Slide 1 Cartography: Modeling GIS data for Mapmaking Aileen Buckley, Ph.D. Senior Cartographic Researcher, ESRI Barbara Buttenfield, Ph.D. Professor, University of Colorado-Boulder Thank you coming to learn more about how to model GIS data for map making. My name is Aileen Buckley, I m a cartographic researcher at ESRI and for the past two plus years I have been working with Charlie Frye on the development of the data model for multipurpose multiscale mapping. Charlie is head of the Cartography Research Team at ESRI and has been working on the development of this data model for almost four years now. (Point out Charlie.) We have also been working with Barbara Buttenfield and her students from the University of Colorado-Boulder. Before we get started, babs and I will take a few minutes to introduce ourselves and then we would like to learn a little more about all of you.
Slide 2 What kind of maps do you make? Special Use Maps Trail Map Bike Map Reference Maps Topo Map Atlas Reference Map Analysis Results Suitability Map Crime Hot Spots Thematic Maps Census Data Atlas Thematic Map Ask the audience about themselves: How many make maps regularly? How many use only GIS software to do that? How many have a database background? How many attended the Cartography: Making Better Maps session in the morning today? Other questions?
Slide 3 Agenda for Seminar 1:30-2:05 Introduction Part 1: Differences between GIS data and cartographic data Part 2a: Describing cartographic workflows 2:05-2:20 Exercise 2:20-2:45 Part 2b: Modeling GIS data to support cartographic workflows 2:45-3:00 Exercise 3:00-3:30 Break 3:30-3:50 Part 3a: Cartographic data model 3:50-4:05 Exercise 4:05-4:30 Part 3b: Supporting multi-scale multi-purpose mapping Part 4: Cartographic workflows in practice 4:30-4:45 Exercise 4:45-5:00 Conclusion Before we get started, let s go over the schedule for today s seminar as well as a few logistics. Schedule Restrooms Break
Slide 4 Pre-conference seminar materials Slides Exercises Cartography Data Model Cartography Tables References Web URL Because we cannot go into as much detail as we would like in the time we have together, we have provided you with lots of resources that you can take with you and refer to later. These are compiled in the book you have in front of you. In additional to all the slides you will see today, and their related notes, you also have the exercise worksheets and all the answers (don t peek at these until after the exercises are complete!), the data model poster and schema, an example Cartography table, and a list of references that you can draw on, including those you can find on the ESRI web site.
Slide 5 Part 1 Differences between GIS data and cartographic data 15 MINUTES
Slide 6 Differences between GIS and cartographic data Processing cartographic data is different from processing GIS data Representations of data for graphic display Symbols Labels Geoprocessing for cartography not spatial analysis Classification Selection by attribute Selection by size Changing the view on the data as opposed to changing the data in the database Let s start by talking about how GIS data and cartographic data are different. One thing that distinguishes cartographic and GIS data are the types of processes used to manipulate the data and more importantly the purposes of the processes. With GIS data, the objective is often either data management of data analysis. With cartographic data, the processes are used to modify the data so it can be displayed graphically, that is, optimized for symbology and labeling. Sometimes we use the same geoprocessing tools as for analysis, such as classification and selection, but the aim for cartography is to get the best representation of the data for display on a map. In essence, the goal is to modify the view of the data but not necessarily modify the original data in the database.
Slide 7 In other words, we change the representation of the features but not the features themselves.
Slide 8 Changing the representation changes the message For analysis dissolve roads so you can validate connectivity in the network For cartography dissolve roads so you can use a continuous road symbol When we change the representation of the data, we change its meaning. The meaning in a cartographic sense is related to how the features are displayed and understood by the map reader. For example, if we were to try and map roads with a dashed line symbol, we might first want to dissolve the roads to make them fewer features. That way, the dashed line will look more continuous. Alternatively, for GIS analysis, you might dissolve the roads to learn more their network characteristics.
Slide 9 Here we can see how changing the representation of roads allows for a better cartographic display.
