NACIS, October 2007

ScaleMaster characterization of types of design change through scale for mapping

from multi-resolution databases

Cynthia A. Brewer, Pennsylvania State University

Jessica Acosta, Penn State ESRI

Barbara P. Buttenfield, University of Colorado—Boulder

Charlie Frye and Aileen Buckley, ESRI—Redlands

a.k.a. Cindy, Jess, babs, Charlie, Aileen

NACIS, October 2007

ScaleMaster

Understand workload balance between– display change (select, eliminate, symbolize) – geometry change (simplify, aggregate, displace, collapse…)

Complement to European MRDB research– reviewed in Brewer & Buttenfield (2007) in CaGIS

NACIS, October 2007

ScaleMaster

Understand workload balance– using databases with different resolutions– through a continuous range of scales – for varied map purposes (topographic, zoning, soils, population)

Return to diagram details later…

NACIS, October 2007

Project context

Ada County government (Boise Idaho) ESRI– clean an existing 1:5,000 base map database– topographic and special purpose map products– mapping scales 5K and smaller

1:5K and 1:1M map segments

NACIS, October 2007

Project context

Work that happens at CU-Boulder– geodatabases for ScaleMaster

• edit, clean and attribute Ada County data • create smaller scale geodatabases (Python, ModelBuilder)

akin to LoDs (level-of-detail preprocessed databases)

See Buttenfield & Frye paper in ICC2007 Proceedings

NACIS, October 2007

Map examples at 1:5K, 15K, 30K 100K, 200K, 700K

NACIS, October 2007

Example topo map at 1:30K from 1:5K data

RA Jess Acosta’s map sets:

Designed for onscreen display

Prepared using ArcGIS 9.2 and

Maplex (no convert to annotation)

NACIS, October 2007

Different map purposes with same base data

TopoZoning

1:30K maps from 1:5K data

NACIS, October 2007

Different map purposes with same base data

TopoZoning

Soils 1:30K maps from 1:5K data

NACIS, October 2007

Different map purposes with same base data

TopoZoning

SoilsPopulation density

NACIS, October 2007

NACIS, October 2007

ScaleMaster topoHorizontal axis is log of scale1:5K to 1:1M

Feature types

grouped along

vertical axis

(hues)

Low tech…Excel page

NACIS, October 2007

ScaleMaster topo

symbols – wide bars

labels – thinner bars

Bars extend through range that symbols/labels for feature are present

NACIS, October 2007

ScaleMaster topo

features off map

Geometry change marked by lightness change

NACIS, October 2007

NACIS, October 2007

14 types of design decisions for scale change

code descriptions – size change* c – color change*p – pattern change (e.g., dash, crosshatch)*t – transparency change*l – modify label appearance (e.g., bold, italic, character

spacing, leading)i – improve label positions in relation to nearby

features or labels (e.g., overrun, remove duplicates, feature weighting)

o – on/off for aspect of symbol or label (e.g. remove outlines for areas with feature type still present)

* visual variables

NACIS, October 2007

14 types of design decisions for scale change

r – reclassify features by attribute (e.g. fewer categories)

f – filter by threshold on feature attribute (e.g., filter on size to remove small parks)

e – eliminate layer or eliminate by feature type (e.g., eliminate intermittent streams)

a – add layer or add by feature type (e.g., add labels for physiographic features classed as large)

x – change layer order in TOC (e.g., roads moved from beneath transparent area to above)

R – use Representations tools (e.g., set endings of dashes with full pattern)

G – geometry change (e.g., new data set or new layer with generalized features)

NACIS, October 2007

procedure

Jess’ designs:5K, 10K, 12K, 15K, 20K, 25K, 30K, 40K, 50K,60K, 70K, 80K, 90K, 100K, 120K, 140K, 175K,185K, 210K, 250K, 350K, 400K, 500K, 600K, 700K, 800K, 900K, 1M

Design individual mapsworking randomly through scales,starting as new project on each,with same geodatabase,then examine designs in series

Procedure:

NACIS, October 2007

Procedure:

Jess’ designs:5K, 10K, 12K, 15K, 20K, 25K, 30K, 40K, 50K,60K, 70K, 80K, 90K, 100K, 120K, 140K, 175K, 185K, 210K, 250K, 350K, 400K, 500K, 600K,700K, 800K, 900K, 1M

Design individual mapsworking randomly through scales,starting as new project on each,with same geodatabase,then examine designs in series

1st

6th

4th

5th

2nd

3rd

etc.

