DMT 2008
The following was presented at DMT’08
(May 18-21, 2008).
The contents are provisional and will be
superseded by a paper in the
DMT’08 Proceedings.
See also earlier Proceedings (1997-2007)
http://ngmdb.usgs.gov/info/dmt/
Rescuing Legacy Digital Data:
Maps stored in Adobe
Illustrator™ format
Andrew L. Wunderlich
Robert D. Hatcher Jr.
Tectonics & Structural Geology Research Group
Department of Earth and Planetary Sciences
and Science Alliance Center of Excellence
University of Tennessee, Knoxville
Introduction
As GIS databases become the standard for
storing spatial data, many organizations may
be struggling to integrate “legacy” digital data
into modern geospatial databases.
Map information that is stored in older digital
data formats without spatial reference or
attribution are in danger of being lost for
future mapping and analysis purposes if the
data are not converted into newer digital
geospatial databases.
What are “legacy” data?
• Digital file formats that do not natively accept
storage of geospatial information and feature
attributes (Adobe Illustrator files)
• Other older, possibly obsolete digital file
formats (programs have been discontinued,
superseded by another format, etc.)
Also:
• Attribute data in spreadsheet files, field
books, etc…
• Analog map formats such as mylar or paper
Our Problem…
• Large collection of legacy geologic maps
covering over 100 7.5-minute quadrangles.
• Maps are accompanied by a wealth of
ancillary information that is not easily
accessed. (station maps, field books
containing detailed structural data and notes,
photographs, etc.)
• Over 40 years of detailed geologic mapping
(24K or larger scale) in the southern
Appalachians is in danger of slipping into
obsolescence...
Our Goal…
Transform and integrate these legacy data into
geospatial databases to:
– Prepare current and future analysis and
mapping projects (single quad maps,
compilations, cross sections, etc.)
– Enhance the chances for long-term
survival of the data
– Make the data available to the larger
geologic community
Our Response…
Develop a method to convert these legacy digital data
from Adobe Illustrator (and others) to a geospatial
database that:
– Makes the process modular, clearly defined, and
repeatable
– Retains as many of the existing attributes in the
legacy files as possible
– Saves time on repetitive tasks by automating
certain procedures with customized software
tools
– Uses a template geodatabase for better
interoperability with other geologic databases
Our Method…
1) The original file is cleaned up and organized into
layers in Adobe Illustrator and exported to
AutoCAD™ drawings. An image of the map is also
exported to be used for attribution.
2) AutoCAD files are converted to shapefiles and
spatially adjusted in ArcMap™ and appended into
a geodatabase.
3) Geologic point data attributes that were not
retained directly are either calculated from the
feature (strike/rotation) or added to the features
semi-automatically with the help of ArcMap utilities
developed in-house (dip/plunge).
Our Method…
4) The entire database is then checked for
topology errors, and if the map being
processed is adjacent to existing data, the
adjoining edges are reconciled.
5) A round of quality control measures are
taken, including correcting mistaken
attribution, after which any special
considerations are addressed.
6) The finished database is then symbolized,
labeled in ArcMap, and exported as a
graphic for placement in a final map layout
for editing and publication.
Method diagram
Step 1
• The original file is cleaned up and
organized into layers according to feature
type in Adobe Illustrator in order to
maintain maximum attribution.
• Line and Polygon features are densified
and straightened to remove Bezier curves
• Each topological type is individually
exported to an AutoCAD™ drawing.
• A TIFF image of the map is also exported
to be used for attribution of dip/plunge in
Step 3.
Polygons layered by rock unit
Illustrator file layering of polygons.
Faults and contacts layered by type
Illustrator file layering of lines.
Symbology layered by type
Illustrator file layering of lines.
Add anchor points to densify features Simplify features to remove Beziers
Densify and simplify
Export selections to AutoCAD DWG Export entire map to TIFF
Export to TIFF and export to AutoCAD
Step 2
• AutoCAD files are checked in ArcCatalog
for errors. (If found, back to Step 1)
• Selected feature classes within each
AutoCAD file are converted to shapefiles.
• The shapefiles are spatially adjusted in an
ArcMap edit session using the quadrangle
boundary as a guide
• The adjusted shapefiles are saved and
appended to their corresponding feature
classes in a geodatabase
Check AutoCAD
exports for content
and attributes
Preview of AutoCAD classes in ArcCatalog. Check exports for content and attributes.
