2008 Joint Meeting of The Geological Society of America, Soil Science Society of America, American Society of Agronomy, Crop Science Society of America, Gulf Coast Association of Geological Societies with the Gulf Coast Section of SEPM

Paper No. 21
Presentation Time: 8:00 AM-4:45 PM

The Effects of Variable Quality Data on the Accuracy of 3-Dimensional Subsurface Models for the McMaster Campus, Hamilton Ontario

MACCORMACK, Kelsey E., School of Geography and Earth Sciences, McMaster University, 1280 Main Street West, Hamilton, ON L8S 4K1, Canada and EYLES, Carolyn H., Integrated Science Program & School of Geography & Earth Sciences, McMaster University, 1280 Main Street West, Hamilton, ON L8S 4K1, Canada, maccorke@mcmaster.ca

As the demand for 3-Dimensional subsurface models has been steadily increasing over the past few years, so too has the dependency on large, readily available digital databases, such as the Ontario waterwell database, in order to produce models in relatively short time frames. However, the most common problem with relying on waterwell data as the primary data source is the variable quality of the input data. This research will show how 3-Dimensional model outputs of the McMaster campus subsurface varied with the quality of the input data and will offer a new method for integrating data from various sources while compensating for the inadequacies of the lesser quality data sets.

Borehole record quality was determined based on the level of detail in the soil descriptions and on the degree of correlation with neighboring wells. Higher quality records came from soil reports and construction reports in which the depth and characteristics of the underlying sediment were the primary focus. Lower quality records were obtained from large digital waterwell databases which typically provide general soil classifications with little accompanying description. It may be necessary to integrate both high and low quality data into a single database but this may cause a significant ‘dilution' effect on the good quality data.

To test the most effective methods of integrating databases of high and low quality, a series of 3-dimensional subsurface models were created from data available for the McMaster campus and surrounding area. Integrating outputs from models of high and low quality data sources creates an output using high quality data to constrain the stratigraphy in localized areas while using waterwell data to expand the model where data are sparse. This ‘integrated' model approach utilizes the strengths from both data sources and can be applied to enhance the quality of regional 3-dimensional modeling studies.