Classification and Application of Cartographic Animation - UO ...

Page created by Tiffany Pham
 
CONTINUE READING
Classification and Application of Cartographic Animation

Amy Lobben
Central Michigan University
The development of a definition and classification scheme for cartographic animation is now warranted, since
focus on the visualization method increased during the 1990s. This article offers a comprehensive definition of
animation, distinguishing between animation, slide shows, multimedia, and hypermedia. In addition, a
classification system identifies and offers specific characteristics for four different methods or types of
cartographic animation: time, areal, thematic, and process. Key Words: animation, cartography, research tool,
visualization.

Introduction                                                 media and hypermedia, as is the case with some
                                                             areal animations, or ‘‘fly-bys,’’ discussed later.
        ith use of cartographic animation now                Often the distinction between multimedia and
W       widespread, different techniques and
types of animation are being developed. The
                                                             hypermedia is blurred and while these two
                                                             are often used together in a presentation, they
purpose of this article is to identify the different         are different methods, serving different pur-
categories of animation and to briefly discuss               poses (for further discussion of the distinction
their potential usefulness to cartography, as                between these methods, see Krygier et al. 1997).
well as to other areas of geography.
   The focus here is specifically animation,
                                                             Background
which will be defined now for clarification
purposes. An animation may be, in simplest                   Animation is now beginning to become widely
terms, any moving presentation—be it through                 used, not only by cartographers, but also by
film, video, or computer—that shows change                   other geographers. Its popularity as a data-
over time, space, and/or attribute. Animations,              display method has grown since Campbell and
which often run multiple frames per second,                  Egbert in 1990 spoke of the lack of incorpora-
differ from computer ‘‘slide shows,’’ which                  tion of animation into geography. In the thirty
are simply a series of graphics presented con-               years before 1990, the field of geography
currently, often employing transitions, such as              was only speckled with papers, including those
fades or wipes, from frame to frame. The                     by Thrower (1959), who thought that popu-
difference between the two is a matter of time-              lar animation had nearly arrived, and Tobler
scale. In animations, the viewer cannot detect               (1970), who first published the results of a
the point at which one graphic in the series                 computer animation in cartography (Campbell
replaces the previous, while in slide shows, the             and Egbert 1990). Fortunately, since 1990,
viewer can identify the point at which graphics              geography has witnessed the beginning of what
change. In other words, each individual slide-               could finally be the blossoming of animation, as
show graphic frame remains on the screen                     more projects, papers, theses, and dissertations
longer than does each animation graphic frame.               continue to be devoted to the subject.
   An animation often consists of thousands                     Terminology has varied through this evolu-
of slightly different individual frames and may              tion of animation. The terms ‘‘cartographic ani-
include sound (narration, music, or sound                    mation,’’ ‘‘map animation,’’ ‘‘four-dimensional
effects), resulting in a multimedia animation.               cartography,’’ ‘‘dynamic mapping,’’ and ‘‘spatio-
Multimedia animation may be incorporated                     temporal display’’ may all be used to describe
into a hypermedia presentation, which allows                 ‘‘cartographic displays having a succession of
users to explore data through the use of on-                 maps pertaining to the same area whose content
screen buttons (Krygier et al. 1997). In some                changes in relation to the independent vari-
cases, animation may be simultaneously multi-                able—time’’ (Weber and Buttenfield 1993, 141).