Export appropriate
CAD feature class
from each drawing
to shapefile
AutoCAD class tree and export to shape dialog
Exported
shapefiles are
loaded into
ArcMap and
spatially
adjusted by
snapping the
corners of the
footprint to the
corners of a
projected
quadrangle
boundary
(HW_Footprint).
Note the small
residual error
(in meters)
Spatial adjustment in ArcMap
Use Append tool to
load data from
adjusted line and
polygon shape-
files to the GDB.
Use the Field Map
to match shape
attributes to GDB
attributes.
Template GDB
with basic
classes to store
geologic data
from legacy maps
ArcCatalog toolbox and Append to GeoDB showing tree
Step 3
• Symbology shapefile is processed in
ArcMap and any unnecessary features are
flagged as unneeded (operator intensive).
• Strike/rotation attribute is calculated for
each feature using an ArcMap Field
Calculator expression.
• Dip/plunge attribute (if applicable) is read
from reference TIFF image and added to
each feature semi-automatically by an
operator with the help of an ArcMap utility
developed in-house (AttributeFeatures tool)
Symbology shapefile needs to be processed to flag
unnecessary features so they are not appended to
database
Use the ‘Length’
attribute to select
features that are
not needed.
Symbology cleanup in ArcMap 1
Don’t delete anything! Use attributes to flag features. Mistakes can be fixed later…
Use contrasting
colors to help
check for errors
more quickly
Symbology cleanup in ArcMap 2
0°
P2
60°
23
P2
P1
300°
Calculated Quadrant Azimuthal
azimuth notation: notation:
N 60 W 300
23 NE 23
With “Keep=1” features
selected (green), use the
Field Calculator on the
‘Strike’ field to determine
the azimuthal rotation
(see the azimuth example
upper-right).
This number is used to
rotate the symbol later
when it is point-based.
ArcMap Field Calculator for strike/rotation
Once symbology shapefile has been processed, it
can be converted to points and appended to the
‘GeologicPoints’ class in the GDB.
Some points will
need to be
snapped to their
station locations
Symbology convert to points in ArcMap
Use the custom tool
‘AttributeFeatures’ to
quickly populate the ‘Dip’
attribute for the selected
features from the image in
the background.
User types in the number,
hits Enter, and the tool
automatically moves to
the next feature.
Attribute Features tool
After attribution, the points are symbolized and the rotation
checked. Note the inversion of the inclined bedding and
overturned antiform symbols. These will have to be corrected.
Symbolize and check for errors
Steps 4, 5, and 6
• The database is checked for topology
errors and edge-matched to existing data
if necessary.
• A round of quality control measures are
taken and any special considerations are
addressed.
• The database is ready for symbolization
and labeling in ArcMap, and can be
exported for placement in a final map
layout for editing and publication.
Use topology to maintain relationships Edgematch adjoining quadrangles
Topology rules/Edgematching
QC and Special Items
• Quality control can consist of:
– Checking tables for <null> or incorrect attributes
and fixing if found.
– Comparing to reference map or checking field
books to confirm correct attributes.
• Special cartographic or other items:
– Labels are converted to annotation so they can
be positioned by hand.
– Cross-section lines, label points, fault names,
fault teeth, etc. are created and stored.
– Any other items specific to a particular map.
QC/Cartographic features added
Map symbology and export
• The data are loaded into a template
ArcMap layout and the layers are arranged
appropriately.
• Layer files allow for implementation of
standards and provide consistent
replication of colors and symbols.
• DRGs, hillshades, or other data can be
added to the map.
• Exports to PDF, Illustrator, or TIFF for
placement in a layout for print or editing.
Map symbology labeling etc
Final map and labels in ArcMap
Finished poster
Final Thoughts
• Modular aspect makes it easier for multiple
operators to work simultaneously.
• Method is less labor intensive than scratch
digitizing, but requires greater care by
operator not to lose or change data.
• Process can be tweaked to work with other
input formats and different outputs (SDE,
ArcIMS, etc.)
Future Developments
• Find a way to link CAD annotation to
features to speed attribution.
• Could be automated further if AI files could
be dissected programmatically.
• Other suggestions?
Thank you for your time
and attention!
Questions?