however, readers can explore the map much more easily. Acrobat Reader's zoom and pan tools allow map readers to readily interpret the text symbols on the map as well as the text in the legend. In the JPEG version, the map appears small and the text is quite hard to read. Summary Some ArcMap settings have no obvious visual impact on a design, but make your map useful in wider contexts. Eliminating rescaling and setting relative path names improve the stability and transferability of a map file. Choosing appropriately among many export options lets people without ArcMap use your map files in formats they are able to view and manipulate. Quality map design is often most important when you want many people to see your map. These more general concerns, that go beyond the details of mapmaking, are an important part of being a competent map designer --------------------------------------------------------------------------------

Designing for map purpose and medium Cartographers begin planning a map by interviewing their clients to learn why they want the map made. They ask questions about expectations and budget constraints. They ask questions such as: What information is being mapped? Is the map content being coordinated with written text? Who will be reading the map? What size and media will be used? A mapmaker who is not knowledgeable about the topic being mapped reads up on it or looks at maps made for similar projects to gauge the detail needed for the new map and learn relevant symbol conventions. As more researchers begin making their own maps using GIS software, they need to ask themselves the same questions listed above, though they have the advantage of already being familiar with the details of the larger context into which the map fits. The impetus to design well comes from a desire to make maps that are clear and convincing to those who will read them. A successful design begins with knowing why the map is being made Designing for the intended audience If you are laboring over map design, you are probably making a map for people beyond yourself and your immediate work group. You intend to communicate the mapped information to others. Who are these expected map readers? If they are expert with the data being mapped, they will expect a rich and multi-layered presentation that adds to their knowledge or thoroughly supports your (the mapmaker's) contention. If the readers are a non-expert audience new to the information being mapped, they may require a simpler presentation. Likewise, if they are people who are too busy to spend much time reading, they will also need a simple overview. The more knowledge and time the audience brings to the task of reading your map, the more information you will be able to include. Maps that have a simple purpose, such as an in-car navigation display showing an address location, demand a simple design. Maps for non-expert or busy people will have a similar look—they should have a single message that focuses the attention of the reader. In contrast, maps for people who already know about the topic can be more complex. These readers are motivated to spend more time examining a map on a topic of interest. The knowledge they bring to map reading allows them to focus directly on key information. Detailed information on the map will support their map reading, rather than distract from it. When designing a map, you should also consider your audience's physical ability to read. If the audience includes older people and others likely to have reduced vision, do not make the map text so small that map reading is a struggle. If the map will be read in near-dark or other difficult viewing conditions, use exaggerated lightness contrasts. You may also choose to design a map to accommodate color-blind readers (who comprise 4 percent of the population). The two example maps below were made with the same set of lines from a map of Joshua Tree National Park in Southern California. In the example on the left, the emphasis is on physical features adjacent to the park: the San Andreas Fault (dashed), the transition zone between the Mojave and Colorado Deserts (brown), and sea level (blue). The map on the right emphasizes cultural features adjacent to the park: roads, interstate highway, and populated places. These two maps are designed from the same lines, but they have very different purposes. Designing for resolution and viewing distance Choosing how to present a map is part of designing for the map purpose. There are many ways you may be planning to display a map. Some possibilities are listed below. Each context will be best served by a different map design. Full computer screen viewed at the reader's desk Computer-projected display supporting a presentation to hundreds of people Color laser prints distributed to a working group Black and white print for a report that concerned citizens will photocopy at the library Large plot pinned up at a planning meeting for viewing across the room Page in a glossy magazine article or book that is professionally printed on an offset color press Huge backdrop at a trade show Supporting information on a documentary television show Black and white fax to an emergency response team 2-inch display on a personal digital assistant (PDA) for route planning Part of an online interface for Web-based data dissemination Your job as the mapmaker is to use design to master the constraints that a particular medium places on a map (not to complain about those constraints). Many of us have attended a talk where the presenter declares that the projector is at fault for the illegibility of the maps. Wrong. The error was made by the presenter who borrowed a design suited for another context or by the map designer who did not account for the final display constraints. If your map is printed in a book, you can use fine lines, small type, and subtle color differences. If you need that map in a projected presentation, redesign it with bolder color differences, larger type, and simpler lines to be sure the main messages hold up at coarse screen resolution, bleached by the projector and the room lights, and seen at a distance. Resolution varies widely among the media on which we see maps. A computer screen may show 72 dots of light per inch across its display. A household television has poorer resolution, about 26 dots per inch for a 27-inch TV (dpi varies with television screen size). A laser print may squeeze 600 dots of toner in an inch to build the image. A litho plate on an offset press can easily reproduce 12,000 dots per inch from an imageset negative. As you can see from the above examples, resolution can vary radically among media. Because different media have very different resolutions, your map designs must change to accommodate each medium that you are using. Map features and type must be much larger to build them with the computer screen's emitted spots of light than to reproduce them