See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/306331480

Pilot Study Using the Augmented Reality Sandbox to Teach Topographic Maps
and Surficial Processes in Introductory Geology Labs

Article · August 2016

DOI: 10.5408/15-135.1

CITATIONS READS

97 2,126

5 authors, including:

Terri Lee Woods Sarah Reed

East Carolina University University of California, Berkeley
15 PUBLICATIONS 304 CITATIONS 9 PUBLICATIONS 182 CITATIONS

SEE PROFILE SEE PROFILE

Sherry Hsi
Biological Sciences Curriculum Study
82 PUBLICATIONS 5,224 CITATIONS

SEE PROFILE

All content following this page was uploaded by Terri Lee Woods on 25 June 2019.

The user has requested enhancement of the downloaded file.

 JOURNAL OF GEOSCIENCE EDUCATION 64, 199–214 (2016)

Pilot Study Using the Augmented Reality Sandbox to Teach

Topographic Maps and Surficial Processes in Introductory Geology
Labs
Terri L. Woods,1,a Sarah Reed,2 Sherry Hsi,3 John A. Woods,1 and Michael R. Woods4

ABSTRACT
Spatial thinking is often challenging for introductory geology students. A pilot study using the Augmented Reality sandbox
(AR sandbox) suggests it can be a powerful tool for bridging the gap between two-dimensional (2D) representations and real
landscapes, as well as enhancing the spatial thinking and modeling abilities of students. The AR sandbox involves a real box of
sand with virtual contour lines and a water flow model created using a three-dimensional (3D) scanning camera, visualization
software, and a projector. It was used in undergraduate, physical geology courses to teach topographic maps and surficial
features and processes. The instructor demonstrated topographic concepts (contour lines, topographic profiles, etc.), and
students engaged in model building of coastal and fluvial environments (drainage basins, cutoffs, longshore transport, sea-
level rise, spits, flooding, etc.). The virtual water flow model was used to illustrate water flow dynamics on surface features.
With the AR sandbox connected to a computer monitor, students could simultaneously see 3D landscapes in the sandbox and
their corresponding 2D images on the monitor. Students used camera phones to capture landscape models they built and
submitted them via e-mail for grading. Exit surveys indicated students were overwhelmingly positive (97%) in their perception
of how the AR sandbox improved their understanding of learning objectives. They also preferred AR sandbox activities when
compared to traditional laboratories that used only topographic maps. Effective classroom use of the AR sandbox required
developing student-modeling exercises that took advantage of real-time feedback, virtual water, and physical modeling
activities. While data are limited and more research is needed, real-time feedback on student models by both the students and
the instructor suggests sandbox models are particularly useful for dispelling student misconceptions. Ó 2016 National
Association of Geoscience Teachers. [DOI: 10.5408/15-135.1]

Key words: augmented reality sandbox, spatial thinking, embodied learning, topographic maps, geology labs, surficial
processes

INTRODUCTION topographic maps, help students understand the link

Spatial thinking is a fundamental skill for discovery and between 2D representations and 3D objects, and improve
problem solving in many disciplines, especially the geosci- student engagement (Lord, 1985; Feldman et al., 2010;
ences. Spatial thinking abilities—such as describing the Kuehn 2012; Atit et al., 2015). In addition, virtual models
shape, position, and orientation of objects; creating and such as interactive stereoscopic visualizations like the
reading maps; and visualizing processes in three dimen- GeoWall (Johnson et al., 2006) can also support spatial
sions—are critical to understanding the complex processes learning, for instance, by allowing students to see through
that take place on Earth (NRC, 2006). Despite its impor- an object and view objects from multiple angles, by
tance, students often have difficulty with spatial thinking and providing environmental context, and by preparing students
face challenges such as understanding scale, symbology, and for outdoor field research (Johnson et al., 2006; Reynolds et
how to connect two-dimensional (2D) representations with al., 2006; Rapp et al., 2007; Keehner et al., 2008). While
their three-dimensional (3D) counterparts (Chang et al., animations and interactive models are common in geosci-
1985; Ishikawa and Kastens, 2005; Liben and Titus, 2012). ences education, the uses of AR are relatively limited but
Educators have used a variety of models—physical, virtual, expanding (e.g., an AR sea-level-rise tool has recently been
and augmented reality (AR) models—in an effort to improve developed for teaching geoscience; Kintisch, 2013). AR
spatial thinking ability. models may make it possible to combine the interactive
benefits of physical models with the flexibility and diversity
Physical models such as raised relief maps and
of virtual tools.
sandboxes have been shown to improve understanding of
Caudell and Mizell (1992) coined the term ‘‘augmented
reality (AR)’’ to describe overlaying computer-generated and
Received 19 November 2015; revised 2 May 2016; accepted 23 May 2016; computer-presented information onto the real world. AR (1)
published online 19 August 2016. combines real and virtual images, (2) presents an interactive
1
Department of Geological Sciences, East Carolina University, East 5th
Street, Greenville, North Carolina 27858, USA image in real time, and (3) displays a scene in 3D, where real
2
The Lawrence Hall of Science, University of California–Berkeley, 1 and virtual objects are accurately aligned (Azuma, 1997). The
Centennial Drive, Berkeley, California, 94720, USA
3
presently pervasive navigational aids used in modern
The Concord Consortium, 6550 Vallejo Street, Suite 101, Emeryville, CA automobiles are examples of AR devices that superimpose
94608, USA
4
Carolinas IT, 1600 Hillsborough St., Raleigh, North Carolina 27605, USA
digital sightlines onto a camera view of a road, while the
a
Author to whom correspondence should be addressed. Electronic mail: plethora of popular video games exploits virtual reality. The
[email protected]. Tel.: 252-328-6014. Fax: 252-328-4391 former is closer to the real environment, and the latter is

