The Effects of Lake Breezes On Weather (ELBOW) 2001 was held in southwestern Ontario during June to August of 2001. Its purpose was to study the role of lake breezes in triggering convection and, in particular, the strong convection which can result in hail, heavy rain and tornadoes.

A mesoscale network was installed to supplement the regular surface network. On intensive observing days serial radiosondes were launched from 3 sites, mobile observations were taken and the NRC Twin Otter flew transects on a line similar to that of the radiosondes.

As part of its support to the project the Canadian Meteorological Centre provided special runs of the GEM model with a horizontal grid spacing of 2.5 km. The model was run in a non-hydrostatic mode and no convection parameterization was used. These model runs were used in conjunction with the regional GEM model and other mesoscale models to plan activities during the project.

GEM 2.5 provided useful estimates of lake breeze locations and possible interactions between lake breezes and with fronts. We will use case studies to illustrate some of the strengths and weaknesses of these forecasts and discuss some implications for severe summer weather forecasting. In particular, we will show a case in which storm outflows interacted with topographically induced boundaries resulting in "backbuilding" of the storms. Some of these storms produced heavy precipitation although no precipitation was forecast by the regional GEM model. GEM 2.5 forecast the sequence of events correctly in this case but the timing was inaccurate.