on press. A map designed for screen display will look clumsy in a magazine and a map designed for print will be illegible on-screen. There are no bad media, only maps that are not designed for their media. Viewing distance affects map design just as resolution does. Features need to be enlarged to be visible from a distance. Letters 2 inches high that are seen at a distance of 14 feet are approximately the same size as 10-point type seen at a reading distance of 1 foot. A line 2 points wide is practically invisible at a far-away distance, so line widths also need to be increased to retain visibility. (Points are small units of measurement; 1 point equals one-seventy-second of an inch). Similarly, color differences need to be stark to distinguish small features, whether they are small in measured dimensions or small because of the viewing distance. The maps below show land use in Clark County, Washington. An enlargement of the left map's inset area is shown on the right. The enlarged map section has been designed with fine lines and type that would be suitable for reproduction in print. Linking layout to map purpose Knowing a map's purpose allows you to decide on a visual hierarchy and projection for the map. The hierarchy structures some elements as foreground and some as background. The foreground elements are more visually prominent—they are the parts of the map you want people to notice first and remember after they finish reading the map. You should design map elements that supply supporting information so that they have decreasing visual importance, in accordance with their importance to understanding the mapped information. Choosing a map projection is also a design decision. You decide how to project the round Earth onto the flat page so that unavoidable distortions in the geography do not interfere with the map reading tasks you intend for the map. Visual hierarchy is probably the most important design principle that cartographers use. The map's purpose determines which visual aspects of the map should be most important. These important elements should be most visually prominent and highest in the visual hierarchy. Information that supports the main message of the map is referred to as base information and should be lowest in the visual hierarchy. The elements from which a map layout may be built are: Main map Smaller scale inset maps showing location Larger scale inset maps showing detail or locations outside the area of the main map Titles Subtitles Legends Scale indicators Orientation indicators Graticule Explanatory text notes Source note Neatline Photos Graphs The list of elements can be extensive for a complex project. For any mapping project, design is largely a process by which you decide how prominent to make each element as you build a hierarchy of information. Hierarchy is established by position on the page, or screen, and the size of an element. A note in small text in the lower left corner would be lower in the hierarchy than a title in large text that is centered across the top of the map. Contrasting colors, line weights, and line detail also establish hierarchy. Elements for a vegetation map of the Democratic Republic of Congo are shown below. The elements are not arranged in any particular order on the page, and this lack of planning produces a cluttered and unclear product. From top to bottom, the elements shown are: Title and location inset map Source note and subtitle for detailed inset map Legend and detailed inset map Scale for main map Orientation indicator (north arrow) and main map with graticule (lines of latitude and longitude) Neatline Using decorative design elements In addition to the many map elements used to build a layout, there are various graphic elements and effects that may be chosen to decorate it, including: Drop shadows Line stylings for frames Background patterns Full compass rose Zoom lines Colorful logos Decorative type fonts Choosing map projections Many mathematicians have been intrigued by the interesting puzzle of projecting the round globe onto a flat map. My students and I had fun physically acting this out using an old globe that I found on the roadside one day: stomping and pulling and tearing this poor old carcass into a flat surface. All projected maps are distorted in some way (and more kindly). Therefore, you should choose a map projection that relegates those distortions to places on the map that are not important for your message. That challenge makes projection selection a design decision, because it depends on the purpose of the map. If you are making a detailed map of a small land area (a large scale map), the particulars of map projection will matter less, unless map readers will need to make detailed measurements from the map. If you are mapping larger areas—all the U.S. states, for example—you should put more thought into projection settings. For country and continent mapping, projection becomes an important decision. If you see a map of the United States that looks like a rectangular slab, with a straight-line U.S.-Canada border across the west, be suspicious of the mapmaker's knowledge of map projection and of interpretations of the mapped data. For example, if you want to understand the road network on a map with a poorly chosen map projection, you will not know whether roads look sparse in an area because it is underdeveloped or because the map is distorted in a way that happens to expand that part of the map. Likewise, maps of point patterns or area densities need equal area base maps for accurate interpretation. The map of western Canada, below, was produced with a Plate Carrée projection. One degree of latitude is equal to one degree of longitude to form a square grid (this is sometimes misnamed "no projection"). That seems like a fine idea until you remember that degrees of longitude get smaller and smaller as you near the poles. The provinces, and especially the northern islands of Canada, are stretched horizontally because of this distortion. East-west scale (degrees of longitude) get bigger and bigger as you go north on this map. Judging the density of roads or the areas of national parks is difficult with a projection that results in distortions like these. Distorted Plate Carrée map projection of a portion of Canada. A more suitable projection of the same area of Canada is shown below. This map was made using an Albers Equal Area projection with two standard parallels (lines of true scale) running through the area of interest (at 50 and 70 degrees latitude). Notice how differently proportioned the provinces are on this map compared to the map above. Notice also how much bigger Wood Buffalo National Park is than the parks to the south, and the openness of the southerly road network. Areas are correct all over this map, so density of features, such as roads and small lakes, can be accurately judged. Albers Equal Area map projection of Canada. The