1089-9995/2016/64(3)/199/16 199 Q Nat. Assoc. Geosci. Teachers

200 Woods et al. J. Geosci. Educ. 64, 199–214 (2016)

closer to the virtual environment. Many augmented or In this paper, we first describe the background and
virtual reality devices use mobile computers, head-worn origins for an AR sandbox designed for use in museums and
displays, and devices for global positioning system and science centers and then share information regarding its
wireless Web access. These systems often overlay computer- design, application, and use for an undergraduate geology
generated information and images onto real buildings, room lab course. We also provide a sample lesson plan (available
interiors, and exterior landscapes, among other settings. in the online journal and at http://dx.doi.org/10.5408/15-
In a systematic review of research and applications of AR 135s1) and explanatory video (ECU, 2016), as well as a
to education, Bacca et al. (2014) describe the uses, technical guide to the sandbox (available in the online
advantages, limitations, effectiveness, challenges, and fea- journal and at http://dx.doi.org/10.5408/15-135s2). Because
tures of AR in education by evaluating 32 studies published of significant recent upgrades to the AR sandbox software
between 2003 and 2013. Their results indicate that only 40% and online instructions, some of the issues we encountered
of the uses of AR in education are in science and that 85% of when constructing and programming our sandbox in fall
these applications have been to explain a given topic and 2014 have been resolved. Still, this supplement provides
augment information about it, as opposed to only 12.5% for insight on many important aspects of AR sandbox construc-
laboratory (lab) experiments. Cai et al. (2013) describe an AR tion that new users might find helpful.
3D technique used in a physics course to conduct an
interactive and integrated convex-lens, image-forming
experiment and an application in chemistry enabling THE AR SANDBOX
students to more readily envision the composition of Lake Visualization 3D (LakeViz3D, 2015b), a National
substances in a microworld (Cai et al., 2014). Andújar et al. Science Foundation–funded project, is a collaboration
(2011) developed augmented online labs for use in among the University of California (UC)–Davis W.M. Keck
engineering education at the University of Huelva. Their Center for Active Visualization in the Earth Sciences
simulated lab experiences are aimed at giving the user the (KeckCAVES, 2015), UC Davis Tahoe Environmental Re-
sensation that lab functions can be handled just as they search Center (TERC, 2015), Lawrence Hall of Science (LHS,
would be in the physical lab classroom. Several European 2015), Ecology, Culture, History, and Opportunities Leahy
projects have also developed AR applications (apps) for Center for Lake Champlain (ECHO, 2015), and Audience
educational purposes. For example, the Augmented Reality Viewpoints Consulting. The project’s primary objective is to
in School Environments (ARiSE) project is developing and increase understanding and stewardship of freshwater lake
piloting AR learning tools in classrooms to enhance student ecosystems using 3D visualizations. As part of the initiative,
understanding of topics such as the human digestive tract the interdisciplinary team built a real, hands-on sandbox
and chemical reactions (ARiSE, 2015). In the realm of K–12 exhibit, overlaid with virtual topography and water created
education, AR apps are becoming more readily available and using a motion-sensing input device (a Microsoft Kinect 3D
include some Earth Science topics (astronomy is particularly camera; Kreylos, 2015b), visualization software, and a data
popular). Kamarainen et al. (2013) described environmental- projector. The first prototype AR sandbox was developed at
education activities with middle-school students combining UC Davis KeckCAVES and allows users to create topo-
AR and probeware technologies—computer-aided, data- graphic models by shaping real sand on which an elevation
collection devices that both capture data with sensors (e.g., color map, topographic contour lines, and simulated water
temperature, pH, and force or acceleration probes) and are projected in real time. As users move the sand, the
analyze it with the connected computer—with a field trip to camera perceives changes in the distance to the sand surface,
a local pond. Until the AR sandbox (Kreylos, 2015a), and the projected colors and contour lines change accord-
however, little AR technology has been readily applicable ingly. When an object (e.g., a hand) is sensed at a particular
to the university geology classroom. height above the sand surface, virtual rain appears as a blue,
Many undergraduate students completing the lab shimmering visualization on the surface below and a flow
section of geology courses have difficulty with interpreting simulation moves the water across the landscape. The virtual
topographic maps and other spatial thinking tasks (Chang et water slowly disappears as if it was infiltrating the soil or can
al., 1985; Ishikawa and Kastens, 2005; Rapp et al., 2007; be drained rapidly with a push of a drain button. The
Clark et al., 2008; Alles and Riggs, 2011). Besides the LakeViz3D project (2015b) then created three AR sandbox
obstacle of the significant amount of math involved in exhibits (called Shaping Watersheds), which were installed
understanding and using scales, determining elevations and at the three science-center partner sites (LHS, TERC, and
relief, and interpolating latitudes and longitudes, the concept ECHO) and one portable AR sandbox for Howard University
of representing 3D landscapes in two dimensions (i.e., Middle School. They continue to be used by the public and
contour lines symbolizing elevations and contour spacing K–12 school programs to explore a variety of topics,
indicating steepness of topography) is challenging for most including geomorphology, hydrology, environmental stew-
students to grasp. The AR sandbox is a compelling tool for ardship, and watersheds. The exhibit software was inspired
bridging the gap between 2D representations and the real by Czech researchers who demonstrated an early prototype
world. When the modeled 3D landscape in the sandbox is called Sandy Station (SmartMania, 2011). Since its develop-
juxtaposed with its 2D image projected on a screen, the ment in 2013, more than 100 versions of this sandbox have
connection between the two is more accessible to novices been created in the U.S. and internationally (LakeViz3D,
than when they study a topographic map and try to envision 2015b; Reed et al., 2014). The original physical sandbox was
the real terrain. Another unique aspect of this particular AR made with plywood and mounting brackets with the help of
visualization is the ability to study the movement of virtual an undergraduate student (Peter Gold). The driving visual-
water interacting with real modeled landscapes to envision ization software (https://tinyurl.com/sandbox-download,
the processes by which landscapes evolve. under the GNU General Public License) was written by
J. Geosci. Educ. 64, 199–214 (2016) Using the Augmented Reality Sandbox in Geology Labs 201

Oliver Kreylos of the KeckCAVES at UC Davis and is based

on the virtual reality user interface (VRUI) development
toolkit (Kreylos, 2008; Kreylos, 2015c), the Kinect 3D video
processing framework, the OpenGL shading language, and
fluid flow models. The water flow simulation is based on the
Saint-Venant set of shallow water equations, which are a
depth-integrated version of the Navier-Stokes equations
governing fluid flow (Kurganov and Petrova, 2007). The
virtual sand surface is the boundary condition, and the
simulation is run such that the water flows exactly at real
speed, assuming a 1:100 scale factor. Details about the
sandbox and its development can be found on the project
Web site (Kreylos, 2015a) and in the technical supplement to
this paper (available in the online journal and at http://dx.
doi.org/10.5408/15-135s2).

Parts and Construction of the Sandboxes at East

Carolina University
Guided by information from the LakeViz3D Web site
(LakeViz3D, 2015a), staff and faculty at East Carolina
University (ECU) Department of Geological Sciences con-
structed and calibrated two AR sandboxes (Fig. 1, see color
versions of the figures in the online version of this
manuscript) in fall 2014 for use in a university lab setting.
The ECU team adapted the original sandbox components to
construct two sandboxes (total cost of about $1,000), one for
each of the two lab classrooms (see the supplemental
materials on the construction, calibration, and programming
process; available in the online journal and at http://dx.doi.
org/10.5408/15-135s2). While the cost can vary enormously
depending on the construction materials, projector, and
computer used, a complete sandbox can be built for under
$1,000 with used parts (not including staff time). For
instance, the ECU team employed some materials already
available in the department, including used computers, a
donated video card, cables and wiring to connect camera
FIGURE 1: ECU AR sandbox linked to display monitor
and monitor, and woodworking tools. Alternatively, private
to permit viewing of both 2D and 3D images simulta-
design firms offer exhibit-grade sandboxes starting at
neously. Wooden framework was designed to be sturdy
$10,000. By not using all recommended equipment, the cost
and portable and to fit through classroom doors. The
was minimized, but that savings resulted in somewhat
short-throw projector is mounted vertically on the
compromised performance. For example, the water simula-
upright wooden board to the back. The right end of
tion is slightly slower than with the recommended GTX970
the Microsoft Kinect camera (suspended at the front of
video card, and the less expensive sand used in the ECU
the overhanging, forward-projecting, wooden mounting
sandboxes is not as good for this application as white
board) can be seen extending a bit to the right of the
Sandtastik sand, resulting in a projected image that is
projector. Some key dimensions are indicated.
slightly less bright and true to color. The time required for
sandbox construction can also vary widely depending on
staff expertise. The ECU sandboxes were built, the software
was installed, and the AR sandboxes were operational in 1 the projector and computer. The operating system then
week. boots, logs in the correct user, and launches the sandbox
When configuring the operating system and software, software, all without requiring input from the user.
the goal was to anticipate the most likely usage pattern for
instructors and to remove probable obstacles. At ECU, most
users of the AR sandbox will be graduate teaching assistants LOCATION AND CONTEXT OF THE PILOT
(GTAs)—usually master of science (MS) candidates, with STUDY
varying degrees of familiarity with computers and usually no To understand the institutional and instructional context
experience with the Linux operating system. To simplify for this work, a brief description of ECU and its student body
sandbox operation (i.e., minimize unnecessary navigation is provided here. This is followed by an overview of the
and remove the need for operation using the command-line exercises and an outline of the individual learning objectives
interface), adjustments were made after installing the for each of the three lab topics (topographic maps, rivers,
sandbox software. Finally, the sandbox program was set to and coastlines).
start automatically upon user login. With these adjustments, ECU is a regional, doctoral-granting (Carnegie classifi-
an instructor can simply walk into the classroom and turn on cation: higher research activity), liberal-arts institution with
202 Woods et al. J. Geosci. Educ. 64, 199–214 (2016)

TABLE I: Topics historically covered in GEOL 1501, and syllabus revisions made to accommodate AR sandbox activities in summer
2015. Labs that were changed in summer 2015 are indicated in bold.
Typical Lab Syllabus Without Sandbox Exercises Syllabus With Sandbox Exercises, Summer 2015
Lab Topic Lab Topic
1 Properties of minerals 1 Properties of minerals
2 Mineral identification 2 Mineral identification
3 Mineral exam and igneous rocks 3 Mineral exam and igneous rocks
4 More igneous rocks 4 More igneous rocks
5 Sedimentary rocks 5 Sedimentary rocks
6 Metamorphic rocks and geologic time 6 Metamorphic rocks and geologic time
7 Exam on rocks and geologic time and introduction to 7 Exam on rocks and geologic time, 45-minute
contouring spot elevations demonstration of the sandbox, and introduction to
contouring spot elevations
8 Basics of topographic maps 8 Basics of topographic maps using topographic maps
and the sandbox
9 Studying rivers using topographic maps 9 Studying rivers using the sandbox and topographic
maps
10 Quiz on topographic maps, hydrographs, flood-insurance 10 Quiz on topographic maps and continuation of river
rate maps, and flood recurrence diagrams exercises using topographic maps and the sandbox
11 Groundwater 11 Short exercise on stream hydrographs and
continuation of river exercises using the sandbox
12 Studying shorelines using topographic maps 12 Groundwater exercises and shoreline exercises using
topographic maps and the sandbox
13 Crustal deformation and earthquakes 13 Continuation of shorelines exercises using topographic
maps and the sandbox
14 Plate tectonics 14 Plate tectonics
15 Final lab exam 15 Final lab exam