     The Professional Geographer, 55(3) 2003, pages 318–328 r Copyright 2003 by Association of American Geographers.
              Initial submission, July 2000; revised submission, November 2002; final acceptance, December 2002.
  Published by Blackwell Publishing, 350 Main Street, Malden, MA 02148, and 9600 Garsington Road, Oxford OX4 2DQ, U.K.
Classification and Application of Cartographic Animation         319
After years of experimenting with the uses of       are distinct, and some are subtle. Usually, the
animation, this method seems to have proven         literature fails to identify an animation as being
valuable in geographic education, specifically in   of a particular type or to place it into a distinct
the development of educational supplements,         category, probably because a comprehensive
including student exercises and instructor aids.    categorization system for animation does not
Educational publishers offer CDs containing         yet exist. This article offers a categorization
geographic applications—which include multi-        scheme for animation and highlights ways in
media, hypermedia, and animation—as supple-         which animation may be used most effectively
ments to textbooks or as independent products.      for representing dynamic data.
  Fortunately, the popular application of ani-         The categorization scheme is based on three
mation, multimedia, and hypermedia in ge-           criteria: time, variable, and space. The dynamic
ographic education is resulting in the continued    or static quality of each criterion provides the
development of design methods as well as of         classification basis. Time is considered to be
technology. As a result, several categories and     dynamic if the dataset includes temporally
uses of animation as data display methods have      varying observations—in other words, if the
been developed and are identified below.            data include observations that change over
                                                    time. Time is considered static if the dataset
                                                    is temporally unchanging—in other words, if
Categories of Animation
                                                    observations are made at a single moment or at
Both the earlier years of design and method         least recorded as a single moment. Variables
experimentation and the current widespread          are considered dynamic if their quality or
development of educational materials have           quantity changes; as discussed further below,
produced a plethora of animations and accom-        these changes may or may not take place over
panying literature. The highest-order classifi-     time. Conversely, variables may be labeled
cation may be based on interactivity, which         static if they and the symbols representing
separates animations into two categories, pres-     them do not change. The space may also be
entation and interactive. Presentation anima-       identified as either static or dynamic, depend-
tions allow the viewer little or no control over    ing on whether the base map itself is held
the progress of the product; the animation is       constant or allowed to change through either
only viewed. If it is seen through a computer,      viewer or animation control.
the viewer may be able to control the speed and        The discussion below will further elaborate
may be able to pause the movement. One              on the characteristics of time, variable, and
advantage of this method is that the graphics       space as well as their static or dynamic nature.
may be viewed in other ways, such as via video      This discussion will allow a reader to identify
or film projector. Interactive animations—also      different types of cartographic animation that
known as hypermedia animations, as men-             may be created based on specific characteristics
tioned above—offer the viewer considerably          of their spatial data. The reader may apply a
more control over the course of the animation.      decision-tree diagram, such as that in Figure 1,
The viewer may direct the animation in several      to assist them in selecting an appropriate
ways—panning, zooming, and rotating—as              animation method. To use the decision tree,
well as engaging in discourse with the computer     the reader must first identify the dynamic or
through inputting data or responding to ques-       static nature of time, variable, and space
tions. The following defined categories do not      characteristics in the dataset they wish to
distinguish based on this factor; they rely on      visualize. The classification scheme is designed
different bases for classification. As a result,    to allow creators and users of animation to
animations falling into one or more of these        choose a specific type of animation that best
categories may be either presentation or inter-     visualizes their unique data. Included in the
active, depending on how the author intends         discussion of each category is a three-axis
the final product to be used.                       graphic, which provides a graphic representa-
   A review of the literature reveals several       tion of the combination of criteria comprising
categories of animation, identified here as time-   the specific category. The three axes in each
series, areal, thematic, and process animation.     diagram represent space, time, and variable.
Some differences between these categories           The shading or absence thereof indicates the
320     Volume 55, Number 3, August 2003

Figure 1 Decision-tree diagram.

static or dynamic nature of each variable. The       Dynamic Time The most important char-
axes do not represent numerical values of            acteristic of a time-series animation may be the
the criteria; instead, they show the potential       depiction of change over time. In this categor-
for change of the criteria. For this reason, the     ization, a variable’s changing location over
diagrams are not drawn as cubes, which may           time is highlighted; the variable appears or dis-
give the visual impression of finite axes. Rather,   appears at varying locations through time.
the open-ended axes of the diagrams included         For example, the geographic extent of a bark-
here indicate the change potential, which is         beetle infestation may be visualized by creating
theoretically infinite.                              dozens of different maps (using the same base
                                                     map), each representing infestation locations
                                                     over twenty years. In this example, data may be
Time-Series Animation                                represented per day, week, month, or year. For
Animating time, ‘‘where the map is held still
and the action played out upon it’’ (Dorling
1992, 218), has been one of the most popular
forms of animation. Change over time normally
may be shown in a time-series animation, in
which time is dynamic (it changes), the subject
or variable remains static (no change in the
variable quality), and the location (base map)
is static. Figure 2 provides a graphic cube
representation of these criteria.

