Abstract

Computer-based, mathematical models that simulate spatially distributed and time dependent environmental processes are increasingly recognized to provide diagnostic and productive outputs for the assessments of environmental factors. Mathematical models have been utilized since 60's to study the water quality, the circulation structure and the seasonal changes of large water bodies such as the Great Lakes. Analysis of changes in water quality related to seasonal cycles such as the formation and the propagation of the thermal
bar requires spatial and temporal models, with high resolution. The overall objective of
this study was to integrate geo-referenced site-specific spatial data to a 3D hydrodynamic
model (ALGE) and compare the outputs with geo-referenced remotely sensed thermal imagery. The motivation of this work is based on the idea that the spatial data provides
a significant advantage of equipping the model with site-specific data to manipulate site-specific circulation patterns. Further, since the input is geo-referenced and site-specific,
the output can be directly compared to geo-referenced remotely sensed imagery for model
validation. Thus, the aspect of the integration of spatial data to hydrodynamic models is an
obvious and promising approach to study lake-wide hydrodynamic processors and seasonal
changes. A specific objective of the study was to examine how different spatial patterns
and weather conditions influence the formation and the propagation of the thermal bar
temporally and spatially. Several preliminary studies were conducted on square lakes and
lakes with false bathymetric profiles, and compared with the results of laboratory studies
and field surveys. Since the outputs compared well with the literature, bathymetric data
of Lake Ontario was integrated to the 3D hydrodynamic model with real time surface and
atmospheric weather data. The formation and the propagation of the thermal bar was
simulated for the years 1997, 1998, 1999 and 1997 (winter) through 1998(fall). The model
was finetuned by changing the initial conditions and weather data to obtain a good approx
imation to the surface temperature derived from satellite imagery using the split and dual
window technique. The lake was also subjected to different weather conditions to study
its influences on the formation and propagation of the thermal bar. The model outputs
from case studies as well as comparisons of the model output with satellite imagery and the
NOAA forecasting model are discussed.