an enrollment of ~28,000 students, of which about 80% are student teaching assistant begins each class with a 15- to 20-
undergraduates and 20% are graduate students. Nearly 90% minute introduction to the topic. Assessment for this course
of ECU undergraduates (average age 22 years) are from typically involves four in-class exams or quizzes (~67%) on
North Carolina originally, and 32% qualify as low income. In minerals, rocks and geologic time, and basics of topographic
2015, composite SAT scores (reading and math) for degree- maps (quiz) and the final exam covering the remaining
or certificate-seeking students ranged from 980 to 1,120 topics (Table I). The remaining 33% of the lab grade is
(Forbes, 2016). derived from prelab exercises submitted at the beginning of
In the Department of Geological Sciences, there are 10 each lab period and daily exercises completed during the lab
faculty members who routinely teach the large (100 period. Other than taking the lecture portion of the course,
students) introductory lecture classes in basic physical prelab reading in the lab manual, and responses to prelab
geology (GEOL 1500: Dynamic Earth) and typically 10 questions, students are not expected to have skills or
graduate students (MS only—no doctoral candidates) who knowledge related to the topics covered.
teach the accompanying lab course (GEOL 1501: Dynamic
Earth Laboratory). The topics for which the sandbox was Topographic Maps Lab
used are currently taught as part of GEOL 1501, which meets Learning to use topographic maps can be challenging
for 3 h each week of the semester. During the 5-week for college students, but facility with these maps is crucial to
summer session, in which use of the AR sandbox was developing an understanding of surficial processes associat-
ed with rivers, glaciers, oceans, and groundwater. This skill
piloted, the lab course met three times a week for 3 h each
can also help students in their everyday lives when using
class period. This course, along with its accompanying
road maps, property maps, and navigational charts. The
lecture, is dominantly populated by nongeology majors
specific learning objectives for this lab include the following:
fulfilling the lab requirement of their general science-
education curriculum. Most students take the lecture portion  Interpret information and understand symbols on
of the course during the same semester in which they take maps
the lab, although some students enroll in the lab during a  Determine latitudes and longitudes for locations on
later semester. The lecture is taught by faculty members, and Earth’s surface
during the fall and spring semesters, labs are taught by  Calculate distances between locations, as well as
graduate students. Typical enrollment in a lab is 23 students. elevations and relief
Because lab students have not necessarily had the lecture  Determine contour spot elevations
material for a specific topic before entering lab, the graduate-  Construct topographic profiles
J. Geosci. Educ. 64, 199–214 (2016) Using the Augmented Reality Sandbox in Geology Labs 203

 Develop the ability to envision a terrain based on its TABLE II: Demographics of students participating in this pilot
topographic map study.
Demographic No. Students
Rivers Lab
In a humid climate such as that in North Carolina, Class rank
students seldom find themselves farther than a few hundred Freshman 2
yards from running water. Understanding fluvial processes Sophomore 5
and features can help them avoid serious loss of life and
Junior 2
property, such as that the state experienced following
Hurricanes Dennis and Floyd in 1999. Due to the geographic Senior 3
diversity of North Carolina, all types of streams are found, Major
ranging from high-gradient mountain rivers flowing in Social science 2
narrow, V-shaped valleys to low-gradient, coastal creeks
on broad coastal plains. North Carolinians must, therefore, Science and technology 5
be familiar with the entire spectrum of fluvial systems. This Fine arts 2
knowledge permits potential property owners to make wise Health and human performance 1
decisions before they purchase and empowers informed
citizens to influence municipal officials as they make Education 2
decisions regarding land use near rivers. As the population Gender
of the state increases, water-supply issues will require even Male 8
more difficult choices to be made about North Carolina
Female 4
streams and rivers. Specific learning objectives for this lab
include the following: Race/ethnicity
White 11
 Recognize factors that affect stream flow, velocity, and
erosion African American 1
 Understand the concept of gradient
 Learn about flooding and how human development
can affect it
 Use hydrographs to describe river flow
 Explore the work that rivers do: erosion, transporta- PREPARATION FOR SANDBOX LABS
tion, and deposition The sandbox exercises were designed and taught by the
 Recognize the erosional and depositional features first author (T.L.W.), who has been teaching this class (and
created by rivers other undergraduate and graduate geology classes) since
 Develop the ability to envision the processes that 1988. The summer lab course had only 9 students, which
create fluvial features offered an ideal situation for exploring the learning potential
of the sandbox and developing sandbox lesson plans to pilot,
Coastlines Lab especially when compared to the larger sections offered
Coastal change has significant political and social during fall and spring semesters. Three students who were
consequences for the U.S. in general and North Carolina enrolled in lecture but not lab also spent several hours
in particular. If current trends in climate change continue, working in the sandbox, thereby exposing a total of 12
coastal regions will experience profound effects from sea- students to this new teaching tool (Table II). All but 1 of
level rise. Flooding of low-lying coastal areas requires these 12 students was of traditional college age. The faculty
citizens to make hard decisions about preserving infrastruc- member teaches both lecture and lab during the summer
ture and property in the coastal zone. As a result, it is semester, so the syllabus (Table I) was designed to allow
important for voters to understand natural coastal processes lecture background on lab topics to be covered before
and the impacts humans have on them. Specific learning students undertook the lab exercises on that topic. However,
objectives for this lab include the following: because several lab students had taken the lecture portion of
the course in a previous semester, the instructor presented a
 Learn the basic terminology and behavior of wind- 15- to 20-minute review of each lab topic.
formed waves Time is precious in 3-h university labs, and several
 Investigate the link between wave refraction and weeks of lesson design, testing potential activities, and
longshore transport acquiring supplies were required to ensure course objectives
 Recognize the types of coastal features formed by were achieved and that AR sandbox activities would work
erosion and deposition and not take too long. A long list of potential features,
 Learn ways humans have attempted to interfere with processes, and terrains to be modeled in the sandbox was
natural coastal processes considered, but many were discarded as impractical. For
 Understand results of human interference with example, the sandbox could potentially be used to explore
natural coastal processes (1) the impact of differing cross-sectional areas, wetted
 Describe how the barrier islands of North Carolina perimeters, and channel roughness on stream flow; (2) the
originated concept of changing base levels; and (3) meandering versus
 Differentiate submergent, emergent, primary, and braided streams, but it was discovered these would not work
secondary coasts because the loose sand used in the ECU sandbox does not
204 Woods et al. J. Geosci. Educ. 64, 199–214 (2016)