Characteristics of Time-Series Animation
Time-series animations portray data that main-
tain three distinct characteristics: dynamic
time, static variable, and static space. These
characteristics are presented in Figure 2 and
discussed more fully below.                          Figure 2 Time-series animation criteria axes.
Classification and Application of Cartographic Animation         321
time-series animations, the time scale is not         the location of a given phenomenon may change
restricted; change may be measured and visual-        through time, usually the whole map area is
ized over seconds or over centuries.                  held constant: the base map itself does not
                                                      change. Therefore, time-series animations are
Static Variable      For the purposes of the          considered to maintain fixed space. However,
classification scheme presented in this article,      two time-series animations may be included
a variable may be considered static if the quality    together in a single presentation. For example,
of the variable, represented by the variable’s        if the bark-beetle data include precise data
symbol on a map, does not change. For exam-           for a small geographic area, then a large-scale
ple, consider the bark-beetle infestation. If the     animated map could be created, as described
data include locations for specific trees that        above. But, in this example, a single, unchang-
were attacked over the course of many years,          ing base map on which the bark-beetle sym-
the cartographer could symbolize each tree            bols appear and/or disappear is maintained.
infestation with a single symbol. The final           Assume that statewide data that show which
animation could show symbols appearing as             counties are inundated by the bark beetle are
trees are infested, or tree symbols could be          also included. The state-level data could be
shown disappearing from the map as each tree          represented by coloring each county as it
dies from the beetle attack. Similar to a dot-        initially becomes infested. The state map—
density map, the static quality of this variable      a relatively small-scale map—which represents
is represented with a constant (unchanging)           area data (county) with area symbols (colored
symbol. Although attribute symbol quality does        enumeration units) could follow the first large-
not change, the animation may portray data            scale map, which visualizes individual, tree-
pattern or density changes. In other words,           level data (point symbols representing point
density patterns may be visually perceived, even      data). But, again, this presentation would
though the attribute’s symbol remains static.         include two separate time-series animations
   In time-series animations, the existence of        presented consecutively.
a variable is documented through discrete
change at varying locations, rather than differ-      Design Considerations and Applications of
ences in quantity, which could be represented by      Time-Series Animation
changing symbols. If a researcher wishes to           Because a time-series animation visualizes a
illustrate quantity differences, this goal could      given phenomenon changing locations through
be accomplished through the use of different          time—a characteristic intrinsic to the categor-
symbols representing different quantities of an       ization of time series—a main focus of the
attribute at varying locations. These quantity        presentation is time. As a result, time changes
differences can be portrayed by creating a thema-     can ideally be represented as occurring at a
tic animation (discussed below), rather than          constant rate, as much as the original data
through a time-series animation. But, while time-     will allow. For example, in a one-hundred-year
series animation data are static, they may be meas-   timetable, a map (animation frame) may be
ured and represented as point, line, or area data.    created to represent change for each successive
   Finally, while a symbol may be used to             year, instead of several frames showing change
represent an attribute and while that symbol          sporadically (at two-year, then one-year, and
remains static, it does not necessarily have to       then three-year intervals). However, maps are
be qualitative (nominal). A single symbol may         created for a specific purpose, and therefore the
be used to represent a quantitative amount of         intended communication goal will influence
a variable (say, one dot representing five trees      the visualization of the data.
infested by the bark beetle). However, the value         Time-series animations portray the chrono-
of the symbol in a time-series animation will         logical and spatial change of phenomena and
not change, as this would represent dynamic           thus are intrinsically geographic, because they
attribute symbolization.                              provide a method for illustrating dynamic
                                                      geographic data (chronologically changing
Static Space Although a variable’s appear-            spatial data). This method has utility in several
ance or disappearance in geographic areas is          areas of geography; a few specific examples are
depicted, and therefore space as it relates to        presented here. Medical geographers can use
322     Volume 55, Number 3, August 2003

time-series animation to display the appearance
of a disease in specific areas through time (point
or area data). Population geographers may
employ this method to visualize the migra-
tion sources of U.S. immigrants through time
(usually area data, such as countries or pro-
vinces). Military geographers may use this
method to illustrate changing political borders
resulting from military campaigns (line data).
Finally, as in the example discussed above,
biogeographers can visualize disease, infesta-
tion, and animal-migration data.