Slide 10 Alternate example: DEM Terrain Analysis For analysis hydrologic flow analysis with flow accumulation to create a stream network For cartography construct contours from DEM
Slide 11 So what happens is this: The GIS data is sort of like the wire frame of the data. The cartographic data is modified so it can be presented on a map, complete with symbols and labels, and potentially also geoprocessed in preparation for graphic display.
Slide 12 Digital Landscape Models Cartographic Modifications and Enhancements Digital Cartographic Models Map Production Map / Atlas Products 1 km km DLM resolution DCM_10Million 10M World Wall Map DCM_250K 250K Road Map DCM_100K 100K Recreation Map 25 25 m DLM resolution DCM_50K 50K Wall Map DCM_Campus10K 10K Campus Map Derived DLM 10m DCM_10Ktopo 10K OS Topo Map Series 5 m resolution DL:M DCM_5K 5K OS Base Map Real World Geoprocessing Cartographic Finishing A longstanding issue for cartographers is that national mapping agencies don t have compiled DLMs the data we work with is compiled from paper maps. There are a couple of impacts from this: the data we use is already tied to a particular map product, and cartographic enhancements and modifications are so deeply embedded in the database that it shard to distinguish the rules for collection from the specifications for mapping.
Slide 13 DLM - DCM Digital Landscape Model or DLM GIS database (wire frame model) compiled from source information registered to the ground several standardized scales accuracy and resolution are of primary concern Digital Cartographic Model or DCM stored in a product database (for a map) enhanced and modified from DLM specific to a map s purpose graphic clarity and visual integrity are of primary concern To help distinguish between these two types of data, we sometimes use the terms digital landscape model and digital cartographic model. The digital landscape model is that wire frame representation of the data. It is the GIS data that was complied from a source that can be referenced to the ground. It may come in several standardized scales that relate to the requirements of the organization or agency collecting the data. For the GIS data, positional and attribute accuracy are of primary concern and adequate resolution of the data is required for its various applications. The digital cartographic model is a product database that is derived for a specific map product. It contains often a subset of data from the digital landscape model that has been enhanced or modified for the map s purpose. Because the data are used for mapping, the primary concern for digital cartographic data is graphic clarity and visual integrity.
Slide 14 Types of maps High Number of map features (and the subsequent complexity of relationships among them) Special Use Maps Trail Map Bike Map Analysis Results Suitability Map Crime Hot Spots Reference Maps Topo Map Atlas Reference Map Thematic Maps Census Data Atlas Thematic Map Low Low Number of users and uses (indicates the level of general application) High Introduce the concept of lots of maps that can come from one DLM.
Slide 15 GIS database Source -> DLM -> DCM Real world example: Orthophotos, satellite imagery, LIDAR and DEM data are examples of the data that would be in the DLM these is used as source data to create the base polygon, vector and line feature datasets. Different types of maps could be created from the DLM data this results in DCM data. The GIS database can contain all of this information.
Slide 16 Multiple DCM products Terrain analysis Traditional topographic views
Slide 17 Multiple DCM products Town planning Infrastructure development
Slide 18 Multiple DCM products Land cover analysis Site suitability
Slide 19 Summarizing Part 1 Processing cartographic data is different from processing GIS data The intent is to change the view on the data as opposed to change the data in the database Changing the representation changes the message DLM = GIS database; DCM = map product database Lots of maps and their DCMs can come from one DLM
Slide 20 Digital Landscape Models Cartographic Modifications and Enhancements Digital Cartographic Models Map Production Map / Atlas Products Derived 50 DLM 50 m Rules Rules DCM_10Ktopo Model Generalization Derived 10 DLM 10 m Rules DCM_Campus10K DCM_10Kweb Selection Rules Generalization Restructuring the data 3 rd rd Party Datasets DLM Archive DB DLM Archive DB Rules DDMS Landline DB Real World Object Generalization Source data Rules MasterMap Ordnance Survey Great Britain example.