eliminating about one-fourth of the topographic-map

questions on shorelines and groundwater.
Sandbox activities did not serve as a replacement for
crustal deformation and earthquake exercises pursued in Lab
13 in previous semesters; those activities were eliminated
from the lab. In the past, the exercises relating to crustal
deformation would have required approximately 1 h of a lab
period, and the computer exercise developed to study the
distribution of earthquakes worldwide required about 2 h.
These were deemed the least crucial of the GEOL 1501
exercises for different reasons. Crustal deformation (faulting
and folding of rocks usually occurring many miles beneath
Earth’s surface) generates features usually only exposed at
Earth’s surface following millions of years of weathering,
erosion, and uplift. Furthermore, crustal deformation is
difficult for most students to visualize, is perceived by them
as irrelevant to their daily lives, and is not effectively dealt
with in the time typically allotted for it, so it was deemed
FIGURE 2: Top view of sandbox terrain constructed for expendable. In contrast, earthquakes are a crucial aspect of
an initial contour-line demonstration by instructor in a internal Earth processes for people living in regions
lab class. The mound was constructed with the steeper experiencing frequent earthquakes. However, these phe-
terrain on the left side and a valley carved at the top to nomena are not perceived as relevant in the everyday lives of
demonstrate the Rule of Vs. Print version of the figure: most ECU students. Also, this topic is well covered in the
Lightest shade in the center left indicates the highest lecture portion of the course, so most students finish GEOL
elevation, whereas the other light shade at the edges of 1500 with a reasonable idea of the distribution of regions
the image surrounding the central mountain represents experiencing frequent earthquakes.
lower elevations. Online version of the figure: Elevations Final lab preparations included placing a whiteboard
are color coded, moving from highest to lowest as color along the back of the sandbox to record sand elevations as
changes from white to green, orange, yellow, and then students create a topographic profile of a 3D sandbox
gray. (Sandbox exterior dimensions are 28 00 · 38 00 .) landscape. A corkboard and pushpins would serve the same
purpose. In addition, a mound of sand (steeper at one end
hold the required shapes. (In the original use of the AR than the other) was built before the beginning of class. The
sandbox using white sand from Sandtastik, a spray bottle of mound was designed to fill most of the box, thereby
water has been helpful to minimize dust and help with minimizing the amount of time required to add the virtual
landform shaping.) Implements that were useful for gener- water needed for the demonstration to follow (Fig. 2). Sand
ating sandbox models included large pieces of wood to in the rest of the box was flattened to limit irregularities in
smooth sand throughout the entire box, small shovels and the contour lines around the mound. In addition, a piece of
buckets, thin sticks with cardboard squares taped to the end transparency film was taped over the sandbox image on the
computer monitor (Fig. 1) to record student observations of
to generate rainstorms on limited areas, wooden models of
the changing positions of the shoreline around the mound as
buildings, rollers for smoothing sand, rocks and wood for
virtual water is drained from the sandbox. Finally, a single
jetties and groins, string, dowels, markers, and transparency
distinctive river valley was carved in one side of the
film.
mountain as a demonstration of the Rule of Vs (the Vs on
Sandbox activities take time that in the past was a topographic map open toward lower elevations; Fig. 2).
allocated for other exercises. Table I outlines the changes In the typical introductory ECU lab on topographic
made to the previous lab syllabus to accommodate sandbox maps, students are introduced to the rules of contour lines
activities. Lab periods highlighted in bold on the right side of and the basic techniques of contouring spot elevation data
Table I involved significant use of the sandbox for and asked to produce a simple topographic map. Although
demonstrations and exercises. To make time for sandbox many have seen contour maps before (e.g., atmospheric
activities, a portion of the questions students would pressure maps on weather reports), most have not thought
previously have been answering (based on their individual critically about what the maps show and have never been
study of topographic maps) was eliminated. Lab 7 is faced with the task of generating contour lines from
traditionally one of the shorter lab periods, so addition of individual data points. Their first attempts almost invariably
the sandbox demonstration did not require elimination of contain numerous, significant errors. In the sandbox-based
any activities pursued in previous years. Sandbox exercises course, this contouring exercise was the initial activity;
involving student construction of simple terrains (described however, immediately afterward, the sandbox was intro-
later) were added to the standard topographic-map activities duced to clarify the concept of contour lines. The activities
pursued in Lab 8, without requiring elimination of any map described next were pursued with the sandbox connected to
exercises. Time for sandbox activities was generated in Labs a computer monitor so that students could see the 2D image
9–11 by eliminating about one-fourth of the topographic- on the monitor at the same time they were looking at the 3D
map questions and shortening the hydrology exercises in landscape in the box (Fig. 1; see the detailed lesson plan for
Lab 11 (by eliminating the flood recurrence activity). In Labs the demonstrations and exercises, available in the online
12 and 13, time for sandbox activities was generated by journal and at http://dx.doi.org/10.5408/15-135s1). (With
J. Geosci. Educ. 64, 199–214 (2016) Using the Augmented Reality Sandbox in Geology Labs 205

FIGURE 3: Oblique view of the instructor’s model used to demonstrate the construction of a topographic profile. A
whiteboard has been placed upright in the back of the box so that students can record changing elevations along the
profile line. Print version of the figure: The light line along the center of the terrain (from left to right) is a string laid
along the profile line at a point representing maximum elevation changes. The mottled gray-and-white pattern along
some edges is virtual water. Online version of the figure: Elevations are color coded, moving from highest to lowest as
color changes from white to green, gray, reddish orange, and then yellow. The mottled blue-and-white pattern along
some edges is virtual water. (Sandbox interior dimensions are 28 00 · 38 00 ). The yellow line along the center of the
terrain (from left to right) is a string laid along the profile line at a point representing maximum elevation changes.

ECU’s hardware configuration [including the screen of a describe what was different about the spacing of contour
computer monitor or a projected image on a screen] the lines at these two ends.
program crashed if the cursor was positioned on the monitor Subsequently, after rapidly flooding the area around the
when a numerical key was pressed. Such random problems mound, the drain button was used to progressively dry up
may result from using other than LakeViz recommended some of the water. Stopping the process after each water-
hardware configurations.) The computer was also connected level decline, successive student volunteers traced the edges
to a printer for capturing screenshots, but the ubiquity of of the mound-to-water contact onto the transparency film
camera phones renders this largely unnecessary. taped over the computer screen. When the transparency film
was removed from the monitor, students recognized that
they had generated a contour map. Many students
INITIAL DEMONSTRATION ON CONTOUR commented immediately how easy it now was to see the
LINES AND TOPOGRAPHIC PROFILES relationship between the sandbox landscape and the lines on
the transparency. They could quantify elevations or water
Contour Lines
depths by counting up or down from a reference point (e.g.,
Before beginning the demonstration on Contour Lines
the bottom of the sandbox) to the water surface or the
and Topographic Profiles (East Carolina University, 2016),
contour line in question. Subsequently, with two mounds in
the instructor gathered the students around the AR sandbox the sandbox—one higher and steeper than the other—the
and explained its parts and how they work together to instructor explained the concept of relief, both local and
generate the projected images. Then students were asked for total, and asked, ‘‘Which mountain has the greater relief?’’
their observations about contour lines. This did not elicit any ‘‘Which mountain is steeper?’’ and ‘‘How would you
response, so they were asked to kneel down until their eyes determine local versus total relief for this landscape?’’ Then,
were level with the rim of the sandbox and told to follow a after scraping off the top of the higher and steeper mound so
single contour line all the way around the mound, noting that it was the same elevation as the smaller one, students
how its position changed with respect to the rim of the were asked again, ‘‘Which has the higher relief?’’
sandbox. Then, to prompt observations, they were asked
questions such as ‘‘Does the line’s elevation change with Topographic Profile
respect to the rim of the AR sandbox?’’ ‘‘Do contour lines After building two mounds with different heights,
close to form loops?’’ ‘‘Do they cross?’’ and ‘‘Do they reliefs, and shapes (Fig. 3), the instructor laid a brightly
branch?’’ After establishing some of the basic characteristics colored piece of string in the sand along a straight line across
of contour lines, students were asked, ‘‘What shape do the the features (parallel to the whiteboard at the back of the
lines in the river valley remind you of?’’ and then ‘‘In which sandbox) to represent the trace of a topographic profile.
direction does the sharp end of the ‘V’ point—uphill or Students took turns placing skinny dowels laid horizontally
downhill?’’ Finally, they were directed to study the steeper from the front to the back of the box—perpendicular to the
and more gently sloping ends of the mound and asked to string and parallel to successive contour lines. One student
206 Woods et al. J. Geosci. Educ. 64, 199–214 (2016)

FIGURE 4: (a and b) Map views of simple terrains and (c and d) student models of these terrains. Print version of the
figure: In (c) and (d), the lightest shade in the center of the mountains indicates the highest elevation, whereas the
other light shade at the edges of the image represents the lowest elevations. The mottled gray-and-white pattern in
some corners of (d) is virtual water. Online version of the figure: In (c) and (d), elevations are color coded, moving
from highest to lowest as color changes from white to green, gray, reddish orange, yellow, and then bluish gray. The
mottled blue-and-white pattern in some corners of (d) is virtual water. (Sandbox interior dimensions are 28 00 · 38 00 ).

kneeled in front of the box and sighted toward the relief, steepness, and location were generally correct; and if
whiteboard at the back, keeping the dowel horizontal at the extent of low- and high-relief sections was generally
the elevation of the contour line. A second student placed a correct, students received full credit. Figures 4(c) and 4(d) are
dot on the whiteboard where the dowel touched it. Finally, typical of the models constructed, indicating that most
the students connected the dots to complete the profile. students could correctly interpret these simple contour maps
Later in lab, the instructor explained how to create [Fig. 4(a) and 4(b)] and visualize the 3D terrain they
topographic profiles for different traverses across the paper represent. As time allowed, the instructor also checked with
topographic maps. students working in the sandbox to address questions or
misconceptions that arose as they worked.