Areal Animation
                                                     Figure 3 Areal animation criteria axes.
Areal animations ‘‘emphasize the existence of a
phenomenon at a particular location’’ (DiBiase
et al. 1992, 206). Unlike time-series animation,     Characteristics of Areal Animation
where the animation is shown from a constant         Three characteristics may be observed in
viewpoint (static base map), the viewpoint in an     areal animations: static time, static variable,
areal animation is dynamic (the base map is not      and dynamic space. Figure 3 provides a graphic
static), while time remains fixed.                   illustration of these characteristics, which are
    DiBiase and colleagues (1992) identify this      discussed more fully below.
location change as visualizing change through
space. Currently, one of the most popular ways       Static Time An areal animation may be
to achieve an areal animation is by using a ‘‘fly-   considered to be a ‘‘snapshot’’ in time. Time is
by.’’ DiBiase and colleagues (1997, 207) define a    static. Therefore, areal animations as classified
‘‘fly-by’’ as ‘‘a sequence of views of a static      here would not show change over time. Instead,
surface or volume in which the viewpoint of the      if a geographer wishes to visualize terrain or
observer changes gradually.’’ Fly-bys can be         other areal changes over time (continental
either presentation or interactive animations.       drift or creation of mountains, for example),
Using a presentation fly-by animation, the           they may choose to visualize the data through
viewer may observe a rotating earth, for             a process animation, discussed below, which
example, where the only control the viewer           may be used to visualize dynamic time and
has over the course of the animation is in           dynamic space.
starting or stopping the rotation or through
adjusting the speed at which the rotation is         Static Variable    While an areal animation
viewed. Using an interactive fly-by animation,       can—and most likely will—include many varia-
such as one that shows terrain, the viewer can       bles, these variables are fixed (unchanging).
control the geographic position from which the       Because areal animations do not show change
animation is viewed—they can view a mountain         through time, specific observations—and
range, for example, from a northwest position        therefore each unique symbol representing
at a 100-foot altitude and then move to a            these observations—remain static on the base
southeast position at a 1,000-foot altitude.         map. As a result, variables are shown with
    Another common example of an areal anima-        neither discrete nor continuous change. How-
tion is the increasingly common interactive          ever, several unchanging symbols may be
Internet street map. Many of these maps allow        included in the animation. For example, a fly-
the viewer to input a general geographic area,       by of a mountain range could include symbols
such as a city or state, but some allow increased    representing several variables, including ele-
specificity by permitting the input of a specific    vation, biological data, and cultural pheno-
address. Once the map is displayed, the viewer       mena (roads, buildings, and so on). Also, while
can change scale, pan, and sometimes rotate.         the variables are static, areal animations may
Classification and Application of Cartographic Animation         323
represent qualitative or quantitative data. Be-        spatially explore the geographic data and
cause the objective of many landscape fly-bys          identify patterns, similarities, or differences
is to visualize terrain differences, the visual dif-   between areas. For example, imagine a map of
ference between the top of the mountains and           the United States constructed so that three-
the bottom of the valleys, for example, illustrates    dimensional prisms represent the populations
elevation differences, which are quantitative          of each state for a given year, with higher prisms
data. One can also imagine a map in which              symbolizing higher values. The population
states are classified qualitatively according to       would appear as clusters of skyscrapers in
the predominant crop grown (e.g., corn, wheat,         the Northeast and on the West Coast, while
oranges), on which a viewer can pan or change          the comparatively sparsely populated middle of
to larger scale to see the same data differen-         the country would contain only an occasional
tiated by county. Both examples animate data           midrise. The animation viewer would be able to
that are distinguished as qualitative or quan-         ‘‘fly’’ through the United States, gaining a
titative, but both may be classified as areal          dramatic and unique view of U.S. population
animations.                                            distribution.

Dynamic Space The most important char-
                                                       Thematic Animation
acteristic of an areal animation is the illus-
tration of spatial variation. While space (base        In thematic animations, location is held con-
map) remains fixed in a time-series animation,         stant (static space) and time and variables may
constant spatial variation is what distinguishes       change their values or attributes (dynamic time
areal from other types of animation. Specifi-          and variable). The emphasis in a thematic
cally, an areal animation may illustrate chan-         animation is attribute, which is highlighted at
ging spatial viewpoints. Panning, zooming,             a particular location and, over time, differen-
rotating, and angle perspective are some of            tiated by value or nominal classes.
the techniques that may be used to create a               Several methods of thematic mapping can be
dynamic space. A viewer may adjust map scale           applied to visualize spatial data, creating static
through zooming in and out, or they may adjust         thematic maps. It follows that some of these
perspective by entering a specific viewing angle       methods can also be applied to create animated
(above ground level, for example). In addi-            thematic maps. For example, an animation
tion, north-east-south-west viewpoints may be          that uses graduated color symbols (choropleth)
changed through rotation or panning through            or graduated point symbols to illustrate the
the base map.                                          magnitude in difference through time at dif-
                                                       ferent locations would be classified as a the-
Applications of Areal Animation                        matic animation. However, as mentioned in
Currently, most areal animations display land-         the time-series discussion, an animation that
scapes or landforms. Therefore, the seemingly          uses dots (similar to a static dot-density map)
obvious beneficiaries may be geomorpholo-              or unchanging solid-area (choropleth) sym-
gists. However, other geographers will find this       bols to show the appearance and/or disappear-
method useful for displaying spatial relation-         ance of an attribute at different locations
ships or patterns between different locations.         through time would be better classified as
While maintaining fixed time, terrain fly-bys          a time-series animation. Therefore, use of a
usually only show landscape changes, but it is         traditional, static thematic method in an anima-
the inclusion of the other variables that makes        tion does not automatically classify it as a
this method appealing to other geographers as          thematic animation.
well. For example, one may look at socio-                 Graduated symbols or choropleth mapping
economic trends among the states: as location          methods may be used, but animations employ-
varies, differences in a variable such as popula-      ing other thematic methods may belong in
tion may be observed. These changes may be             other categories. For example, an animation
achieved by using color, shading, or prisms            based on a series of dot-density maps would
to represent numeric values or nominal catego-         perhaps be better classified as a time-series
ries as they differ spatially at a single given        animation, since the variable’s attributes are
time. Areal animations allow the viewer to             maintained (one dot size and color), time varies,
324     Volume 55, Number 3, August 2003