projection used also affects the shape of map areas, which in turn constrains the size and layout of the map. You can see how much of northern Canada is not shown in the map above that accurately represents areas. If northern Canada was relevant for this map, a larger frame or a smaller scale would be needed to suit the map purpose. For thematic mapping, remember to design your map with an equal area map projection. Despite all the fun we could have with projection distortions, this is really the only piece of information I want you to remember: If you are mapping data distributions, choose an equal area projection. If you are mapping the mainland of the United States, the Albers Equal Area projection customized to the U.S. is a common projection choice. ArcMap makes it easy to use Albers and also offers customized Albers projections for Alaska and Hawaii. These versions of Albers reposition standard parallels so that no part of the area of interest gets far from these lines where there is no distortion. Again, these projections are customized to suit the map purpose, so their use is a map design decision. Another basic category of map projection is conformal. Large scale reference maps often use conformal projections. Conformal projections are better for showing routes or shapes because they preserve angles rather than areas. You can find additional information about map projections in the ArcGIS Desktop Help Planning a layout Geographic areas are often irregularly shaped, and a novice designer may be tempted to fill the corners and voids in a display with the remaining elements of the map. Unfortunately, some designs evolve like this: "I see a big hole in a lower corner of my map, so I will use a large compass rose to fill in that problem area." If that sounds like familiar thinking, your future maps will benefit from design practice. The problem with the "fill-in" strategy is that the resulting overly large or bold map elements are at the wrong level in the visual hierarchy of the map. Experimenting with design can reveal new and more effective arrangements of elements for a map display. Finishing touches to the details of how elements align and fit together can transform a sloppy piece into an authoritative piece. And knowing how to get the most out of others' critiques lets you finish a project with confidence. Balancing empty spaces If the goal of page layout is not consistent filling in, what is it? Page layout is the act of balancing empty spaces. If you have an empty space on the page in one corner, you can position other map elements to produce empty spaces that are similar in size in other parts of the page to balance that gap. Learn to see and use the empty spaces between elements when you are designing a page layout. Unnecessary boxes around map elements often produce gaps and spaces that interfere with designing an attractive and balanced layout of map elements Refining a layout A map layout makes better sense if elements that are conceptually related to one another are laid out so they are physically closer to one another. This seems obvious but, in a layout with many map elements, it can be difficult to accomplish. For example, with one large map and a few small inset maps, you may end up with the scale bar for the large map closer to an inset than to the main map, producing a confusing association. A series of maps, each with explanatory text, is well laid out if spacing unambiguously relates each text block to the map it describes. A general explanation for the entire layout also functions well if it stands on its own—not isolated, necessarily, but not visually associated with a particular element through proximity. The scale bar in the portion of the map below might be seen as part of a list of information about the main map on the right. Its location is more likely to be confusing because it is equally close to three maps, each at a different scale. This layout fails because the importance of proximity is not considered. Benefiting from layout experimentation and critique In addition to planning hierarchies and balancing empty spaces, a good dose of experimentation often improves a map design. Novice designers tend to place the map elements in positions that seem obvious and workable. They may adjust these positions or change the sizes of elements slightly to improve the layout, but the initial arrangement of elements on the page is not questioned. Before you start making small adjustments to improve a layout, push yourself to think of some arrangements that are radically different from the first one that you are assuming will work. Change the page orientation from portrait to landscape and see how elements fit together. Move elements from the top of the page to the bottom. Try pulling them into a more compact arrangement with overlapping elements. For example, overlay titles and text blocks on some conveniently open areas of the map. You may come back to the first layout in the end, but this experimentation is an important first step in map design. The vegetation map seen in previous examples is shown in portrait (tall) and landscape (wide) orientations below. Both arrangements are well balanced and have a minimal difference in visual hierarchy. Equally important as experimentation is asking other people to judge your draft layout for a map. When you ask a person to critique your work, your job is to be quiet and let them do what you asked. A critique is not an opportunity to explain and defend your decisions. You will adjust or discard many of their suggestions, but do that after you hear them out. During the critique, ask them to elaborate on the reasons behind their ideas and interpretations but do not spend time debating them. A draft map usually has unfinished aspects, such as incomplete text, nonsense colors, and downright errors. The person doing the critique will always zero in on these details first. Acknowledge that the work is a draft and encourage them to look at the big picture, the overall layout. Help them get past the details. Details are easier to edit than understanding the larger scope of a project. The person you have pushed into a quick critique is not immersed in the design or map content as you are; do not despair if you cannot get them beyond details. Thank them for their sharp-eyed edit, then choose another person who is less distracted and perhaps able to engage the larger challenge of making sense of the page. You should ask a few people for suggestions and balance their critiques. Pay attention to their reasoning and suggestions, but be aware that points of confusion can be improved by making changes different than the ones your critics suggest. For example, one critic may suggest that legend boxes be made larger so they are more visible and another may suggest spacing the boxes. You may decide that changing the position of the legend so the boxes are not as close to the colorful