Daejeon, July 27–31
Pre-modern maps, particularly Japanese Ezu (pictorial maps), contain rich historical information regarding maritime routes, topography, and political boundaries. Unlike modern cartography, these maps are characterized by non-scale representations and artistic distortions that reflect the subjective spatial understanding of their creators. Consequently, integrating them into standard Geographic Information Systems (GIS) has traditionally been challenging.
This study presents the development of the “Digital Archive for Maritime Routes,” a system designed to manage, visualize, and compare large-scale historical maps. Theoretically grounded in the concept of “Deep Mapping” (Bodenhamer 2010), our system seeks to capture the rich narrative layers embedded in pre-modern cartography rather than treating maps solely as navigational tools.
We engage directly with the conference theme by demonstrating how digital technologies can bridge the gap between historical “memory” (archival documents) and modern spatial understanding. Specifically, the project focuses on the comparative analysis of East Asian maritime history, linking Japanese historical materials with broader regional contexts, including significant Korean historical sources.
Existing platforms such as Old Maps Online (Kowal / Přidal 2012) have made significant strides in the georeferencing and discovery of historical maps. However, these systems primarily manage maps as single primitive objects. They often lack the granularity to manage and analyze specific “place names” depicted within the maps as independent data entities linked to semantic information.
Our project addresses this limitation by adopting an approach similar to the Pelagios project (de Soto 2014), which emphasizes the semantic annotation of place references. We aim to construct a content management system (CMS) specialized for map data that not only displays images but also manages the semantic relationships between the text on the map (toponymy) and actual geographic locations.
Addressing the sub-theme of “Annotating,” we propose a data model that bridges the gap between IIIF (International Image Interoperability Framework) resources and structured geographic data. Instead of relying on “Big Data” approaches that might obscure nuances, our model focuses on the detailed interpretation of “Small Data” — specifically, over 700 curated place names found within historical maps.
Figure 1. The proposed data model linking IIIF resources with geographic semantics.
The data model consists of three interconnected layers as shown in Figure 1. The Descriptive Layer (IIIF Presentation API) manages metadata for the map images (titles, identifiers, rights) using standard IIIF Presentation API definitions. The Annotation Layer (Web Annotation) uses oa:hasSelector to define specific regions (polygons) on the historical map images, where each annotation corresponds to a semantic entity such as an island, port, or shipping route. The Geographic Semantic Layer links the image annotations to modern geographic coordinates (latitude/longitude) and external authorities. We utilize the Japan Search Utilization Schema (National Diet Library 2020) to structure this data. By assigning unique IDs to these place names and linking them to modern gazetteers such as GeoLOD (Kitamoto et al. 2024), we enable the cross-referencing of historical place names that have changed over time.
To implement this model, we utilized Drupal as the core CMS due to its flexibility and API-first architecture. We developed custom modules to generate IIIF Manifests directly from the content data. The “Geofield” module manages coordinate data, while the “Feeds” module allows for the import/export of data via CSV, facilitating data curation by historians who may not be technical experts.
For visualization and engagement, we developed a Comparative Analysis Tool using Mirador. Building upon comparison support methodologies previously developed for IIIF-based historical maps (Nakamura / Yamada 2024), we customized the Mirador viewer (v3) with a custom “sync-windows” plugin.
A unique feature of this implementation is the use of Triangulated Irregular Networks (TIN). By setting common ID points (control points) on both maps, the system uses barycentric coordinates to precisely morph and synchronize the viewport of a historical map with another historical map or a modern map. Unlike simple affine transformations, this method respects the non-linear distortions inherent in hand-drawn maps, allowing for a dynamic, geometric comparison of spatial distortions and visualizing how the “empty space” (e.g., sea routes) between landmarks was perceived differently across eras.
A focal point of this project is the comparative analysis of the Shoho Ryukyu Kuni Ezu (Pictorial Map of the Ryukyu Kingdom from the Shoho Era, 1649, Japan) and the Haedong Chegukki (1471, Korea). This comparison directly engages with the conference’s “Remembering” sub-theme by preserving and re-examining the entangled maritime history connecting Korea and Japan.
The Haedong Chegukki (Records of Countries Across the Sea to the East), compiled by the Joseon scholar-official Sin Sukju in 1471, serves as a crucial baseline for understanding pre-modern East Asian maritime relations. By comparing this Korean source with the Japanese Shoho Ryukyu Kuni Ezu (1649) using our TIN-based visualization, differences in spatial memory are revealed.
Figure 2. Comparative analysis of Yoron Island. Left: Shoho Ryukyu Kuni Ezu (1649). Right: Haedong Chegukki (1471).
As illustrated in Figure 2, the depiction of Yoron Island differs significantly. The Haedong Chegukki depicts the island primarily in the context of navigation routes and connectivity. In contrast, the Shoho Ryukyu Kuni Ezu, created under the so-called isolationist policy (sakoku) of the Edo period, emphasizes political control. It portrays the island with specific references to guard posts (Tōmiban) and detailed depictions of coral reefs acting as natural barriers to hinder foreign vessels.
This digital environment allows scholars to visually verify how “official memories” of territory and security differed between the Joseon Dynasty and the Tokugawa Shogunate, and how these memories were encoded into the landscape of the maps.
This system demonstrates that digital archives can be more than repositories; they can be active workspaces for historical inquiry. By integrating CMS capabilities with IIIF and GIS, we provide a platform where narrative, history, and geography converge. This approach not only preserves the “memory of the world” contained in these fragile documents but also provides a methodological framework for engaging with the spatial subjectivities of pre-modern East Asia.
Future work will focus on expanding the dataset and implementing the IIIF Georeference Extension to further standardize the spatial linking of these resources.