location (base map) remains static, and the dots     an areal animation, time is static. In the case of a
appear and disappear at different locations over     thematic animation, time can be either static
time. Choropleth maps may be used in either          or dynamic, depending on the specific nature
time-series or thematic animations, the main         of the animation. Borrowing from Peterson’s
difference being that in a time-series animation,    (1993) classification, animations can be used
the symbols used to represent a variable do not      to compare and contrast different methods
change: they are presented as one class of data      of classification, as well as different levels of
(recall, however, that the class may represent       generalization within a single classification.
either qualitative or quantitative data). A thema-   Datasets can be classified using several different
tic animation, though, will include several          systematic and acceptable methods. However,
categories of data, which may also be qualitative    in most cases, the resulting classification will
or quantitative. These characteristics are dis-      differ by class number, individual class ranges,
cussed in more detail later.                         and class members. Moreover, not only do
   Peterson (1993) identifies four ways in which     different classification methods result in differ-
animation may be used, all of which may be           ent maps, but different calculations using the
classified as thematic animations. Classification    same classification methods can also result in
animation is used to show and compare the            different maps. For example, when using equal
different methods of classification. General-        intervals, one can choose different intervals for
ization animation shows the different levels of      different classifications. Also, when employ-
data generalization as the number of classes         ing the quantile method, quartiles or quintiles
change for the same data on a given map.             would most likely result in different classes and
Geographic trend can be observed when viewing        would then map differently. Different breaks
a shifting of values shown in a choropleth           can be chosen if using the natural-breaks
animation depicting changing land values, for        method, and user preferences abound when
example. Finally, animation is used in comparing     using the arbitrary method. Therefore, one
distributions within a series of maps.               could choose a single dataset (fixed in time and
                                                     space) and complete a myriad of classifications
Characteristics of Thematic Animation                that might all be visualized through a thematic
The matic animations share the following three       animation.
characteristics: static or dynamic time, dynamic
variable, and static space. Figure 4 illustrates     Dynamic Variables        One of the most sig-
the time, variable, and space characteristics that   nificant characteristics of thematic animations
define an animation as thematic.                     is the dynamic variable. The symbol assigned to
                                                     a specific value or class and represented at a
Static or Dynamic Time Recall that in a              given location on the base map may change its
time-series animation, time is dynamic, and in       quality through time. Consider the population
                                                     example discussed above. A single state could
                                                     change hue values several times as the anima-
                                                     tion progresses due to population fluctuations
                                                     over time. Thus, the symbol represented at a
                                                     specific location will physically change. The
                                                     change may be observed as discrete or con-
                                                     tinuous change—for example, a discrete jump
                                                     in an enumeration unit on a choropleth map
                                                     from light blue to dark blue, or the continuous
                                                     growth of a graduated symbol. Contrast this
                                                     characteristic with the static variable utilized in
                                                     a time-series animation. In that case, a symbol
                                                     may discretely appear or disappear in given area
                                                     (presumably as the data in that given location
                                                     changes), but the quality of the symbol itself
                                                     does not change. Although static thematic maps
Figure 4 Thematic animation criteria axes.           may represent quantitative or qualitative data,
Classification and Application of Cartographic Animation            325
thematic animations may represent quantitative        dynamic geographic data, it has the unique ability
data. The dynamic symbols in a thematic               to dramatize and highlight geographic trends and
animation represent differences in magnitude          demonstrate the effects of spatial autocorrelation,
at given locations (point, line, or area). Quali-     a subject of study that is uniquely geographic. A
tative data may be better represented through a       similar animation could highlight the changing
time-series animation.                                influence of a single component of economic
                                                      base, such as the service industry, in several cities
Static Space Like time-series animations,             over time. In this example, a graduated symbol
thematic animations maintain a fixed base map;        could be used to represent the changing percen-
zooming and panning around the base map are           tage of service as part of the economy through a
not characteristics of this method. Therefore,        given period of time.
while data may vary from location to location,
the overall base map itself is static. However,
                                                      Process Animation
as can be accomplished when employing a
time-series animation, two thematic anima-            Process animation takes a unique perspective.
tions with two different base maps can be             Although some animations—even some of
shown consecutively, or even concurrently. For        the examples mentioned above—may not differ
example, a dataset that maintains population          dramatically from a series of static maps, pro-
data per state for each year between 1,600 and        cess animation offers two important attributes
1,999 could be animated, resulting in thematic        that static representations do not: motion and
animations. An animated choropleth map could          trajectory. Motion is defined by Webster’s Desk
be created by systematically classifying the data     Dictionary as ‘‘the action or process of moving
for each year and assigning different values of a     or changing place or position,’’ while trajectory
hue to each class. Because of the settlement          has been defined as ‘‘the direction of travel of a
patterns of the United States, two animations         moving object’’ (Reiber 1991, 318). Therefore,
using two different base maps may be appro-           by creating an animation that emphasizes both
priate. A base map of the eastern seaboard could      motion and trajectory, the viewer will be able to
be used for about the first 200 years (resulting in   visualize a phenomenon that cannot be visua-
a 200-frame animation, for example), followed         lized in static maps, ‘‘the continuous evolution
by another animation in which the base map            of [a] process’’ (Epstein 1990, 54).
includes the entire 50 states (300 additional            Motion and trajectory in animation are used
frames, with one frame representing each year         very effectively in fields in which the viewer
of data). The resulting animations would most         must understand a specific continuous action
likely show states changing hue values as their       or be able to visualize objects as they appear in
individual classification changes. Again, though,     real life. For instance, in engineering, ‘‘where
the resulting product would contain two dif-          . . . rotation or portrayal of three-dimensional
ferent thematic animations shown consecu-             objects are involved and required, the use of
tively—although, depending on the dataset             animated graphics . . . has been found to be
and mapping purpose, two animations could             appropriate and is recommended’’ (Asoodeh
also be shown concurrently, essentially result-       and Clark 1993, 33). For example, an anima-
ing in a split-screen.                                tion illustrating an offshore oil spill would be
                                                      considered a process animation if the animation
Applications of Thematic Animation                    showed the continuous flow of spill, the speed,
Thematic animation would be useful for car-           and the shape changes. This animation would
tographers and other geographers who wish to          show the constant movement and the process
represent temporal and variable changes in a          that the spill experiences from the initial point
particular location, such as the changing pre-        of spillage to when the oil eventually settles or
dominant economic bases of U.S. cities. In this       is removed. A process animation such as the
example, the location is constant (one city); time    above example of the oil spill illustrates at least
varies (chronological change); and the varia-         two dynamic characteristics—time (from mo-
ble changes (varying economic bases such as           ment of spillage through time to the settlement
agriculture, manufacturing, or service). Because      or cleanup) and attribute (the oil is the variable
this method provides a platform for displaying        and is represented by a symbol that constantly
326     Volume 55, Number 3, August 2003