CLASS EXERCISES Modeling Fluvial Features and Processes

Constructing Terrains From Simple Contour Maps For the subsequent lab on rivers, the sandbox was
As graded exercises, students constructed simple ter- used for two different but intimately interwoven purposes:
rains in the sandbox using 2D diagrams of these terrains as a (1) to have students build different fluvial features and (2)
guide [Figs. 4(a) and 4(b)]. This part of the exercise was done to help them better understand the interaction of river
when the students could get a timeslot, in pairs, with the water with the landscape. They were encouraged to use
sandbox, while other students were working on the standard the water flow feature to make their models more realistic
topographic-map exercises. Students took photos of their and more useful as representations of the natural world.
resultant terrains and e-mailed them to the instructor as part They based their models on lecture materials, class
of their daily lab grade. readings, textbooks, and lab manuals about these features
For the purposes of grading, the detailed shape of the and processes, as well as what they were seeing on the
feature and its absolute relief were not of primary concern. topographic maps of rivers they were studying. Again,
Instead, if the landforms were properly oriented; if their size, students worked in pairs, taking photos of their efforts and
J. Geosci. Educ. 64, 199–214 (2016) Using the Augmented Reality Sandbox in Geology Labs 207

FIGURE 5: Print version of the figure: Top views of student models built to represent (a) stream divide and (b)
meander (left) with an oxbow lake (right). (a) Objects running top to bottom down the center of the stream divide are
rock chips marking the points of maximum elevation along the stream divide. Elevations increase up and away from
both the meandering stream and the oxbow lake in all directions. The mottled dark-and-light patterns to the left and
right of the stream divide and within the meandering channel and oxbow represent virtual water. Online version of
the figure: Top views of student models of fluvial features. (a) Stream divide with elevations color coded, moving
from highest to lowest as color changes from greenish white to green, gray, reddish orange, and then yellow. Blue
rock chips mark the points of maximum elevation along the stream divide. (b) Meander (left) and oxbow lake (right)
with elevations color coded, moving from highest to lowest as color changes from green to reddish orange. The
mottled blue-and-white pattern in both photos is virtual water. (Sandbox interior dimensions are 28 00 · 38 00 .)

e-mailing them to the instructor for grading. Each pair had  Erosion and deposition on the outside and inside of
to construct a model showing a drainage divide and had to meanders
use the water flow model to study effects of rain events  Formation of an oxbow lake
and flooding [Fig. 5(a)]. A primary goal was to encourage  Formation of deltas and alluvial fans
students to think about how water moves through  Headward erosion and stream piracy
drainage basins and whether water can naturally move
on the surface from one drainage basin to the neighboring Some of those modeling flooding built a stream with a
one. Subsequently, they had a choice of which features to wide floodplain and a stream with virtually no floodplain
model, such as follows: and then investigated how far flood water moves away from
 these two streams. They also modeled two paths for a
The contiguous 48 states, its two major continental
hurricane (perpendicular and parallel to drainages) to show
divides, and the Mississippi River Basin
 Point bars, relict point bars, and bars within streams different flooding patterns resulting from different storm
 Stream with natural levees and a yazoo tributary paths. In addition to submitting images to the instructor for
 Difference between channel length and valley length grading, a model-building competition between groups of
to explain sinuosity students was introduced into these activities as an instruc-
 Cutoff, oxbow lake [Fig. 5(b)], and abandoned tional device. This teaching tool was effective in the small
meanders summer lab (only nine students), during which it was
 Various drainage patterns possible to use both sandboxes for the same class. Two pairs
of students at a time were charged with modeling a specified
Their choices for processes included the following: terrain or process. Students not involved in building that
particular landscape then studied the models and selected
 Formation of a meander, starting with a straight the winner. All students in the class were building models in
channel and putting an impediment in the way of the box and studying topographic maps showing landforms,
water flow so no rubric was provided for their determination of the
208 Woods et al. J. Geosci. Educ. 64, 199–214 (2016)

FIGURE 6: Top view of student models of backshore and foreshore regions along a beach at (a) high tide and (b) low
tide. Print version of the figure: The mottled dark-and-light pattern in both images is virtual water projected onto the
sandbox. Beach elevations increase toward the top of the photograph, and the ocean is at the bottom. Online version
of the figure: Elevations are color coded, moving from highest to lowest as color changes from green to pale green,
reddish orange, and then yellow. The mottled blue-and-white pattern in both photos is virtual water. (Sandbox
interior dimensions are 28 00 · 38 00 .)

winners of these competitions. They were asked to decide

based on which model was generally the most recognizable,
accurate, and complete. Somewhat surprisingly, although
they were not provided with a rubric, the decision in all cases
was unanimous.

Modeling Coastal Features and Processes

Students followed the same approach for coastal
geology as they did for fluvial features and processes. They
constructed models of features such as follows:
 Barrier islands of North Carolina with major inlets
 Shoreline showing backshore, foreshore, and offshore
and their appearance during high tide [Fig. 6(a)]
versus low tide [Fig. 6(b)]
 Western U.S. coast with sea stacks, tombolos,
headlands, and bays
 Coastline with barrier islands, ebb and flood tide
deltas, and spits
FIGURE 7: Top view of student model showing the effect
on beach width of longshore transport in the surf zone. Coastal processes modeled included the following:
Print version of the figure: Model shows sand buildup  Groins and their effect on beach width upstream and
on the upstream side (left) of a coastal groin (represent-
downstream (Fig. 7)
ed by a piece of granite) and beach erosion on the  Cape Hatteras National Seashore, North Carolina,
downstream side (right). The mottled dark-and-light
pattern to the bottom of the image is virtual water. showing where the sand eroded from the former
Higher beach elevations are at the top of the photo- lighthouse location goes as it is carried southward by
graph. Online version of the figure: Model shows sand longshore transport
 Sea-level rise eroding the shallowly sloping North
buildup on the upstream side (left) of a coastal groin
(represented by a piece of granite that appears green Carolina mainland faster than the barrier islands
because the camera senses it as a region of higher migrate landward
elevation). Beach erosion on the downstream side (right)  Longshore transport closing the mouth of an inlet
is indicated by the beach migrating landward (toward  Conversion of a spit into a bay-mouth bar
top of photo). The mottled blue-and-white pattern is  Barrier island migration by sediment overwash
virtual water. (Sandbox interior dimensions are 28 00 ·  Evolution of a coastal river into an estuary as sea level
38 00 .) rises
J. Geosci. Educ. 64, 199–214 (2016) Using the Augmented Reality Sandbox in Geology Labs 209

TABLE III: Exit survey administered to lab and lecture

students in summer 2015.
Questions (N = 10, 7 From Lab and 3 From Lecture)
1. Have you ever seen an augmented reality model such as
this? Specifically, have you ever seen anything about this
particular model on the Web?
2. How would you describe the sandbox to someone who has
never seen it?
3–6. Students were asked to respond with 1–5, where 1 =
strongly disagree and 5 = strongly agree
3. The sandbox helped me learn about topo maps, contour
lines, relief, etc.
4. The box helped me learn about Earth surface features and
processes such as barrier islands, point bars, longshore
transport of sediment, floods, etc.
5. I learned some things using the box that will help me in
my everyday life.
6. I will tell others about the exercises I did with the
sandbox.
7. If you chose 4 or 5 for question 3 above, how in particular
FIGURE 8: Top view illustrating wave refraction pattern has it helped you understand topo maps better? Specifically,
projected onto sand by water flow software. The what are your most vivid recollections of how the sandbox
sandbox has been completely flooded with virtual water allowed you to better understand them?
for this image. Wave fronts can be seen to refract around
8. If you chose 4 or 5 for question 4 above, how in particular
the rock island in the center, especially above and below has it helped you understand them better? Specifically, what
the island. Print version of the figure: Longer, dark are your most vivid recollections of how the sandbox
curving lines represent wave fronts. Online version of allowed you to better understand these features and
this figure: Longer, dark-blue, curving lines represent processes?
wave fronts. (Image measures approximately 27 00 · 36 00 .)
9. If you chose 4 or 5 for question 5 above, what in particular
have you learned that will help in your everyday life?
 Straightening of a coastline by erosion of headlands 10. What did you like best about the sandbox exercises?
and deposition in bays
 California beach compartments and offshore flow of 11. What did you like least about the sandbox exercises?
sand to submarine canyons 12. What would make this more useful for future students?
13. What would make this more interesting for future students?
One particularly useful aspect of the water flow
simulation is that it is possible to show how waves are 14. Do you have any suggestions on how the sandbox activities
could be adapted for a larger lab class (23 students) taught
reflected from vertical surfaces and refracted around shallow by a single graduate-student teaching assistant? During a
spots (Fig. 8). normal 3-h lab, the teaching assistant has trouble
answering all questions and helping that large number of
students. Any ideas on how the teaching assistant could
STUDENT LEARNING AND COMMENTS enhance student learning without input of excessive
At the end of the semester, the 12 students (9 from the amounts of time?
lab and 3 from the lecture only) were asked to complete an 15. Did you enjoy the sandbox exercises more than working
exit survey (Table III), providing feedback on their percep- with hard copies of the topographic maps, and do you
tion of the value of the AR sandbox, as well as insights on think you learned more with the sandbox activities?
how to improve the activities and scale them up for the 16. Which of the activities was the most affective? Which the
normal enrollment of 23 students. The objective of the least?
survey was to help improve instruction to maximize student 17. Please write any other comments you have about these
learning using this new technology. Seven of the nine lab exercises.
students and all three lecture-only students responded, for a
total of 10 replies.
None of the respondents had ever seen an AR model sand and a thing is on top of it to measure the height of the
like the sandbox. When asked to describe the AR sandbox, sand. This can create topography maps. (It’s awesome),’’ and
some responses were fairly explicit, such as ‘‘The sandbox is ‘‘A sandbox with a brain!’’
basically a projection of hundreds of still images recorded To quantify students’ overall response to this new
and projected from above onto a sand area below. Contour teaching tool, numerical evaluations (1 = strongly disagree
elevation lines (and different colors for different elevations) through 5 = strongly agree) of their responses to Questions
are projected according to how you shape the sand. It is 3–6 were solicited (Table III). Students were universally
helpful if you are a visual, 3-dimensional learner.’’ Others positive (97%) in their perception of the helpfulness of the
were enthusiastic, if not very informative: ‘‘It is a box with AR sandbox for understanding topographic maps and
210 Woods et al. J. Geosci. Educ. 64, 199–214 (2016)

