Category : Science and Education
Archive   : WDBDISK1.ZIP
Filename : MWDBII.DOC

Micro World Data Bank II (MWDB-II)

Introduction
============

The full WDB-II is a digitial map data base produced by the
Central Intelligence Agency (CIA) and distributed by the National
Technical Information Service (NTIS), U.S. Department of Commerce,
5285 Port Royal Road, Springfield, VA, 22161. Micro WDB-II is a
highly compressed version which is suitable for use on micro
computers and was put in this configuration by Micro Doc. Before
describing Micro WDB-II a few words should be said about the source
file - WDB-II. The following paragraphs are paraphrased from the
NTIS overview of WDB-II.

WDB-II is a digitial representation of the world
coastlines and boundries suitable for use in automated mapping
systems. It contains approximately six million discrete
geographic points and was digitized using all available sources
of information. Map scales used range from 1:750,000 to
1:4,000,000 with a nominal scale of 1:3,000,000. These points
are grouped by and identified as describing (1) coast lines,
(2) country boundries, (3) state boundries (USA only), (4)
islands, (5) lakes, and (6) rivers. Each of these groupings is
further broken down into features and subordinate
classifications/ranks. These ranks are hierarchically
structured, and are also used for plotting symbol definition.

WDB-II, as provided by NTIS, is in a 20 character format
on five 9 track EBCDIC one-half inch magnetic tapes. This data
base consists of two types of records, one for the line segment
identifier data, and the other for the latitude and longitude
values of each discrete point making up the line segment. In
this format latitude and longitude values are recorded only as
integers in degrees, minutes, and seconds. WDB-II is available
for $660.00 (Order Number PB-271 874 SET/HBG).

Clearly, WDB-II is an excellent data source when making large (4
x 6 foot) plots on a mainframe or minicomputer. It is, however,
somewhat large (150 - 200 megabytes) for use on microcomputers. Even
on the larger commercial online graphics systems, many points have to
be filtered out before generating displays. For this reason many
people have spent considerable time over the past to filter and
compress this data into a form which could be used in desktop
computers. To these people we wish to express our appreciation -
particularly Antonio Riveria who provided us with a copy of the
latest download of the file.


Background
==========

The present version began as a three megabyte ASCII text file
which contained some 179,000 points selected from all six of the line
types described above. This file was then converted into a sixteen
bit integer format which reduced the size to just over one megabyte.
Since this was still a little large for most five inch disk formats
the file was divided into six files - one for each of the six line
types. The coast line file was further divided into two files as it
was over 400 KB. With this processing completed the file was
configured such that it could be readily moved to most desk top
microcomputers using the MSDOS disk format. The following table
provides a brief statistical overview of the six line types which are
contained in the data base.


Micro WDB-II File Composition
=============================

Detail COAST COUNTRY STATE ISLAND LAKE RIVER Total
Level .PNT .PNT .PNT .PNT .PNT .PNT

1 29150 8573 798 13781 5886 11146 69339
2 35077 10172 959 16077 7080 13217 82599
3 5470 1496 153 2524 1023 1817 12526
4 4000 1249 101 1945 820 1254 9502
5 1478 869 248 844 309 760 5365
Total 75175 22359 2259 35171 15118 28194 179331

Lines 208 301 111 344 103 196 1263

Size (KB) 449 134 14 211 90 169 1067

The above table shows the number of points which are in each
file and for each level of detail. It also shows the number of line
segments in each file and the file sizes in kilobytes.

Detail level 5 contains the fewest points and produces the least
detailed graphics image. The points at each level of detail are
additive to the points at all lower levels. For example, when using
detail level 4 the points from both levels 4 and 5 must be used /
retrieved. Therefore, if one were drawing coast lines at detail
level 4, a total of 5478 (4000 + 1478) points would be processed.
These 5487 points will describe 208 line segments of various lengths.

Note that the same number of line segments exist at all levels
of detail. For example, regardless of which level of detail is
selected, there will always be 208 line segments described by the
coast line data file. In some cases, mostly in the islands file,
line segments degenerate to single points at the lower levels of
detail. Therefore, programs which process this data must check to
see whether each line segment header record is followed by a point
record or another line segment record. More will be said about this
later.