changes size, shape, motion, and trajectory)           of a spatial database (population migration over
—while space may be dynamic or static in               time, for example) in which time data are
this case.                                             recorded for phenomena. Or time may be part
                                                       of a developed theory that is illustrated through
Characteristics of Process Animation                   a process animation. For example, while con-
The previous classifications include a combina-        tinental drift is a generally accepted theory,
tion of static and dynamic characteristics. In         specific data were not recorded at given times as
fact, in identifying the characteristics of the        the process occurred; thus, a compiled dataset
time-series, areal, and thematic classifications,      that includes dynamic time is not represented.
the dynamic nature of one of the characteri-           As a result, process animation would not
stics defines the classification to some extent.       necessarily be employed to visualize the dataset;
For example, with time-series animation, the           rather, it would be employed to visualize a
significant consideration is that time is dy-          theoretical process (in which time is a factor in
namic, while space and attribute are static.           the theory development).
Space is dynamic (while time and attribute are
static) with the areal animation. Attribute is         Dynamic Variable         As discussed above, a
dynamic, space is static, and time is either           variable is dynamic when the symbol designed
dynamic or static (depending on the specific           to represent an attribute changes its quality
dataset or purpose) in a thematic animation.           through time. With process animations, the
With process animation, all three criteria may         variable exhibits continuous (though not neces-
be dynamic, with none of the three more                sarily constant) change. Consider the example of
significant than others in defining this category      the oil spill. The symbol representing the
of animation. Figure 5 illustrates the dynamic         process of the oil spill (the oil moving through
nature of all three characteristics.                   and over the water) changes as the motion
                                                       and trajectory of the spreading oil changes. Or
Dynamic Time         A process animation visua-        imagine spilling a glass of water onto a blue and
lizes processes or concepts occurring and/or           white tile floor. The water may flow over the
changing over time. As a result, according to the      tiles as a single puddle or may separate into
classification structure identified in this article,   different puddles, depending on the obstacles
dynamic time is a characteristic of process            encountered over the course of movement. The
animation. The time may be measured as part            number of variables should not necessarily
                                                       factor into the classification. A single (albeit
                                                       changing and dynamic) variable may be included
                                                       in a process animation, or the symbols of several
                                                       different variables (again dynamic) may repre-
                                                       sent different data. In addition, the variable data
                                                       may be qualitative or quantitative. However, in
                                                       the case of a process animation, the variable does
                                                       not change to represent differences in magni-
                                                       tude at a given location (a growing graduated
                                                       symbol or changing value of a hue), as would
                                                       be seen in a thematic animation. Instead, the
                                                       dynamic nature of the variable shows change,
                                                       spread, or trajectory of a process.