surficial features and processes. A score of 4.6 out of 5 education recently identified by the Natural Research Council
suggests they would tell others about these exercises, and 4 (Schweingruber et al., 2012) and deeply integrated in the Next
out of 5 indicated they perceived they had learned Generation Science Standards (NGSS; NGSS Leads States,
something using the AR sandbox that would help them in 2013). Scientific modeling is a process that involves producing
their everyday life. Comments related to the latter included concrete representations of abstract ideas and evaluating and
‘‘Where to build a house! Definitely not on the beach or the revising the representations based on testing and observations
bottom of a valley,’’ and ‘‘As elementary as it might sound, it (Gobert and Buckley, 2000). Because the sandbox is a mixed-
helped me to grasp elevations better. I would like to hike reality environment in which students can interact with a
more with my father whose eyesight is starting to fail him. If physical model overlaid with abstract diagrams, while testing
I am better equipped at reading maps, it could aid us in our and exploring specific science concepts, it may support the
journeys together.’’ development of scientific model-building skills. That is,
Students also expanded on their numerical evalua- because the sandbox experience can merge scientific content,
tions, giving specific information about how they per- practice, and knowledge development, it may be an important
ceived the sandbox helped them learn. Many of the more tool for helping students understand how scientists create
general responses indicated how well they thought the models that explain Earth systems and begin to practice model
sandbox allowed them to visualize what the 2D contour building themselves (Louca and Zacharia, 2012; Bryce et al.,
lines were trying to describe about the 3D landscape. 2016). The sandbox, then, may also be a useful means by
Some mentioned specifically that they believed the AR which to address the focus on model-development practices
sandbox made it easier to visualize flow direction of outlined in the NGSS. These practices are developed via the
streams and steepness of slopes compared to contour lines means of learning progressions: ‘‘research-based cognitive
on maps. Two student comments are particularly illustra- models of how the learning of scientific concepts and practices
tive of this outcome: unfolds over time’’ (Duncan and Rivit, 2013). Because research
on learning progressions is in the early stages, the AR sandbox
‘‘The first time we were introduced to the sandbox, something could not only become a productive tool for geoscience
clicked in my brain. I was able to learn and interpret contour education but also could be used to understand the
lines and elevations more accurately in that 15 minute development and evolution of scientific reasoning.
introduction than in an hour by myself just using a Many students were particularly intrigued by the
topographic map. I am a visual learner, and it was such water flow model, and their comments suggested they
an asset having this tool in the classroom.’’ believed it had helped them better understand surficial
processes and features. Instructor observations and
student comments clearly indicated students enjoyed the
‘‘The sandbox helped me better understand topo maps from a sandbox exercises more than the traditional topographic-
visual perspective. Sometimes it’s hard to picture what the map exercises. Speaking for herself, the first author can
terrain of the land actually looks like based on a bunch of state that using the sandbox in summer 2015 vastly
lines on a map. The sandbox allowed [me] to actually improved her enjoyment teaching this portion of GEOL
visualize how the contour lines matched up with hills, 1501. Selected responses on this and other topics are
mountains, etc.’’ compiled in Table IV. Most students used the word ‘‘fun’’
somewhere in their other comments on sandbox activities,
Any perceived or actual learning gains resulting from and many descriptions included these adjectives: refresh-
use of the AR sandbox are likely in part due to the direct ing, incredible, creative, interactive, hands-on, and real-
link between the 3D landforms and processes and their 2D istic. Not surprisingly, when asked whether (1) they
representations that the sandbox makes possible. In enjoyed the sandbox exercises more than working with
particular, the sandbox may help with spatial learning by hard copies of the topographic maps and (2) they thought
providing an opportunity for embodied learning. The they learned more with the sandbox activities than with
impact of action on learning has long been recognized by the maps, the response was a unanimous ‘‘Yes.’’ Com-
developmental psychologists (Piaget, 1952; Held and ments such as this one were common: ‘‘The sandbox
Hein, 1963). Embodied cognition refers to the notion that activity was most effective for me. Looking at regular
an individual’s abstract representations of a concept are diagrams can be confusing and boring.’’ One of the
often based on a somatic experience of the concept lecture-only students remarked, ‘‘Looking forward to lab
(Niedenthal, 2007). Research is demonstrating that em- in order to use the sandbox again!’’ When asked what they
bodied learning shows promise to support powerful liked least about the sandbox, a few students mentioned it
learning experiences (Lindgren and Johnson, 2013; Abra- was ‘‘messy.’’ Others mentioned an issue with the rain
hamson and Lindgren, 2014). Being able to interact with function in that it takes practice locating the ‘‘cloud’’ (as
real (model) landscapes may help students develop a represented by the user’s hand or a cardboard cutout on
sense of scale, experiment with changing reference points, the end of a stick) to get the rain to fall in a particular spot.
improve penetrative ability, and recognize a range of In addition, if students left the sandbox program (e.g., to
slopes (Birchfield and Megowan-Romanowicz, 2009; transfer files to thumb drives), when they returned, they
Liben and Titus, 2012; Atit et al., 2015). It may also prove had trouble reactivating the water features using number
to be a useful tool in helping novice users more carefully keys 1–4 that drain and add water throughout the box. A
investigate and partition complex 2D to 3D translation few comments described the software as ‘‘sensitive’’; in
tasks (Ishikawa and Kastens, 2005). actuality, the program is quite stable, although it is
Furthermore, the sandbox may support the development occasionally necessary to wiggle the mouse to get the
of model building, one of eight core practices in science projection back.
J. Geosci. Educ. 64, 199–214 (2016) Using the Augmented Reality Sandbox in Geology Labs 211

TABLE IV: Selected student responses to exit survey. engagement. In the future, we would implement pretests and
Comments on the Water Flow Feature of the AR Sandbox
posttests of introductory geology knowledge and course
and What Students Learned From It engagement, using an appropriate sample size.
 It was also really helpful to show how water reacted to
different terrains, especially w/ ocean features.
LESSONS LEARNED AND PLANS FOR
 The sandbox helped me better understand the earth’s FUTURE USE
processes because of the water features. Modeling different As mentioned previously, many potential activities could
structures and adding water to it, allowed me to visualize
how water flows in relation to landforms.
not be pursued because loose sand does not sufficiently hold
the shape of certain terrains. When asked which of the
 By creating different land formations you can get real-time sandbox exercises were the most effective, students com-
response on what the resulting effect will be. If I built a mented, ‘‘The creation of ocean features [was most effective],
mountain with a valley I can see free flow of water.
the way rivers behaved for me was just challenging to
 Specifically longshore current and how water reacts to actually make happen,’’ and ‘‘Oceans most, rivers least
certain coastal features. ‘cause [sand] wouldn’t hold together well enough to form
 I liked the feelings of clarity that I received after working river features.’’ Upgrading the sandboxes by replacing
with the sandbox, but in particular I thought the way it regular sand with more moldable sand should permit
modeled water flow was incredible! construction of many additional models and features. Other
 The sandbox helped me to understand water run-off and ideas include having a potter make clay models of
river flow. The flooding function is very accurate. It was landscapes for the box or using a 3D printer to generate
also beneficial in demonstrating shorelines. such landscapes from paper topographic maps. ECU is
What Students Liked Least About the AR Sandbox currently generating features with a 3D printer to be used for
the introductory demonstration described previously.
 Sand got into my shoes. But it’s a sacrifice I was willing to Several issues must be resolved to incorporate these
make.
activities into fall- and spring-semester labs that typically
 I kept spilling the sand. have 23 students and are taught by a GTA. GTAs sometimes
 I wish it was bigger. struggle to answer all standard topographic-map questions