Slide 21 Part 2a Describing cartographic workflows 10 MINUTES
Slide 22 Describing cartographic workflows The workflows are intended to transform information DLM -> DCM -> final map product As you go through the transformations, you: move from multi-purpose to targeted use may move across limited scale ranges Design the workflow and model the data before you go into production
Slide 23 Information transformations in cartography DATA COLLECTION CARTOGRAPHY MAP USE Transform 1 Transform 2 Transform 3 Geographical environment Geospatial data Map Map image Census Ground survey GPS Remote sensing Digitizing Selection Generalization Symbolization Labeling Reading Analysis Interpretation
Slide 24 Cartographic data modeling for map production Geographical environment Conceptual data model ArcGIS geodatabase 250K Road Map 10K Campus Map GIS data model + Workflow Map products 10K OS Topo Map Series 5K OS Base Map Map production process
Slide 25 DLM DCM data production workflow Small R4 R8 R7 DLM DCM Derived data model Map scale R5 R6 R3 R2 R1 Large Small Number of users Large
Slide 26 Map customer and map purpose Graphic ideation Graphic plan Map conceptualization process Graphic refinement REFINEMENT Map conceptualization Map Publication Final product Workflow Map production process Cartographic data model ArcGIS geodatabase Conceptual data model Geographical environment In summary, we can see that it is now possible to see how cartographic data modeling and workflows fit together with map conceptualization. Map making is not a linear process, and simply stringing together the decisions made to conceptualize the map and model the data does not usually result in a product that is satisfactory to the cartographer the first time through. Once the results of the various graphic decisions can be seen simultaneously, multiple iterations of the map are often made in an effort to refine the graphic display. The map maker may choose to refine the initial concept (in which case the loop cycles back through the map conceptualization process) or the map symbolization (in which case the loop cycles through the map production but not the conceptualization process). With this more complete picture of fluid process of map making, it is now possible to apply this understanding in the automation of map production. It is can then be determined to what extent the information model for maps supports map making.
Slide 27 Exercise Discuss how mapping is integrated into your GIS production workflow Map product specifications GIS data or cartographic data? To what extent is your cartographic workflow modeled? By whom?
Slide 28 Part 2b Modeling GIS data to support cartographic workflows 25 MINUTES
Slide 29 Modeling GIS data to support cartographic workflows What does this all look like in an ArcGIS environment? What themes are relevant for the map purpose? What cartographic features? What cartographic attributes? How do you modify the DLM to create the DCM? We talked about GIS data to cartographic data This is where the cartographic geoprocessing happens
Slide 30
Slide 31 Cartographic attributes Any Feature Class Geom ID Attributes... Type 0 0 1 2 3 4 1 2 4 2 Coded Value Domain 0 1 2 3 Description... Prison Description... Park Description... Cemetery Description... Mall 4 Description... Strip Mall Each Type of Cartographic Feature has a number and a description in the Coded Value domain. Integer value in the feature class is faster to display. In an ArcGIS environment -- how do you enrich the data? Through attribute enhancement. This leads to the development of Coded Value domains. At this point, you can take a look at the example domains at the end of this handout.
Slide 32 Symbols Labels In a relational environment, you do this through table joins and relationship classes. For representation, you would attach symbols and labels to the features through these types of joins and relationships.
Slide 33 Example: Cultural CFTID CFTID CFTID CFTID CFTID CFTID Example from one particular theme Cultural features.
Slide 34 Cultural Cultural Features (Buildings, Areas, Points of Interest) Type (how to symbolize and for some kinds of cultural features, it is the label) Name (for label) Size or Area (used to set threshold for which features are shown) You may have the attributes you need in your database through a relationship to a table but for mapping, you need to pull them directly into the DCM so they draw more quickly. For example, export the DLM data to a new feature class delete the fields you don t need, add the fields you do need and populate them, make sure they are in the format that you need them in (e.g., concatenate the fields if you need to).