                                                       Dynamic or Static Space In a process ani-
                                                       mation, space may be static or dynamic. The
                                                       base map may be fixed while the action of
                                                       the process is represented on it. Alternatively,
                                                       the viewer may be granted control over zoom-
                                                       ing or panning around the map as the process
                                                       plays out on the base map. The cartographer
Figure 5 Process animation criteria axes.              may also create the animation as a presentation
Classification and Application of Cartographic Animation       327
animation and control the panning, providing
base-map-shifting as the process changes
course or the action intensifies from location
to location.

Construction and Application of Process
Animation
This type of animation, which emphasizes
movement, is very different than previously
discussed applications. Relative to the other
three animation methods, process animations,
which visualize motion and trajectory, may be
considerably more complicated to construct,
even given improving and more user-friendly        Figure 6 Summary chart.
animation software. Constructing this type
of animation often involves combining char-
acteristics of area, variable, and time. Since        An effective animation is a useful animation.
movement is an intrinsic factor in process         This data display offers geographers an alter-
animation, the base map may be constantly          native to traditional display and research
changing, as in an areal animation. In addition,   methods. Although animation is presently used
this areal change is compounded by chro-           by many geographers as a way to enhance still
nological change—which is the key factor in        images, it can be used for so much more than
time-series animation—to produce areal change      merely enhancing these images. ‘‘The appeal-
through time.                                      ing feature of [animation] . . . is not only the
  Geographers can use process animation to         capability of showing surfaces from different
help explain continental drift, stream meander-    viewing points but also the ability to show
ing, or any other spatial phenomena in which       change through time in a way that is more
depicting movement over time is an intrinsic       powerful than the creation of a series of static
part of understanding the phenomena. Process       maps’’ (Moellering 1980, 67).
animations may be considered to represent             Animation may continue to gain popularity
reality more closely or to be more like a film     as a display tool for geographic data, which will
when viewed. Using process animation could         further the development of improved technol-
represent an accurate interpretation of the        ogy and design techniques. Just as geographers
behavior of the attribute; the viewer might be     realized the research potential of geographic
able to visualize the entire spread of a phenom-   information systems and remote sensing, the
enon or a theory or process.                       desire to keep current technical research as
                                                   part of our science will provide the impetus for
                                                   geographers to continue to explore techno-
Conclusion
                                                   logical research methods. This movement may
The classification system discussed above offers   allow researchers to realize the possibilities of
a possible method for determining an appro-        animation and take a leading role of incorpo-
priate animation that may be used to visualize a   rating it as a bona fide scientific research
given dataset or theory. To apply the classi-      method. For example, a geomorphologist may
fication system and to work through the            wish to test their new theory of landform
decision-tree diagram, a geographer should,        development. Since many landforms develop
ideally, possess a clear and complete under-       over the course of thousands or millions of
standing of a dataset’s characteristics. Before    years, it may not be possible to view the entire
implementing the decision-tree diagram, the        theory at work through field observation.
geographer should consider the dynamic and/        Instead, an animation illustrating the theory
or static nature of time, variable, and space in   may allow the geomorphologist to ascertain the
their dataset. The summary chart in Figure 6       theory’s strengths and weaknesses, leading to
provides a succinct view of the data character-    modification and improvement in the geo-
istics of each method.                             morphic theory. While it may still not be
328     Volume 55, Number 3, August 2003

