First of all, there are two typos in this paragraph: it should be "primitive equations" and "discussed".
That said, this passage is an abstract of a technical paper on the basics of predicting or describing the behavior of oceans using math models. Primitive equations are the series of equations used as a basis for this math model. The problem is, there's a great deal of unneccesary terminology being thrown into this abstract.
Resolution is how fine accuracy of this model is at any given point.
Parameterization is process that the person analyzing this model uses to decide which conditions are important in characterizing the behavior of the ocean.
Numerics in this case is being used in a general sense and doesn't add value to the sentence. Mathematically, you need to define your curve equations (primitive equations), your accuracy (roughly, your resolution in this case) and your assumptions (parameterization). These are all numerical, so saying "numerics" is redundant.
The second sentence talks about how fine or coarse the grid that the person analyzing this ocean model selects (or chooses) to graph his model can affect how accurately he/she can predict or define the behavior of the ocean. Again, this is an unnecessarily technical jargon-laden sentence to say: "How grid sizing affects how the curve equations are displayed (what is diplayed and what is left out).
Sorry, I'm not sure who wrote this, but this kind of poor technical writing is what gives us PhDs a bad name.