main map makes them more visible, addressing both concerns without making either change that was suggested. A critique is raw material that pushes you to experiment and refine your decisions. It also keeps you honest—it prevents you from going forward with convoluted solutions that you have thought about too much. Balancing empty spaces If the goal of page layout is not consistent filling in, what is it? Page layout is the act of balancing empty spaces. If you have an empty space on the page in one corner, you can position other map elements to produce empty spaces that are similar in size in other parts of the page to balance that gap. These open areas are useful; they offer a welcome break from the visually dense information of your map and text blocks. They open up a complex page so that relationships between elements can be understood at a glance by the groupings that the open areas separate. Summary When you are creating a page layout, you should size each map element relative to its importance for the map purpose. Think about the logic of the position of each element relative to other elements. Then step back, squint your eyes, and look explicitly at the arrangement of empty spaces on your page. Designing the positions and shapes of those empty spaces is a key to good page layout. Numerous graphic effects can be produced in ArcMap. Your decisions to use them, change them, or (most importantly) turn them off are guided by the visual hierarchy of information you choose to construct as you design the map. A clear understanding of the hierarchy for map elements to suit your map purpose is the essence of map design. Designs that do not follow a logical hierarchy are cluttered, confusing, and hard to read. Map design is easy when you know hierarchy; the result is crisp, organized, inviting, and to the point

1. What is the defining characteristic of an “exact” interpolator? Every parameter value of the interpolator is user-specified. Its predictions are exactly equal to the data value for sample data locations. The root mean square error for the interpolated surface is zero. It has a stochastic component. 2. Which of the following is a disadvantage of the Inverse Distance Weighting interpolator? It creates “bull’s-eyes” around sample data locations. It assumes that the data comes from a stationary process. It is hypersensitive to sample points that lie near the edge of the study area. It can be difficult to determine a good search neighborhood. 3. Root mean square error is a summary measurement of the difference between predicted and actual values for a data set. True False 4. Global polynomial and local polynomial are smooth interpolators. True False 5. What is the range of values for a probability map? 0 - 10 0.01 – 0.210 0 – 100 0 – 1 6. Which map shows the upper or lower limits on true values for locations? Prediction map Quantile map Probability map Standard Error map 7. Spatial autocorrelation is the statistical correlation between spatial random variables of the same type where the correlation depends on the distance between locations. How can this idea be expressed in simple terms? The rate at which things change over distance tends to be the same everywhere on the earth’s surface. If things change very little over a long distance, any change that comes is likely to be drastic. Things close together in space tend to be more alike than things far apart. A prediction that is accurate for one place will probably be accurate for any similar place. 8. One assumption Kriging makes about random error is that it is spatially autocorrelated. What is another? It is more accurately modeled by Universal Kriging than by Ordinary Kriging. It is the same no matter which semivariogram model is used. It does not significantly affect predictions unless there is a heavy directional bias in the data. On average, it is zero. 9. In a normal distribution, what is the probability that the true value of a location is within one standard error of the predicted value? 50% 32% 68% 100% 10. The values in a prediction map are lower than the corresponding values in a .10 quantile map. True False 11. What is the name of the assumption that the autocorrelation of random error depends only on distance and not on the location of data? Cross-correlation Covariance Stationarity Standard error 12. What is the probability that a location’s true value exceeds the value assigned to it on a .95 quantile map? 5% 50% 95% 21% 13. In local polynomial interpolation, you can move a slider bar to make the interpolation less local and more global. One effect of this is that the search neighborhood becomes larger. What is another? The order of the polynomial used to model the trend increases. Sample points nearer the prediction location become relatively more influential. The order of the polynomial used to model the trend decreases. Sample points farther from the prediction location become relatively more influential. 14. For a one-sector search neighborhood, Neighbors to Include is set to 5 and Include at Least is set to 3. What happens if a certain neighborhood contains only two points? The search stops at the neighborhood boundary and only two points are used. The search is expanded beyond the neighborhood until one more point is found. The search is expanded beyond the neighborhood until three more points are found. The search stops at the neighborhood boundary. The mean value of the data set is arbitrarily weighted at 10% and used as the “third point.” 15. If Location 1 has a higher predicted value than Location 2, it must also have a higher probability of exceeding any given threshold value. True False 16. Assume that a location with a very low prediction map value has a very high quantile map value. Which of the following is a plausible explanation for this? The location has a high probability threshold value. The location has a large standard error value. The sample data is not normally distributed. The surface has negative spatial autocorrelation. 17. Standard errors are smaller near sampled data points and larger in areas where this is little sample data. Which principle explains this? The bell curve. Spatial autocorrelation. Anisotropy. Stationarity. 18. In a search neighborhood, one sample point is red, one is brown, and eight are dark green. This evidence suggests that local influence on the prediction location decays slowly. True False 19. A “local outlier” is a data point whose value is extreme compared to nearby points but not compared to the full range of data values. Which tool is good for identifying local outliers? A trend analysis graph. A semivariogram. A cross validation chart. A histogram. 20. Histograms are used to visualize the overall distribution of a data set. What else are they used for? To find anisotropic tendencies in the data. To identify outliers in the data. To normalize skewed data. To estimate the spatial autocorrelation of the data.