that are asked, so introducing open-ended sandbox exercis-
That the technology was a little sensitive.
es, which will probably generate a lot of unpredictable
 It was very difficult to make anything other than a round questions, may be difficult for them. Therefore, to use the
hill because the dry sand doesn’t form well. sandbox successfully in introductory geology labs, GTAs will
 It might be useful if it were a bit easier to control the ‘‘rain’’ need to feel comfortable in this less prescribed setting—for
function. instance, by developing a database of potential student
 Having an updated computer attached to it, so that the questions and answers and by faculty members coaching
software wasn’t so sensitive. GTAs on how to work through questions for which they
don’t immediately have an answer. However, at the least,
the basic introductory contour line and topographic profile
LIMITATIONS TO STUDY demonstration described earlier should work well. For this
This paper presents a pilot study that tests the use of the introductory demonstration, the GTAs can be given the
AR sandbox in an introductory geology lab and summarizes detailed lesson plan and shown how to set up the sandbox.
self-reported assessments of student engagement and learn- This approach was implemented by GTAs in large labs at
ing. While the results suggest that the AR sandbox may ECU in the fall 2015 and spring 2016 semesters, and
increase student understanding of complex geologic processes anecdotal feedback from them suggests it worked well. For
and enhance spatial thinking skills, to objectively assess example, here are two comments volunteered by GTAs after
sandbox learning gains, the following limitations to the study they used the sandbox in lab:
design would need to be addressed. Student comments were
collected from only a small group of 10 students. A larger ‘‘The sand box exercise was not only personally my favorite
number of students responding in a similar pattern could give class to teach, but also for the students to learn. For my final
more validity to the positive responses received. In addition, exam extra credit question I asked, ‘What was your favorite
limited data were collected from the students (in the form of a experience in my class?’ Almost all of the students
short exit survey). Conducting follow-up interviews and focus incorporated one way or another how useful the sandbox
groups with students was difficult because the course was in a was, both visually and conceptually, understanding geologic
summer intensive session, significantly limiting time available process and topographic maps. Incorporating the sand box
for assessments and additional data collection about student into the lesson plan is by far the best idea and I cannot wait
perceptions of learning. This second limitation can be to see where this model goes into the future because its
remediated in future studies of students who take the regular potential is beyond belief. Lastly, thanks for involving me
course during the academic year by integrating the assess- with this experience, it has been very rewarding!! I love it.’’
ments and research into the course activities. In addition,
incentives can be offered to students to encourage them to be
interviewed in depth about specific ideas and concepts they ‘‘I introduced the class by guiding them through drawing
gained from interacting with the AR sandbox. Finally, because topographic lines. I used the whiteboard and they followed
the data collected were self-reported, there is no external on their own papers. After we had drawn about 6 topo lines,
objective test of the impact of the sandbox on learning and I asked them to describe to me the landscape we were looking
212 Woods et al. J. Geosci. Educ. 64, 199–214 (2016)

at. Silence. We had drawn a stream valley between two es learning about fluvial and coastal features and processes
ridges, but the most detailed responses I got were, ‘A hill.’ Or and the ability to visualize 3D landscapes from 2D maps.
‘A creek.’ Then we went to the sandbox and replicated the Another invaluable aspect of the sandbox is the opportunity
map on the whiteboard. This is where it really clicked for for students to study movement of virtual water interacting
them. They could see the connection between the topo lines with modeled landscapes.
and the actual topography. For the next ~20 minutes, I Small labs of 10 or fewer students are probably the
would draw a new topo map on the whiteboard and pairs of optimal size to make most effective use of the sandbox, and
students would do their best to replicate in the sandbox what it will be necessary to adapt exercises, which were successful
I had drawn. Their interest was captured. From there, I with small groups, to benefit larger lab sections. The ubiquity
showed them on the sandbox how to draw a topographic of camera phones is an advantage in terms of saving time
profile, then passed out practice topo profiles for them to and submitting lab exercises for grading. Retraining of
draw themselves. They performed beautifully.’’ teaching assistants will be required to improve their comfort
level with these open-ended activities and allow them to
Students in the summer 2015 lab provided several good make effective use of the technology.
suggestions for scaling up these exercises to larger classes: Students were unanimous in their perception that
‘‘It might be cool to have a question box set up, with the sandbox activities helped them understand topographic
sandbox. Maybe that way, even if all questions can’t be maps and surficial features and processes better than just
answered at the time, the GTA can address them later with studying topographic maps alone. A comment of one
the whole class,’’ and ‘‘Create a series of YouTube videos student summarizes the overall response of many:
answering possible FAQs. This could benefit an ECU
classroom as well as any other school with this technology.’’ ‘‘I felt so lucky that we were given the time and opportunity
One exercise students enjoyed was the model-building to experiment with this helpful tool. If sandboxes like the one
competition. This teaching tool was effective in the small we were exposed to were put in every public school, I feel it
summer lab (nine students), during which it was possible to would go leaps and bounds in helping students of all ages
use both the sandboxes for the same lab class. Two teams of more easily grasp some of the basic concepts of geology. Fund
students at a time were charged with modeling a specified this! The cost seems so low to construct the model, it should
terrain or process. The whole class then studied these and open up many opportunities in the educational community.
voted on which model was the most effective and correct. It And this is coming from someone who has really struggled
would probably be beneficial to expand the use of this with geology!!’’
strategy whenever multiple sandboxes are available. How-
ever, during the fall and spring semesters, ECU teaches two
introductory-geology labs at once in different rooms, so as of Acknowledgments
now, it will not be possible for a lab class to use more than The ECU authors thank the Department of Geological
one box. Another exercise worth trying in the future would Sciences for funding this project and Rob Howard for
be to choose sections from local topographic maps for the invaluable computer assistance and the donation of a camera
students to re-create in the sandbox. In addition, combining and video card. Students in these summer 2015 courses
the use of model building with Google Earth Timelapse contributed significantly with their enthusiastic participation
(which highlights imagery showing the effects of fluvial and in the exercises and their thoughtful comments. The
coastal processes over time) may help students compare following students also helped with developing exercises
their models with the actual processes. For instance, Google and making the videos: Mark Akland, Kristen Daniel, Devon
Earth Timelapse (Ferrell, 2013) portrays erosion of the Outer Reed, Caroline Smith, Luke Stevenson, and Jeremy Robbins.
Banks. Additional processes can be explored at https:// The AR sandbox was created as part of a project called
earthengine.google.org/timelapse. LakeViz3D, which is supported by the National Science
Foundation under grant 1114663. The development of the
AR sandbox software was also supported by the National
CONCLUSIONS Science Foundation under grant 1135588. The LakeViz3D-
Using instructions and software downloaded from the affiliated authors (S.R. and S.H.) gratefully acknowledge
LakeViz3D Web site (LakeViz3D, 2015a), it was possible to members and advisers of the LakeViz3D project, a collab-
construct and implement a version of the AR sandbox at a oration of scientists, science educators, evaluators, museum
teaching university with faculty and staff members who have professionals, and media developers from UC Davis TERC,
basic knowledge of carpentry, computer hardware and UC Davis KeckCAVES, UC Berkeley LHS, ECHO Leahy
software, and Linux operating systems. Then, several weeks Center for Lake Champlain, and Audience Viewpoints
of preparation were required to develop, pilot, and evaluate Consulting. Finally, the comments of reviewers and editors
which potential activities fit into the existing course structure for the Journal of Geoscience Education significantly improved
and to work through issues associated with landform this manuscript.
building and software and hardware challenges. A sandbox
demonstration lasting about 45 minutes appears to be an
extremely valuable addition to introductory geology labs on REFERENCES
topographic maps and surficial processes, especially to clarify Abrahamson, D., and Lindgren, R. 2014. Embodiment and
the concepts of contour lines, terrain steepness, Rule of Vs, embodied design. In Sawyer, R.K., ed., The Cambridge
and topographic profiles. Based on a self-reported student handbook of the learning sciences, 2nd ed. Cambridge, UK:
survey given to 10 students (Table III), students perceive that Cambridge University Press, p. 358–376.
hands-on time spent with the sandbox significantly enhanc- Alles, M., and Riggs, E.M. 2011. Developing a process model for
J. Geosci. Educ. 64, 199–214 (2016) Using the Augmented Reality Sandbox in Geology Labs 213