Actual use of the data in these files has shown that most
displays which cover a reasonably large area do not need all of the
detail provided at level 1. A large area would be a major portion of
the U.S. For large area plots level 3 or 4 is normally sufficient
and greatly reduces the number of points which must be processed. To
support generation of large area displays, four composit files were
created. These are WDB3.ALL, WDB4.ALL, WDB5.ALL, and WDB5.CCI. The
first three contain all line types for the indicated detail level
(WDB3.ALL is level 3, etc.). WDB5.CCI contains only country, coast
line, and island line segments and is intended for 'quick and dirty'
testing work.

Micro WDB-II Composite Files
============================

Detail WDB3 WDB4 WDB5 WDB5
Level .ALL .ALL .ALL .CCI

3 12480 0 0 0
4 9365 9502 0 0
5 4315 5365 4315 3664
Total 26160 14867 4315 3664

Lines 1263 1263 1263 853

Size (KB) 157 82 25 17



Micro WDB-II File and Record Structures
=======================================

All Micro WDB-II files are sequential files composed of fixed
length six byte records which can be accessed randomly. Each record
is composed of three 16-bit signed integer fields which can be read
by all microcomputer languages. There are no record seperators used
in these files. The record layout is :

| Code | Latitude | Longitude |

In Pascal, the record definition is :

POINT = RECORD
CODE : INTEGER;
LATITUDE : INTEGER;
LONGITUDE : INTEGER;
END;

Within the MWDBII files the records are stored as shown below.
In some cases there may be hundreds of points in a single line
segment. In other cases there may be consecutive header records
which locate individual points to be plotted.

|Header|Point1|Point2|PointN|Header|Point1|Point2|PointN|Header
|Point1|Point2|Header|Header|Header|Point1|Point2|PointN|


The same format is used for both the line header records and
point records. The usage of each field is as follows :


Field Definitions
=================

CODE - Line Header Records.

The CODE field is used to :

1. Identify the type of line segment which is described by
the point records which follow it,

2. Identify the line or point within the line type, and

3. Provide the coordinates if the record describes a point.

In header records the CODE field is always a positive four digit
number in the form of : ABBB where

A = BBB = CODE =
1 = Coast lines 001 -> 208 1001 -> 1208
2 = Country boundries 001 -> 301 2001 -> 2301
4 = State boundries 001 -> 111 4001 -> 4111
5 = Islands 001 -> 344 5001 -> 5344
6 = Lakes 001 -> 103 6001 -> 6103
7 = Rivers 001 -> 196 7001 -> 7196


CODE - Point records.

In point records the CODE field contains the detail level. The
detail level is a value in the range of 1 through 5. To properly
retrieve points at any detail level use the following comparison :

IF CODE >= DESIRED_DETAIL_LEVEL THEN ...


LATITUDE and LONGITUDE - In both header and point records.

The latitude and longitude fields contain the coordinate values
expressed in signed minutes. The values in these fields are computed
with the formula COORD_VALUE := (Degrees * 60) + Minutes. Thus, the
ranges of coordinate values in MWDBII are :

90 00N = 5400
90 00S = -5400
180 00E = 10800
180 00W = -10800

MWDBII uses the standard four quadrant cartesian coordinate
system in which north, south, east, and west are referenced to the
equator and the Greenwich meridian.

When a header record is followed by point records the
coordinates in the header record are the beginning of the line. When
a header record is followed by another header record then the
coordinates in the header record should be plotted as a single point.
When working at low levels of detail it is common to find multiple
consecutive header records which locate isolated points. There may
also be gaps between some of the lines.

This documentation and the associated files are placed in the
public domain and may incorporated into other products without fee so
long as appropriate credits are included. An appropriate credit line
would be - "This product contains/uses data and/or code placed in the

public domain by Fred Pospeschil and Antonio Riveria. Original
coordinate data was created by the Central Intelligence Agency."

All programs were compiled with MS Pascal Version 3.20.
Graphics support was from Flexi-Graph, New Orleans General Data
Services, 7230 Chadbourne Drive, New Orleans, LA 70126 and Micro Doc,
3108 Jackson St., Bellevue, NE 68005.

ALL USERS OF THESE MATERIALS ARE TOTALLY RESPONSIBLE FOR THEIR
USE AS THESE MATERIALS ARE PROVIDED WITHOUT EITHER EXPRESS OR IMPLIED
WARRANTIES OF ANY KIND.