1. What is the defining characteristic of an “exact” interpolator? Its predictions are exactly equal to the data value for sample data locations. is correct! Refer to Inverse Distance Weighted 2. Which of the following is a disadvantage of the Inverse Distance Weighting interpolator? It creates “bull’s-eyes” around sample data locations. is correct! Refer to A synopsis of the interpolators 3. Root mean square error is a summary measurement of the difference between predicted and actual values for a data set. True is correct! Refer to A tour of Geostatistical Analyst (Part 2) 4. Global polynomial and local polynomial are smooth interpolators. True is correct! Refer to Exact versus smooth interpolators 5. What is the range of values for a probability map? 0 – 1 is correct! Refer to Probability map 6. Which map shows the upper or lower limits on true values for locations? Probability map is incorrect. Refer to Quantile Map 7. Spatial autocorrelation is the statistical correlation between spatial random variables of the same type where the correlation depends on the distance between locations. How can this idea be expressed in simple terms? Things close together in space tend to be more alike than things far apart. is correct! Refer to Kriging: More about autocorrelation 8. One assumption Kriging makes about random error is that it is spatially autocorrelated. What is another? On average, it is zero. is correct! Refer to Kriging 9. In a normal distribution, what is the probability that the true value of a location is within one standard error of the predicted value? 68% is correct! Refer to Quantile map 10. The values in a prediction map are lower than the corresponding values in a .10 quantile map. False is correct! Refer to Quantile map 11. What is the name of the assumption that the autocorrelation of random error depends only on distance and not on the location of data? Stationarity is correct! Refer to Assumptions: Remind me about stationarity 12. What is the probability that a location’s true value exceeds the value assigned to it on a .95 quantile map? 5% is correct! Refer to Quantile map 13. In local polynomial interpolation, you can move a slider bar to make the interpolation less local and more global. One effect of this is that the search neighborhood becomes larger. What is another? Sample points farther from the prediction location become relatively more influential. is correct! Refer to Use Local Polynomial and Radial Basis Functions: What does the slider bar do? 14. For a one-sector search neighborhood, Neighbors to Include is set to 5 and Include at Least is set to 3. What happens if a certain neighborhood contains only two points? The search is expanded beyond the neighborhood until one more point is found. is correct! Refer to Use Inverse Distance Weighted: Searching for neighbors by sector 15. If Location 1 has a higher predicted value than Location 2, it must also have a higher probability of exceeding any given threshold value. False is correct! Refer to Probability map 16. Assume that a location with a very low prediction map value has a very high quantile map value. Which of the following is a plausible explanation for this? The location has a large standard error value. is correct! Refer to Quantile map 17. Standard errors are smaller near sampled data points and larger in areas where this is little sample data. Which principle explains this? Spatial autocorrelation. is correct! To say that standard error is smaller at locations near sample data is the same as saying that predictions are more accurate at locations near sample data. Without the principle of spatial autocorrelation, we would have no reason to suppose that this was true--the distance between a sample point and a prediction location would be irrelevant. 18. In a search neighborhood, one sample point is red, one is brown, and eight are dark green. This evidence suggests that local influence on the prediction location decays slowly. False is correct! Refer to A tour of Geostatistical Analyst (Part 2): Interpreting the data point colors 19. A “local outlier” is a data point whose value is extreme compared to nearby points but not compared to the full range of data values. Which tool is good for identifying local outliers? A semivariogram. is correct! Refer to A tour of Geostatistical Analyst (Part 1) 20. Histograms are used to visualize the overall distribution of a data set. What else are they used for? To identify outliers in the data. is correct! Refer to A tour of Geostatistical Analyst (Part 1): Interpreting a histogram

20. Assume that a location with a very low prediction map value has a very high quantile map value. Which of the following is a plausible explanation for this? The location has a large standard error value. is correct! Refer to Quantile map Question Assume, as in the example above, that a location’s prediction value is 0.109, the standard error for the surface is 0.0193, and the error distribution is normal. Which of the following values has a less than 68 percent probability of being the location’s true value? 0.088 Correct. 0.127 0.095 0.111 8. In Kriging, the two components of a prediction are the trend and random error. True is correct! Refer to Kriging 16. Standard errors are smaller near sampled data points and larger in areas where this is little sample data. Which principle explains this? Spatial autocorrelation. is correct! To say that standard error is smaller at locations near sample data is the same as saying that predictions are more accurate at locations near sample data. Without the principle of spatial autocorrelation, we would have no reason to suppose that this was true--the distance between a sample point and a prediction location would be irrelevant. 