An adjoint data assimilation approach to diagnosis of physical and biological controls on Pseudocalanus spp. in the Gulf of Maine-Georges Bank region

The underlying scientific objective here is to determine the mechanisms tha t control seasonal variations in the abundance of Pseudocalanus spp. in the Georges Bank-Gulf of Maine region. It is postulated that the observed dist ributions result from the interaction of the population dynamics with the c limatological circulation. The problem is posed mathematically as a 2-D adv ection-diffusion-reaction equation for a scalar variable. Given an initial distribution of animals, we seek the population dynamics source term R(x,y) such that integration of the forward model will result in predictions3 tha t minimize the sum of squares of differences with observed concentrations a t a later time. An adjoint data assimilation technique has been designed fo r these purposes. This approach has been used to invert for the population dynamics associate d with the transition between bimonthly (i.e. for 2 months) climatological Pseudocalanus spp. distributions derived from MAR-MAP data. Vertically aver aged velocity and diffusivity fields diagnosed from hydrodynamical simulati ons of the climatological flow are specified. Solutions converge rapidly, a nd the procedure reduces the cost function by an order of magnitude within 50 iterations. The resulting population dynamics vary considerably in space and time, as does the balance between local tendency, physical transport a nd biological source terms, Generally speaking, the patterns in population dynamics are not inconsistent with current knowledge concerning potential c ontrols such as predation and food limitation Analysis of the solutions ind icates that the Pseudocalanus spp. population centres located in the wester n Gulf of Maine and on Georges Bank may be self-sustaining, in contrast to Frier studies which characterize the former as a source region for the latt er.