Slide 35 Boundaries Boundary Lines (any level) Type (how to symbolize) L_Name (left/upper label) R_Name (right/lower label) Annotation must be produced to ensure proper identification Legal Areas (any type) Type (how to symbolize and what kind of label) Name (for label) WithinMSA (Boolean for drawing/labeling) City Points Name (how to label) Size (how to symbolize and label) Special Type (how to label or symbolize) WithinMSA (Boolean for drawing/labeling)
Slide 36 Transportation Roads Class (how to symbolize) Type (determine correct kind of label) Name (for text labels) Highway Route Number (for highway labels) UrbanRural (for multipurpose data which may be represented differently within urban areas) Label String (store exactly what is to be labeled and then the Name does not need to be edited/compromised) Railroads Type (how to symbolize) Owner (how to label)
Slide 37 Hydrography Area Features Name (how to label) Type (how to symbolize and which features to label, and sometimes is the label) Area (used to set threshold for which features are shown) Line Features Name (how to label) Type (how to symbolize) Size, Discharge, or Level (how to symbolize or as a filter for which features to show dendritic branch count and Stream Order are not useful here)
Slide 38 Surface Overlays Land cover/land use/zoning/flood zones/homeland security (risk), etc. Type (how to symbolize) Abbreviated Type (for labeling, e.g., soils, surface geology, etc.) Area (used to set threshold for which features are shown or get labels) Uncertainty (used to modify symbology of either features or labels to show reliability, etc.)
Slide 39 Terrain Contours Elevation (for labels) Units (for labels or conversion when mix of data is present) Type (for symbol, includes index, intermediate, depression, etc.) LabelYN (optional: for determining whether to draw a gap or a line, gaps are where labels get placed) Spot Elevations Type (how to symbolize) Elevation (part of label) Units (part of label)
Slide 40 DEM Elevation Raster Elevation (how to symbolize) Hillshade Raster Value (how to symbolize)
Slide 41 Summarizing Part 2 The workflows are intended to transform information DLM -> DCM -> final map product As you go through the transformations, you: move from multi-purpose to targeted use may move across limited scale ranges Design the workflow and model the data before you go into production Modeling GIS data to support cartographic workflows What does this all look like in an ArcGIS environment? What themes are relevant for the map purpose? What cartographic features? What cartographic attributes? How do you modify the DLM to create the DCM? We talked about GIS data to cartographic data This is where the cartographic geoprocessing happens What are cartographic attributes for each of the feature classes?
Slide 42 Exercise 2:45-3:00 Scenario specify the DCM requirements for a bike map One of these maps tourists, participants, organizers Each group describe all themes Compile the answers together
Slide 43 Break 3:00-3:30
Slide 44 Part 3 Cartographic data model 20 MINUTES
Slide 45 Cartographic Data Model Cartographic Feature Types Symbols and Labels Cartography Table Links the feature to its map representation (symbol + label) Construct the Basemap Data Model together Let s construct the Cartography Data Model together. The Cartography Table joins to the DCM feature classes that will be on the map.
Slide 46 Cartographic Feature Type A type of feature on the map that has unique symbology and/or labeling
Slide 47 Cartographic Feature Type ID (CFTID) in the Cartography Table Each cartographic feature type has a Unique ID Unique description The cartographic feature type is a database feature that is used in cartographic abstraction to communicate the position and characteristics of geographic features and their relationships. The special thing about these features is that they are the features on the map which may or may not be the same as the features you use for spatial data inventory or analysis. For example, on a topographic map, you will often see the hypsography symbolized with contour lines. The rules for this representation of elevation -- that is, color and width of the lines and font size and style of the labels -- are stored in the Cartography Table. On the other hand, you might be interested in using hypsography for analysis, say in watershed analysis -- in this case, you would likely use the digital model of the terrain the DEM -- instead of the cartographic representation. The Cartographic Feature Types in the table are unique and each cartographic feature type is identified by a CFT_ID. Associated with each feature type is a unique description. These CFTs and their descriptions encompass what is sometimes called a complex domain that is, it contains the definition of both features and their valid attributes. This table provides the content for the coded value domain that can be used with the geodatabase and the table is also separated into domains for each feature types
so that there is no duplication and there is completeness, so the table is normalized overall.
Slide 48 Cartographic Feature Types (CFTs) in the Cartography Table Feature class Feature geometry Cartographic Feature Type (CFT) ID & description Each row defines one CFT it has a feature geometry, it has a unique description and you can see what would be shown on the map.