17. In a radial basis functions interpolation, the parameter value sets the weighting of neighborhood sample points. How is the optimal parameter value determined? It is the value that results in the lowest root mean square (RMS) error in cross validation. is correct! Refer to Use Local Polynomial and Radial Basis Functions: What does optimizing the parameter value do? 18. A “local outlier” is a data point whose value is extreme compared to nearby points but not compared to the full range of data values. Which tool is good for identifying local outliers? A semivariogram. is correct! Refer to A tour of Geostatistical Analyst (Part 1) 9. A single plot point in a semivariogram cloud represents which of the following? A pair of sample point locations. is correct! Refer to A tour of Geostatistical Analyst (Part 1): Interpreting the semivariogram/covariance cloud 10. For a one-sector search neighborhood, Neighbors to Include is set to 5 and Include at Least is set to 3. What happens if a certain neighborhood contains only two points? The search is expanded beyond the neighborhood until one more point is found. is correct! Refer to Use Inverse Distance Weighted: Searching for neighbors by sector 11. The values in a prediction map are lower than the corresponding values in a .10 quantile map. False is correct! Refer to Quantile map 12. What is the name of the assumption that the autocorrelation of random error depends only on distance and not on the location of data? Stationarity is correct! Refer to Assumptions: Remind me about stationarity 14. If a data set has strong positive spatial autocorrelation, then semivariogram plot points that are near zero on the x-axis will also tend to be near zero on the y-axis. True is correct! Refer to A tour of Geostatistical Analyst (Part 1): Interpreting the semivariogram/covariance cloud 17. If Location 1 has a higher predicted value than Location 2, it must also have a higher probability of exceeding any given threshold value. False is correct! Refer to Probability map 18. Histograms are used to visualize the overall distribution of a data set. What else are they used for? To identify outliers in the data. is correct! Refer to A tour of Geostatistical Analyst (Part 1): Interpreting a histogram 19. Standard errors are smaller near sampled data points and larger in areas where this is little sample data. Which principle explains this? Spatial autocorrelation. is correct! To say that standard error is smaller at locations near sample data is the same as saying that predictions are more accurate at locations near sample data. Without the principle of spatial autocorrelation, we would have no reason to suppose that this was true--the distance between a sample point and a prediction location would be irrelevant. 20. In a search neighborhood, one sample point is red, one is brown, and eight are dark green. This evidence suggests that local influence on the prediction location decays slowly. False is correct! Refer to A tour of Geostatistical Analyst (Part 2): Interpreting the data point colors 1. Which map estimates the values of a spatial phenomenon at unsampled locations on the basis of known values at sampled locations? Prediction map is correct! Refer to Prediction Map 3. Inverse Distance Weighted is a fairly inflexible interpolator—it has few parameters to set and therefore the user has little control over the output. What is another inflexible interpolator? Global polynomial is correct! Refer to Flexibility 5. What is the defining characteristic of an “exact” interpolator? Its predictions are exactly equal to the data value for sample data locations. is correct! Refer to Inverse Distance Weighted

2002 8:38:38 PM | Luis Nunes] 1. Global polynomial and local polynomial are smooth interpolators. True is correct! Refer to Exact versus smooth interpolators 2. In Kriging, the two components of a prediction are the trend and random error. True is correct! Refer to Kriging 3. Inverse Distance Weighted is a fairly inflexible interpolator—it has few parameters to set and therefore the user has little control over the output. What is another inflexible interpolator? Global polynomial is correct! Refer to Flexibility 4. “Random error” can also be described as which of the following? Fluctuation from a trend in an unknown direction. is correct! Refer to Kriging 20. Standard errors are smaller near sampled data points and larger in areas where this is little sample data. Which principle explains this? Spatial autocorrelation. is correct! To say that standard error is smaller at locations near sample data is the same as saying that predictions are more accurate at locations near sample data. Without the principle of spatial autocorrelation, we would have no reason to suppose that this was true--the distance between a sample point and a prediction location would be irrelevant. 8. Root mean square error is a summary measurement of the difference between predicted and actual values for a data set. True is correct! Refer to A tour of Geostatistical Analyst (Part 2) 10. The values in a prediction map are lower than the corresponding values in a .10 quantile map. False is correct! Refer to Quantile map 11. A single plot point in a semivariogram cloud represents which of the following? A pair of sample point locations. is correct! Refer to A tour of Geostatistical Analyst (Part 1): Interpreting the semivariogram/covariance cloud 12. What is the probability that a location’s true value exceeds the value assigned to it on a .95 quantile map? 95% is incorrect. Refer to Quantile map 13. The larger the standard error, the greater the uncertainty about the true value of a prediction location. True is correct! Refer to Ordinary Kriging methods 14. What is the name of the assumption that the autocorrelation of random error depends only on distance and not on the location of data? Stationarity is correct! Refer to Assumptions: Remind me about stationarity 15. Assume that a location with a very low prediction map value has a very high quantile map value. Which of the following is a plausible explanation for this? The location has a high probability threshold value. is incorrect. Refer to Quantile map 16. In a radial basis functions interpolation, the parameter value sets the weighting of neighborhood sample points. How is the optimal parameter value determined? It is the value that results in the lowest root mean square (RMS) error in cross validation. is correct! Refer to Use Local Polynomial and Radial Basis Functions: What does optimizing the parameter value do? 17. A “local outlier” is a data point whose value is extreme compared to nearby points but not compared to the full range of data values. Which tool is good for identifying local outliers? A histogram. is incorrect. Refer to A tour of Geostatistical Analyst (Part 1) 18. If Location 1 has a higher predicted value than Location 2, it must also have a higher probability of exceeding any given threshold value. False is correct! Refer to Probability map 19. In a search neighborhood, one sample point is red, one is brown, and eight are dark green. This evidence suggests that local influence on the prediction location decays