NACIS, October 2007

Total 122 map projects

NACIS, October 2007

NACIS, October 2007

ScaleBrewer

5K 30K 210K 500K

NACIS, October 2007

NACIS, October 2007

NACIS, October 2007

Four map purposes at 1:150KTopo

Zoning

Soils

Population

NACIS, October 2007

Small foursome

Topo

Zoning

Soils

Population

NACIS, October 2007

Small foursome

Topo

Zoning

Soils

Population

NACIS, October 2007

Small foursome

Topo

Zoning

Soils

Population

NACIS, October 2007

Small foursome

Topo

Zoning

Soils

Population

NACIS, October 2007

histogram

Topo

0

20

40

60

80

100

120

140

5K-10K

11-25

26-50

51-100

101-250

251-500

501-1M

Scale ranges

Num

ber o

f cha

nges

other changes

elimination

size

label position

Characterize frequency of changes for all four map series

NACIS, October 2007

histogram

88142026319Pop'n10413142735120Soils12717242413203Zoning1731310232311201Topo

Total #

501-1M

251-500

101-250

51-100

26-50

11-25

5K-10K

26-35

16-25

10-15

0-9

no mapping% changes

Percent of total changes by scale range

Type ofmap

NACIS, October 2007

ScaleBrewer

5K 30K 210K 500K

Potential as a GIS map design tool…ScaleBrewer?

NACIS, October 2007

ScaleBrewer

5K 30K 210K 500K

Click here to see styled feature set:e.g., roads with line styles by class

Potential as a GIS map design tool…ScaleBrewer?

NACIS, October 2007

ScaleBrewer

5K 30K 210K 500K

Click here to see processing detail:e.g., aggregation tolerances in LoD

Potential as a GIS map design tool…ScaleBrewer?

NACIS, October 2007

ScaleBrewer

5K 30K 210K 500K

Select to deliver: - Table of specs- .style file, .lyr file- OGC SLD spec

Potential as a GIS map design tool…ScaleBrewer?

NACIS, October 2007

Summary

Workload balance between – geometry changes (simplify, displace, collapse,

aggregate…) – display changes (symbolize, select and eliminate)

You can design through an entire scale range with – few data captures– key geometry pairs (e.g., centerlines that correspond to

polygons) – few LoDs for sensitive feature types (e.g., hydrography)

NACIS, October 2007

Next steps

How general is ScaleMaster?

– Examine distributions of types of change through scale• assist design decisions• assist data download choices

– Establish effective workload that minimizes geometry change • reduce layer integration burdens• reduce processing time• different (cheaper) worker skills

NACIS, October 2007

Next steps

How general is ScaleMaster?

– Determine where preprocessed LoDs and new data captures most needed

• reduce workload/cost• assist decision to incorporate others’ existing datasets (are

data close enough to be considered redundant scale?)

NACIS, October 2007

Regular research meetings

Kiyoshi, babs, Charlie, Aileen, Clint, Jess

December 2006 in Redlands

NACIS, October 2007

ESRI PSU CU

March 2007 at Penn State

babs, Jess, Aileen, Charlie

NACIS, October 2007

Student acknowledgements

Research Assistants/Interns, 2005 to 2007:PSU: Jessica Acosta, Mamata Akella, Zachary TardivoCU: Torrin Hultgren, Kiyoshi Yamashita, Anna Harman

Applied Cartographic Design, PSU undergrads, Spring 2006:Steve Cline, Rosie Daley, Emily Dux, Derek Foll, Katie Holmes, Allen Huber, Alex Krishchyunas, Adam Naito, Joe Palchak, Ed Petronsky, Ryan Stoyek, Matt Sudac, Jessica Acosta (grad)

Cartography Seminar, PSU graduate students, Fall 2005:Tanuka Bhowmick, Steve Gardner, Adrienne Gruver, Tania del Mar López Marerro, Sohyun Park, Anthony Robinson, Steve Weaver

NACIS, October 2007

ESRI acknowledgements

Funding from ESRI: Professional Services Agreements 2003B6345, 2003B6346, 2003B6347, and 2006B2964

ScaleMaster Excel file and paper posted on Mappingcenter:blogs.esri.com/Support/blogs/mappingcenter/

archive/2007/08/22/scalemaster.aspx