When we talk about an 1D-NMR spectrum (2-dimensional graph), what does the ordinate axis show?

Normally, the ordinate (y-axis) just shows the intensity. These intensities are relative to each other, so two common things you can get from that are:

• The question, "how strong is a significant peak?" --- This will become more apparent in context. Basically, which ones would you have to identify, and which ones are too weak to be significant? This is just something to consider to save time.
• The proton splitting pattern if it's proton NMR. Look for symmetry in a splitting pattern; it follows Pascal's triangle.

If you let the magnitude of each number indicate the relative intensities, you'll have the actual pattern.

For example, a 1331 (quartet) splitting pattern indicates that a proton is "communicating" with three neighboring protons. Therefore, those three would be "communicating" back, and giving a doublet (11) splitting pattern. The 1331 pattern would have two center peaks of about identical height to each other and two outer peaks of identical height to each other, while the 11 pattern would have two peaks of about identical height, if the shimming was done correctly.

Also, if you compare the distance between each peak in the 1331 pattern to the distance between each peak in the 11 pattern, they should match. This distance is called a coupling constant (#J_(HH)# or similar). Normally these are reported in #"Hz"#, though, so if you want to convert:

#"ppm" * "MHz" = "Hz"/cancel("MHz")*cancel("MHz") = "Hz"#

Use the #"MHz"# of your NMR instrument; #300 " MHz"# is an example for a Bruker NMR, and #7~9 " Hz"# is an example of communicating protons in an aromatic environment in an ortho relationship, like on Benzene.