slowly. False is correct! Refer to A tour of Geostatistical Analyst (Part 2): Interpreting the data point colors 20. Standard errors are smaller near sampled data points and larger in areas where this is little sample data. Which principle explains this? Spatial autocorrelation. is correct! To say that standard error is smaller at locations near sample data is the same as saying that predictions are more accurate at locations near sample data. Without the principle of spatial autocorrelation, we would have no reason to suppose that this was true--the distance between a sample point and a prediction location would be irrelevant. [edit] [7/29/2002 8:19:46 PM | Luis Nunes] 1. What is the range of values for a probability map? 0 – 1 is correct! Refer to Probability map 3. In Kriging, the two components of a prediction are the trend and random error. True is correct! Refer to Kriging 4. Inverse Distance Weighted is a fairly inflexible interpolator—it has few parameters to set and therefore the user has little control over the output. What is another inflexible interpolator? Global polynomial is correct! Refer to Flexibility 5. Global polynomial and local polynomial are smooth interpolators. True is correct! Refer to Exact versus smooth interpolators

In this lesson, you will: Learn about the four types of surfaces you can produce with Kriging: prediction, standard error, quantile, and probability Create each of these surfaces Review the Geostatistical Analyst interpolation methods and compare them to each other Surface types Since you are working in a GIS, your ultimate goal of data analysis is to produce an interpolated map from your point data. However, Geostatistical Analyst allows you to make several types of maps, including prediction, standard error, quantile, and probability maps. This topic discusses these various maps so that you can make the best choice before you begin your analysis. Prediction Map You often want to make predictions from sampled data. This data consists of measurements of a geographic phenomenon taken at several locations. From these measurements, you can predict values for locations that have not been measured. A map of these predictions is called a prediction map, or an interpolated map. (Remember—interpolation is the mathematical prediction of unknown values on the basis of known values.) Here are a few examples of sampled data that is used to make predictions: A federal agency takes air quality samples from various locations around the state to monitor pollutants. An oil company drills samples into the ground to measure the amount of oil-producing rock at various locations. A farmer takes small soil samples from various locations around his field and measures them for important nutrients and minerals for his crops. The goals of these examples are similar. We would like to know the magnitude of a certain phenomenon across, above, or below a landscape. Due to practicality and cost, however, we can only measure the occurrence at a limited number of locations. A Real Example The US Environmental Protection Agency is responsible for monitoring atmospheric ozone concentration in California. Ozone concentration is measured at monitoring stations throughout the state. The locations of the stations are shown below. There are 193 ozone monitoring stations in California. [Click to enlarge] We know the concentrations of ozone at each station, but we would also like to know the concentration for any other location in California. However, it is impossible to put monitoring stations everywhere. Therefore, we will establish spatial relationships between the known values at our observed locations and use these relationships to make predictions at unobserved locations. We can do this everywhere in the state and create an interpolated map of the predicted values of ozone. An ozone prediction map is shown below. The prediction map shows a predicted ozone value for every location in the state. [Click to enlarge] After creating the prediction map, we can use the Identify tool to show the predicted value for any location. Standard Error Map We just learned about prediction maps. A natural question to ask is, “How good are the predictions?” It is unlikely that our predictions are exactly the same as the real values (if we had actually measured them). A standard error map quantifies the uncertainty of our predictions. The larger the standard errors, the more uncertain are our predictions. Those who are familiar with regression and basic statistics will remember that confidence intervals are based on standard errors. A standard error is simply the square root of the variance of a prediction or estimate. Here is a standard error map for the ozone prediction map. This standard error is associated with the prediction value of 0.109154 for this location (shown in the previous concept). Notice that there are no data locations nearby, and the standard error is quite high. [Click to enlarge] Notice that the standard errors are smallest near sampled data locations (the darker the color, the higher the standard error value). This makes sense. Our predictions will be more accurate the closer they are to a location where we have observed and measured the values. Here is another example: This standard error is associated with a prediction location that is surrounded by other nearby data locations, and the standard error is much lower. [Click to enlarge] The location indicated by the arrow has a similar predicted value (0.109) to the location in the previous graphic, but here the standard error is much smaller (0.0128 as opposed to 0.0193). The reason is that the location is surrounded by more sampled data. Quantile map We have just learned about predictions and standard errors. Together, they provide the information we need to make a quantile map. The values of a quantile map reflect the upper or lower limits of the true values. The quantile map does not try to show our best guess about the value of a location; rather, it shows the values we are pretty sure the true values do not exceed (or the values we are pretty sure the true values do not fall below). "Pretty sure" refers to a level of confidence that we choose. A .95 quantile map shows values that are upper limits on the true values for 95 percent of the cases. (The true value will only be greater than the quantile map value one time in twenty.) Similarly, a .90 quantile map shows values that are upper limits for 90 percent of the cases. If