2002 DFDC Workshop, Nottingham – abstracts

Capturing digital data in the field - geological field data capture using palm-based RDBMS and PC-based GIS

Ernst M Schetselaar, Department of Earth System Analysis, International Institute for Geoinformation Science and Earth Observation, The Netherlands

Personal digital assistants (PDA) have in the last five years become popular for the digital capture of geological field observations on the outcrop. This has not only yielded large efficiency gains in the dissemination and production of geoscientific databases and maps, but also has provided support to model and represent field knowledge in a more vigorous manner. Field observations, traditionally captured by handwriting in the form of notes and checklists, are directly captured on site and stored with their positional attributes using a relational data model. This approach promotes the consistent usage of discipline-specific terminology and the efficient transfer of direct observations to GIS environments.

Since 1995 field data capture approaches have been used at ITC to support the fieldwork components of curricula in the earth sciences, such as geological mapping, joint groundwater resource assessments, field spectroscopy and interpretation of RS datasets. In 2001 personal digital assistants (PDAs) were introduced for on-site digital field data capture. Instead of the laptop used in past exercises, palmtops were equipped with RDBMS software for direct data acquisition in the field. A relational database was maintained by synchronising daily observation to laptop GIS environments (ArcView v 3.2 and ILWIS v. 3.0) at field camps, tailored for geological database management and spatial operations.

The palm-based approach was implemented using inexpensive off-the-shelf software products (HanDbaseTM v. 2.7 and ThinkDB v 2.5). These relational database tools can be easily tailored to the needs of geological mapping and support: (1) common database field types, such as text, numerical, calculation, pop-up lists, check box, join, memo, radio button and image field, (2) extensive form design functionality, (3) one-to-many relationships up to three hierarchical levels, (4) uni- or bi-directional synchronisation for multiple users with MS Access and ODBC and (5) sketching functionality. ThinkDB, in addition, provides a plug-in for logging position from palm-connected GPS devices through NMEA protocol. The approach proved to be cost-effective and flexible in allowing fieldworkers to modify data entry forms and dictionaries as needed to accommodate for the evolving insight and understanding of the study area, while maintaining a common database structure for the mapping exercise.

The fieldworker starts by activating the 'station' form to log station number, date, time and position coordinates (x, y or x, y, z). From the 'station' form he or she can access labeled tab sheets of the 'lithology', 'minerals', 'structures', 'sample' and 'sketch' forms. The 'lithology' theme provides data structures to model the spatial relationships between lithological units. Field-observed 3-D topologic relationships, such as footwall–hanging wall relationships could be captured without the necessity to capture the detailed geometry. This provided a practical solution for developing representations of spatio-temporal reasoning mechanisms underpinning geological map interpretation and models. The approach also proved effective in supporting novel methods to model geological field data in a GIS environment. Examples here presented include: (1) the 2D representations of spatial relationships between rock fabric elements for mapping in gneissic terrain, (2) the 3D representation of rock fabrics with respect to a digital terrain model to support kinematic analysis of nappe complexes in Alpine orogenic belts (3) the automated hierarchical classification of lithological units in reconnaissance mapping and exploration projects.