possible to observe the theory in the field,         Campbell, Craig, and Stephen Egbert. 1990. Ani-
the animation may allow the scientist to ‘‘see’’ a     mated cartography: Thirty years of scratching the
simulation of the theory in action and deter-          surface. Cartographica 27 (2): 24–46.
mine which specific parts of the theory may or       DiBiase, David, Alan MacEachren, John Krygier, and
may not actually work.                                 Catherine Reeves. 1992. Animation and the role of
   This method may also be used to search for          map design in scientific visualization. Cartography
theories, rather than test existing theories. For      and Geographical Information Systems 19 (4): 201–14.
example, an urban geographer studying the            Dorling, Daniel. 1992. Stretching space and splicing
intraurban migration of various socioeconomic          time: From cartographic animation to interactive
groups may create a time-series or a thematic          visualization. Cartography and Geographical Infor-
animation for a given time frame. The result-          mation Systems 19 (4): 215–27.
ing animation might lead the geographer              Epstein, Melvin. 1990. Animation in the classroom: A
to discover socioeconomic migration-pattern            low-cost, educationally effective approach. T.H.E.
variations that were previously undetectable           Journal June:54–58.
through viewing static maps or data tables.          Harrower, Mark. 2001. Visualizing change: Using
Cartographers can use animation as a tool for          cartographic animation to explore remotely sensed
searching out the most effective map design            data. Cartographic Perspectives 39:30–42.
before production (Lavin, Rossum, and Slade          Krygier, J. B., Catherine Reeves, David DiBiase, and
1998). Or animation can provide a visual first         Jason Cupp. 1997. Design, implementation, and
look at remote-sensing imagery, possibly allow-        evaluation of multimedia resources for geography
ing a scientist to see patterns or change that         and earth science education. Journal of Geography in
may be difficult to discern from static images         Higher Education 21 (1): 17–38.
(Harrower 2001).                                     Lavin, Stephen, Sonja Rossum, and Shawn R. Slade.
   Finally, animation has become such a popu-          1998. Animation-based map design: The visual
lar spatial-data display method that GIS and           effects of interpolation on the appearance of
remote-sensing software packages are offering          three-dimensional surfaces. Cartographic Perspec-
the capabilities of animation creation. How-           tives 29:26–34.
ever, funneling a mass of different datasets         Moellering, Harold. 1980. The real-time animation
through the same default animation commands            of three-dimensional maps. The American Cartog-
may not necessarily result in the ideal anima-         rapher 7 (1): 67–75.
tion being created in each instance. Even if the     Peterson, Michael P. 1993. Interactive cartographic
classification scheme presented above is not the       animation. Cartography and Geographic Information
one employed to select an appropriate anima-           Systems 20 (1): 40–44.
tion method, determining the characteristics         Reiber, Lloyd. 1991. Animation, incidental learning,
of the dataset as well as the mapping purpose          and continuing motivation. Journal of Educational
will aid in developing the animation that most         Psychology 83 (3): 318–28.
effectively represents the data or theory. The       Thrower, N. 1959. Animated cartography. The
categorization presented above allows geog-            Professional Geographer 11 (6): 9–12.
raphers to identify an appropriate type of ani-      Tobler, W. 1970. A computer movie simulating urban
mation based on the characteristics of their           growth in the Detroit region. Economic Geography
datasets. An animation is most likely created to       46:234–40.
represent data. Just as a cartographer selects the   Weber, Christopher, and Barbara Buttenfield. 1993.
most appropriate thematic-mapping method               A cartographic animation of average yearly surface
based on the spatial data to be mapped, the            temperatures for the 48 contiguous states. Car-
geographer should ideally be familiar with their       tography and Geographical Information Systems 20:
dataset and know their mapping purpose before          141–50.
selecting an animation method.’
                                                     AMY K. LOBBEN is an assistant professor in the
                                                     Department of Geography at Central Michigan
Literature Cited                                     University, Mt. Pleasant, MI 48859. E-mail: Amy.
                                                     Lobben@cmich.edu. Her research interests include
Asoodeh, Mohannad, and Donald Clark. 1993.           cognitive maps and mapping, issues and methods
  Visualization of design concepts. The Technology   exploring navigational map-reading ability, and
  Teacher April:33–34.                               tactile maps.
You can also read