Please check the pictures at the following link:
http://memg.ocean.dal.ca/bagniews/NAB_forecasts/pics_forcing/
These figures show the model output for the period of the experiment. I used forcing data for three different years: 2005, 2006 and 2007. The model uses glider and float data for T, S, Oxygen and Chl a from 4 April - 25 April.

Similar figures are shown in the links below, where I varied the initial concentrations of phytoplankton, and some biological parameters. For these figures I used forcing data for 2007.

Model output for different initial phytoplankton concentrations:
http://memg.ocean.dal.ca/bagniews/NAB_forecasts/phy/
Variations in the initial phytoplankton concentration have a large effect on the magnitude of the bloom. Note that the state variables differ only during the first 1.5 months. From June on, the initial condition doesn't matter anymore.

Model output for different values of the slope of the P-I curve (alpha):
http://memg.ocean.dal.ca/bagniews/NAB_forecasts/alpha/
You can see that varying alpha changes the time when the spring bloom occurs by about 2 weeks.

Model output for different detritus sinking rates:
http://memg.ocean.dal.ca/bagniews/NAB_forecasts/w_d/
Detritus sinking affects phytoplankton only after about 2 months, when enough detritus gets produced.

Model output for different detritus remineralization rates:
http://memg.ocean.dal.ca/bagniews/NAB_forecasts/r_d/
Similar as for variations in detritus sinking. Of course, both sinking and remineralization have a big effect on detritus.

Model output for different values of maximum phytoplankton growth rate:
http://memg.ocean.dal.ca/bagniews/NAB_forecasts/mu_0/
The choice of different growth rates makes a big difference, especially during a bloom.

Note that none of the parameter changes had a large effect on oxygen concentrations.