you look at the other end, a .10 quantile map shows values that are lower limits on the true values for 90 percent of the cases. (The true value will only be smaller than the quantile map value one time in ten.) Quantile maps assume that the difference between the values on a prediction map and the true values are normally distributed. So now we need to know a little about statistical distributions. In our prediction map, we used the Identify tool to show the prediction at the location with the coordinates (-2152374.05, 80189.19). The predicted value was 0.109 with a standard error of 0.0193. In the figure below, the prediction forms the center of the bell-shaped distribution, and the standard error forms the width. A normal distribution and its relationships to predictions, standard errors, and quantiles. [Click to enlarge] This is a normal distribution. The distribution of the true value from the predicted value is shown by the curve. The total area under the curve is equal to one. The area between the predicted value (0.109) and ± 1 standard error is about 68 percent of the total area. This means that the true value should be within this interval (predicted value ± 1 standard error) about 68 percent of the time. Find the value with the lowest probability Now, suppose that you want a surface where the true value is less than the surface value 95 percent of the time. This means finding values like the one given by the red line in the figure above, where 95 percent of the area is to the left of the red line. For this example, that value turns out to be 0.141. If we find these values for the whole surface, we get a quantile map like that shown below. A prediction map compared to a 0.95 quantile map. [Click to enlarge] The 0.95 quantile map has considerably higher values than the prediction map because we are trying to find a surface that has a high probability of being above the true value. The quantile surface values depend on two factors: the predicted value and the standard error. If either one is high, it tends to make the quantile surface value high. Thus, in northern California, where the predicted values are quite low, the quantile map values are still high because the lack of data cause the standard errors to be large. When we talked about standard error maps, we compared two locations that had prediction values of 0.109. The first had a standard error of 0.0193 and the second had a standard error of 0.0128. Because the second location has a lower standard error, it will also have a lower 0.95 quantile value. In the figure below, the probability distribution for the first point is shown in blue; the distribution for the second point in orange. The area under each curve is equal to 1, but the width of the curve is determined by the standard error. When the standard error is smaller, the curve is narrower and the limiting values are closer to the predicted value. [Click to enlarge] We can also form a 0.05 quantile map as a “lower bound,” which is shown below. The values of the 0.05 quantile map are lower than the prediction map, just as the 0.95 values are higher. [Click to enlarge] Together, the quantile maps above act like confidence intervals in traditional statistics. That is, for each location we are 90 percent certain that the true value lies between the lower (0.05) quantile map and the upper (0.95) quantile map. The prediction map is our best prediction. Quantile maps can be a little confusing, so let’s summarize the four essential features of a quantile map. For each location: The prediction forms the center of the bell-shaped distribution. The standard error forms the width of the bell-shaped distribution. A probability between 0 and 1 is chosen. Call this value x. Steps 1 and 2 above characterize the bell-shaped distribution, so the appropriate quantile is determined from the probability that you choose. The quantile map value is that value with 100x percent probability that the true value of the location is less than the quantile map value. Probability map A probability map is the flip side of a quantile map. For a quantile map, we choose a probability to obtain a quantile. For a probability map, we chose a value (often called a threshold) to obtain the probability. Probability maps answer the question, “How certain am I that the true value is above (or below) some threshold value?” Suppose that ozone values above 0.11 cause certain pollution controls to go into effect. Before enacting controls for unmeasured locations, we might ask how certain we are that the ozone concentration exceeds 0.11. Below is a map of the probability that ozone exceeds 0.11. Note that probabilities are expressed as values between 0 and 1. The probability of exceeding a threshold is determined from predicted values, the error distribution, and the specified threshold value. Examples are given for four locations where the threshold value is 0.110. [Click to enlarge] Notice that for Location 1, the normal distribution is quite wide compared to Location 2. Thus, even though the predicted value is lower for Location 1 than for Location 2 (0.080 vs. 0.091), we are less certain about the value at Location 1—so we must admit that there is a higher probability that Location 1 could be above the threshold of 0.11. Consider another example. Locations 3 and 4 have the same predicted value (0.114), but because Location 3 is surrounded by more data, it has a smaller standard error (and thus a narrower normal distribution). We are more certain of the value at Location 3, so it has a higher probability of exceeding 0.11. Like quantile map values, probability map values depend on both the predicted value and the standard error. For a given standard error, higher predicted values (relative to the threshold value) increase the probability that the true value exceeds the threshold. If the predicted value is below the threshold, then smaller standard errors decrease the probability that the true value exceeds the threshold; if the predicted value is above the threshold, then smaller standard errors increase the probability that the true value exceeds the threshold. So, here are the 4 essential features of a probability map. For each location: The prediction forms the center of the bell-shaped distribution. The standard error forms the width of the bell-shaped distribution. A data value, called a threshold, is chosen. The probability map values give the probability that the true value exceeds the threshold.