visual penetrative ability: Qualitative inquiry in geoscience trouble with maps and other spatial representations. Journal of
education research. Geological Society of America Special Paper, Geoscience Education, 53.2:184–197.
474:63–80. Johnson, A., Leigh, J., Morin, P., and Van Keken, P. 2006. GeoWall:
Andújar, J.M., Mejı́as, A., and Márquez, M.A. 2011. Augmented Stereoscopic visualization for geoscience research and educa-
reality for the improvement of remote laboratories: An tion. IEEE Computer Graphics and Applications, 26:10–14.
augmented remote laboratory. IEEE Transactions on Education, Kamarainen, A.M., Metcalf, S., Grotzer, T., Browne, A., Mazzuca,
54:492–500. D., Tutwiler, M.S., and Dede, C. 2013. EcoMOBILE: Integrat-
Atit, K., Gagnier, K., and Shipley, T.F. 2015. Student gestures aid ing augmented reality and probeware with environmental
penetrative thinking. Journal of Geoscience Education, 63:66–72. education field trips. Computers and Education, 68:545–556.
Augmented Reality in School Environments (ARiSE). ARiSE Keehner, M., Hegarty, M., Cohen, C., Khooshabeh, P., and
project. Available at http://www.arise-project.org/index. Montello, D.R. 2008. Spatial reasoning with external visuali-
php%3Fid=30.html (accessed 16 July 2015). zations: What matters is what you see, not whether you
Azuma, R. 1997. A survey of augmented reality. Presence: Tele- interact. Cognitive Science, 32:1099–1132.
operators and Virtual Environments, 6:355–385. Kintisch, E. 2013. Visualizing sea level rise with augmented reality.
Bacca, J., Baldiris, S., Fabregat, R., Graf, S., and Kinshuk. 2014. Available at https://www.youtube.com/watch?
Augmented reality trends in education: A systematic review of v=PeytXlQBHi8. (accessed 6 April 2016).
research and applications. Educational Technology and Society, Kreylos, O. 2008. Environment-independent VR development. In
17:133–149. Bebis, G., et al. eds., Advances in visual computing, vol. 5358.
Birchfield, D., and Megowan-Romanowicz, C. 2009. Earth Science Heidelberg, Germany: Springer-Verlag, p. 901–912, doi:10.
learning in SMALLab: A design experiment for mixed reality. 1007/978-3-540-89639-5_86.
International Journal of Computer-Supported Collaborative Learn- Kreylos, O. 2015a. Augmented reality sandbox. Available at http://
ing, 4.4:403–421. ARsandbox.org (accessed 27 June 2016).
Bryce, C.M., Baliga, V.B., De Nesnera, K.L., Fiack, D., Goetz, K., Kreylos, O. 2015b. Kinect hacking. Available at http://idav.ucdavis.
Tarjan, L.M., and Ash, D. 2016. Exploring models in the edu/~okreylos/ResDev/Kinect/ (accessed 16 July 2015).
biology classroom. American Biology Teacher, 78:35–42. Kreylos, O. 2015c. Vrui VR toolkit. Available at http://idav.ucdavis.
Cai, S., Chiang, F.K., and Wang, X. 2013. Using the augmented edu/~okreylos/ResDev/Vrui/ (accessed 16 July 2015).
reality 3D technique for a convex imaging experiment in a Kuehn, S.C. 2012. The topographic sand box, a tool for improved
physics course. International Journal of Engineering Education, understanding and visualization of topographic maps. Pre-
29:856–865. sented at the Geological Society of America Annual Meeting,
Cai, S., Wang, X., and Chiang, F.K. 2014. A case study of Charlotte, NC.
augmented reality simulation system application in a chemistry Kurganov, A., and Petrova, G. 2007. A second-order well-balanced
course. Computers in Human Behavior, 37:31–40. positivity preserving central-upwind scheme for the Saint-
Caudell, T.P., and Mizell, D.W. 1992. Augmented reality: An Venant system. Communications in Mathematical Sciences,
application of heads-up display technology to manual manu- 5:133–160. Available at http://math.tulane.edu/~kurganov/
facturing processes. In Proceedings of 1992 IEEE Hawaii Kurganov-Petrova_CMS.pdf (accessed 27 June 2016).
International Conference on Systems Sciences, p. 659–669. Lake Visualization 3D (LakeViz3D). 2015a. Complete installation
Chang, K.-T., Antes, J., and Lenzen, T. 1985. The effect of instructions. Available at http://lakeviz.org/forums/topic/
experience on reading topographic relief information: Analyses complete-installation-instructions/ (accessed 27 June 2016).
of performance and eye movements. Cartographic Journal, Lake Visualization 3D (LakeViz3D). 2015b. LakeViz3D project.
22:88–94. Available at http://lakeviz.org (accessed 1 October 2015).
Clark, D., Reynolds, S., Lemanowski, V., Stiles, T., Yasar, S., Lawrence Hall of Science. 2015. Lawrence Hall of Science,
Proctor, S., Lewis, E., Stromfors, C., and Corkins, J. 2008. University of California, Berkeley. Available at http://www.
University students’ conceptualization and interpretation of lawrencehallofscience.org/ (accessed 16 July 2015).
topographic maps. International Journal of Science Education, Liben, L.S., and Titus, S.J. 2012. The importance of spatial thinking
30:377–408. for geoscience education: Insights from the crossroads of
Duncan, R.G., and Rivet, A.E. 2013. Science learning progressions. geoscience and cognitive science. Geological Society of America
Science, 339:396–397. Special Papers, 486:51–70.
East Carolina University (ECU). 2016. Sandbox demo. Available at https:// Lindgren, R., and Johnson-Glenberg, M. 2013. Emboldened by
mediasite.ecu.edu/MS/Play/e579f009dbca41e79f0d84d7207a714a1d embodiment six precepts for research on embodied learning
(accessed 27 June 2016). and mixed reality. Educational Researcher, 42:445–452.
Ecology, Culture, History, and Opportunities (ECHO) Leahy Center Lord, T.R. 1985. Enhancing the visuo-spatial aptitude of students.
for Lake Champlain. 2015. ECHO Leahy Center for Lake Journal of Research in Science Teaching, 22:395–405.
Champlain. Available at http://www.echovermont.org/. (ac- Louca, L.T., and Zacharia, Z.C. 2012. Modeling-based learning in
cessed 16 July 2015) science education: Cognitive, metacognitive, social, material
Feldman, A., Cooke, M.L., and Ellsworth, M.S. 2010. The classroom and epistemological contributions. Educational Review, 64:471–
sandbox. Science Teacher, 77:58. 492.
Ferrell, J. 2013. Outer Banks, NC (Google TimeLapse satellite loop National Research Council (NRC). 2006. Learning to think spatially:
1984–2012). Available at http://www.youtube.com/watch? GIS as a support system in the K–12 curriculum. Washington,
v=DKUUNgMKGI8 (accessed 27 June 2016). DC: National Academies Press.
Forbes. 2016. America’s top colleges. Available at http://www. Next Generation Science Standards (NGSS) Lead States. 2013.
forbes.com/colleges/east-carolina-university/ (accessed March. Next Generation Science Standards: For states, by states.
2016) Washington, DC: National Academies Press.
Gobert, J.D., and Buckley, B.C. 2000. Introduction to model-based Niedenthal, P.M. 2007. Embodying emotion. Science, 316:1002–
teaching and learning in science education. International 1005.
Journal of Science Education, 22:891–894. Piaget, J. 1952. The origins of intelligence in children, vol. 8, no. 5.
Held, R., and Hein, A. 1963. Movement-produced stimulation in New York: International Universities Press.
the development of visually guided behavior. Journal of Rapp, D.N., Culpepper, S.A., Kirkby, K., and Morin, P. 2007.
Comparative and Physiological Psychology, 56:872. Fostering students’ comprehension of topographic maps.
Ishikawa, T., and Kastens, K.A. 2005. Why some students have Journal of Geoscience Education, 55:5–16.
214 Woods et al. J. Geosci. Educ. 64, 199–214 (2016)

Reed, S., Kreylos, O., Hsi, S., Kellogg, L., Schladow, G., Yikilmaz, framework for K–12 science education: Practices, crosscutting
M.B., Segale, H., Silverman, J., Yalowitz, S., and Sato, E. 2014. concepts, and core ideas. Washington, DC: National Acade-
Shaping watersheds exhibit: An interactive, augmented reality mies Press.
sandbox for advancing Earth Science education. In Proceedings SmartMania. 2011. Interactive Kinect sandbox. Available at https://
of the American Geophysical Union (AGU) Fall Meeting, www.youtube.com/watch?v=8p7YVqyudiE. (accessed 1 Octo-
Abstract no. ED34A-01. ber 2015).
Reed, S., Hsi, S., Kreylos, O., Yikilmaz, B., Kellogg, L., Schladow,
Tahoe Environmental Research Center (TERC). 2015. TERC,
S.G., Segale, H., and Chan, L.B. (in press). Shaping our worlds
through tangible exploration. Eos. In press. University of California–Davis. Available at: http://terc.
Reynolds, S.J., Piburn, M.D., Leedy, D.E., McAuliffe, C.M., Birk, ucdavis.edu/ (accessed 16 July 2015)
J.P., and Johnson, J.K. 2006. The hidden Earth—Interactive, W.M. Keck Center for Active Visualization in the Earth Sciences
computer-based modules for geoscience learning. Geological (KeckCAVES). 2015. KeckCAVES, University of California–
Society of America Special Papers, 413:157–170. Davis. Available at http://keckcaves.org/ (accessed 16 July
Schweingruber, H., Keller, T., and Quinn, H., eds. 2012. A 2015).

View publication stats