Water quality model

Dear EFDC team and users: I use the water quality model,and the calibration plots show that the result of point which is near the boundary cells is more accurate than which is far from the boundary cell. How to change this situation? (I set benthic flux.) Another problem,one of the calibration result is shown in the picture. In fact ,I set the initial concentration of total P,but in the beginning of the simulation ,the concentration of total P is very low ,very close to zero . The same is true at the end . The concentration seems very unstable. I wonder why did this happen? (This point is far from the boundary cell).