Historic Preservation at Rohwer WWII Relocation Center Cemetery


During the spring semester of 2012 researchers from CAST are collaborating with the University of Arkansas Landscape Architecture Department to survey and historically document the cemetery of the WWII Japanese Relocation Center located in Rohwer, Desha County, Arkansas.

Robyn Dennis, a GIS Specialist at CAST, explains the theory and technique in acquiring GPS points at the Rohwer site


This effort is being led by Landscape Architecture Assistant Professor, Kimball Erdman. Professor Erdman’s short format history placed third in the national 2011 Historic American Landscapes Survey (HALS) Challenge.

The Historic Landscape Preservation class is continuing this HALS documentation with the standard format, including an in-depth and measured drawings for the Library of Congress archives. This documentation is part of a larger conservation effort being led by Dr. Johanna Miller Lewis, the Graduate Coordinator for the Masters in Public History program at the University of Arkansas at Little Rock (UALR).  Dr. Lewis received a National Park Service (NPS) grant for the 2011 Japanese Confinement Sites Preservation Program to stabilize and restore the Rohwer Relocation Center Cemetery’s headstones and monuments.  The University of Arkansas landscape architecture class has partially matched this grant by volunteering their services for the HALS report.  In addition to the documentation and conservation efforts, UALR is working with Arkansas State University (ASU) Heritage Sites program to incorporate interpretive mapping and recordings to guide visitors through the site.  CAST is contributing to this effort with a high-density survey (HDS) of the cemetery and the surrounding site using the Z+F 5006i scanner and the Leica C10 Scan Station scanner.  In addition, the Center applied GIS methods to aid in the assessment of the current condition of the landscape and to globally locate significant features on the site.  Finally, CAST utilized photogrammetric techniques to examine very low-relief art work that was discovered on a deteriorating bench located near the gravestones.  CAST’s involvement was made possible through equipment and funding provided by the National Science Foundation (NSF) awarded CI-TRAIN project, Grant 0918970.  The data that was collected is currently being used by the University of Arkansas’ landscape architecture class to assist them in understanding and accurately documenting the site for the HALS archiving.



Established in 1942, the Rohwer Relocation Center was one of two World War II internment camps located in Arkansas.   These camps came into creation after President Franklin D. Roosevelt signed Executive Order 9066, establishing militarized, incarceration-type camps for Japanese American citizens in reaction to growing fears of a Japanese invasion.  Originally covering 500 acres and reaching over 10,000 men, women and children at its peak population, the hospital smoke stack and the cemetery that was created by the internees are all that remain of the Rohwer Camp today.  The cemetery gravestones and monuments were designed and cast in concrete by internees in the camp.  They honor those who died in the camp as well as the Japanese American soldiers from the camp who died fighting for America in WWII.  The majority of the grounds of the camp have returned to agricultural fields and the cemetery monuments are in an advanced state of deterioration.  Added to the National Register of Historic Places in 1974, and then designated as a National Historic Landmark in 1992, the cemetery is in great need of the documentation and conservation that is currently being pursued.

The cemetery was a challenge to HDS for several reasons.  The majority of significant features are gravestones, which are less than a meter in height.  These low-lying features are not ideal candidates for mid-range scanning when perpendicular angels to the scanner are preferred.  In order to overcome this obstacle while also documenting the three large monuments, scans at multiple angles and heights were conducted.  A 50-foot lift allowed researchers to capture the overall site as well as the low-lying gravestones.  This height also allowed for the capture of some of the terrain, including water features and bridges created in the original design of the cemetery.


 A second challenge involved accessing the site and time constraints.  In order to provide the class with the data early in the semester so that it could be utilized throughout the documentation project, the survey had to be conducted in January.  Wet and windy conditions resulted in increased complications for using the equipment and the 50-foot lift.  With these conditions and a four-day window for data collection, close range scanning was eliminated as an option.


 Students survey the site through traditional methods  while CAST researchers utilize a 50-foot lift to collect overall site scans


Instead, CAST researchers decided to use the Z+F 5006i scanner to document the large monuments and gravestone faces at denser point spacing and the Leica C10 Scan Station scanner for the overall site at sparser point spacing.  The increased fields of view and speed of these scanners allowed for the entire cemetery to be documented in the time allowed.  The point cloud for the two primary monuments has a point spacing of approximately 3-5 millimeters while the point cloud for the overall site has a point spacing of approximately 15-20 centimeters. Photogrammetric techniques were also used in an attempt to document the deteriorating concrete bench on the site.  The inscriptions on the bench are an example of one of the lowest levels of relief found on surfaces within the cemetery.  As this data is further processed and used, options for returning to the site to scan specific details at a higher resolution with a close range scanner and additional photogrammetry are also being considered. 


Top-down, planimetric view of the 3D scan data of the cemetery collected
with the C-10 Scanner

During this project, CAST is attempting to address a recurring objection to preservationist scanning that asserts rapid data collection creates a disconnection between the observer and the site being recorded.  For this reason, on-site traditional analysis and recordation was conducted concurrently with the use of technologies on the site.  This allowed for the class to intimately study the site while also giving input to the researchers to identify and prioritize features and elements for more in-depth recordation.  This made it possible for time to be used efficiently on site with extended investigations being carried out later in the lab.

In addition to the laser scanning and photogrammetry, Global Positioning System (GPS) Equipment was also used to survey the site.  A handheld Trimble GeoExplorer XH was used to locate the entry drive, each gravestone, the monuments, and each bollard of the original fence.  The trees and vegetation on the site were also located, identified and described with a current condition and size. These coordinates, with sub-decimeter accuracy, will be used to globally locate the scan and photogrammetric data.


Perspective view of the overall site scanning data


Since the scanning data has been collected, CAST researchers have performed basic processing including cleaning and aligning the scans.  This data, combined with the GPS locations, is now being used by the class to assist in their drawings and analysis of the site.  In accordance with current Library of Congress standards, two-dimensional drawings and photographs are required for the HALS archival documentation.  This means that the class must use a variety of techniques to transform the three-dimensional information into two-dimensional drawings.  The 3D data allows for highly accurate measurements.  It also allows for cross sections to be extracted and for volumetric and depth analyses to be performed.  These more advanced analyses, which are not possible on the real world objects, are easily performed on the point cloud data and they provide an additional level of intimate knowledge about the site being studied.  While three dimensional data provides certain advantages, the transformation from three dimensions to two dimensions presents a unique set of challenges as well.  Traditional issues such as the drawing’s projection plane and the level of detail being recorded take on a different meaning in the point cloud environment.  Decisions about whether to draw the site in 3D and then “flatten” the line work into a 2D representation introduces one set of distortion and error, while using a projection plane within the 3D environment to draw lines directly into a 2D representation introduces a different set of distortion and error.  The ability to actively reference millions of points during the drawing process raises the need for clear definitions about levels of detail and precision that are being pursued.  Understanding that error, distortion, and the need for decisions will be present and will change as projects change, CAST is attempting develop a system for tracking and recording these factors with metadata and consistency throughout the project.  This reflects a continuing area of research for the Center investigating ways that cutting edge technologies can be actively used in conjunction with traditional methods and standards to best utilize the advantages found in each.  The Center hopes that in addition to concurrent data collection on the site, continuing the discussion and the exchange between the class and the researchers will allow for a better understanding of how the three-dimensional data can enhance and increase the thoroughness of the final documentation.