Slide 49 Label specifications Contain the rules for labeling a specific Cartographic Feature Type
Slide 50 Label specs in the Cartography Table Feature class Feature geometry Cartographic Feature Type (CFT) ID & description Label text Label specifications
Slide 51 Symbol specifications Contain the rules for symbolizing a specific Cartographic Feature Type
Slide 52 Symbols Contained in a Microsoft Access database Named for each Cartographic Feature Type (CFT)
Slide 53 Symbol Specs in the Cartography Table Feature class Feature geometry Cartographic Feature Type (CFT) ID & description Symbol specifications Label text Label specs
Slide 54 The completed Cartography Table in the Base Map Data Model
Slide 55 Summarizing the Cartographic Data Model Identify the cartographic features Assign a unique CFTID and description Assign symbology Assign labeling Compile the info into the Cartography Table
Slide 56 Exercise 3:50-4:05 Bike map define the symbology and labeling specifications Define these: Race route Roads Landmarks Facilities (restrooms, restaurants, toilets, First Aid, etc ) Terrain
Slide 57 Part 3b Supporting multi-scale multi-purpose mapping 10 MINUTES
Slide 58 Supporting multi-purpose multi-scale mapping Cartography Table content can be extended across scales and across purposes Add rows and columns to Cartography Table Rows -- Add new cartographic feature types Columns -- Alternate symbology and labeling multi-scale & multi-purpose SQL query strings Definition queries for filtering Label strings for labeling We have the Cartography Data Model and now we want to extend this to support multiscale multi-purpose mapping. This is simple. Simply add new columns (for symbol and label changes) and rows (for new features on the map). And you may need to create new styles.
Slide 59 Scales Labels/ Anno Features Notes First conceptualize what you want on the map at the various scales. The point is to conceptualize before you start going through the process. First figure out what gets added at what scales, then what the representation is pt, line, poly, text. You do this outside of a database environment this is a conceptualization exercise.
Slide 60 Surface Cover And you do this for all the map layers that you need for your map.
Slide 61 Other specifications in the Cartography Table Definition Queries This selects the features to be shown on the map at a particular scale SQL Query Strings This defines which features are being labeled Label Expressions This defines how a label looks and which attributes are used to construct it Scale Range This specifies the scale at which the features will be shown We are essentially separating design from production hopefully leads to less individual refinement. Much of this work goes on outside the ArcGIS environment with a view towards the ability to implement it within ArcGIS at some point.
Slide 62 Part 4 Cartographic workflows in practice 15 MINUTES
Slide 63 Example Here s what you selected Here s how you want to symbolize them Here s the Cartography Table looks like Here s what the CFTIDs look like Here s what the style and the spec in the CT looks like Here s what the text and the spec in the CT looks like Here s the final map This will now allow you to move from the original base map data model to be able to create multiple maps with multiple scales and purposes.
Slide 64 Exercise Discussion Q&A Exit poll What parts of this process might be the most problematic in your organization? What parts do you think might be most challenging for noninstitutional reasons? Go through various examples Walk through the lineage of DLM to some carto feature types (contours, interstate hwy from the Census database from CFCC add name, extract some features using calculate statement)
Slide 65 CONCLUSION 5 MINUTES
Slide 66 Parting words Have a plan -> start from a product Incorporate mapping into your current workflow Data compilation / analysis / production Work with cartographers / project manager / DBAs
Slide 67 Resources Existing resources Slides Exercises Through the ESRI web site - Base map data model web page Cartographic Data Model Cartography Tables Geoprocessing tools, models, scripts also on Geoprocessing web page Case studies Best cartographic practices User community Case studies / best cartographic practices Geoprocessing tools, models, scripts Model & schema
Slide 68 Other sessions For data modeling: Cartographic data model session Data Modeling for Large-scale Maps ands Map Production Wed 7/27 1:30-3:00 Other data model sessions Address data model for names management -- Wed 7/27 10:30-12:00 ArcHydro data model for understanding a data model and seeing how it can be consumed cartographically For map production: Cartography technical sessions PLTS sessions Production Line Tool Set
Slide 69 http://support.esri.com > Downloads > Data Models http://support.esri.com > Downloads > Geoprocessing