I’ll let this sit a while then post the answer.
Galvanic
Nope. Galvanic corrosion would be disimilar metals in contact.
There is a definite pattern to it. it’s interesting that the corrosion is occurring on the high points at every third row (appears to be at the seam overlap). Also there is no similar patterning at the panels of the adjacent roof area.
Could it be that these panels were stacked on edge and exposed to something before they were installed? Manufacturing defect in the zinc coating at the edge the panels?
Good start, Chuck. What type of corrosion is it and how does it work?
I would almost come out and say wear and tear of ice build up sliding down the roof.
wearing the coating down
Well, I might as well jump in and guess like everyone else. :mrgreen:
A corrosive environment (for example, a swimming pool nearby).
Dirt and debris collects under the flashings and roofing overlaps and this encourages moisture which leads to corrosion.
This can occur from damp salt, dust or sand deposits catching under the ridges.
Just my guess. 
Other facts on metal corrosion;
Metal in its natural state is in the form of ore. It requires a great amount of energy to make metal from ore. Metals are actually in a temporary state. That is, at some point, they will revert back to their original state of ore. This may take several years or it could take up to hundreds or even thousands of years. The more energy required to transform an ore to metal, the faster the metal returns to its original state. Corrosion is the natural mechanism by which metal returns to its original state of ore. Corrosion of metal is an electro-chemical process -the flow of electrons from high-energy areas of metal to low energy areas through a solution on the surface of the metal, capable of supporting corrosion. Corrosion will not take place without a conducting solution. The solution is called an electrolyte, caused by water, rain, moisture, and humidity. Only a small amount of an electrolyte is needed to cause corrosion. Even 65% relative humidity will form an electrolyte, which can cause corrosion. The high-energy area of a metal is called an anode. This is the area of metal where current leaves to enter an electrolyte. The low energy area is a cathode. Current leaves the electrolyte here and returns to the metal. The number of pairs of cathodes and anodes are specific to each metal. Metals, which have more pairs, are more susceptible to corrosion. For instance, machined metal surfaces have more pairs. In some cases, corrosion can occur within minutes.
The corrosion process results in the formation of oxidation on a metal surface. If the metal layer is physically sturdy, the corrosion process can be delayed while the layer acts as a buffer between the electrolyte and the metal. If the layer is fragile or porous, the corrosion continues unhindered. For instance, in the comparison of aluminum and steel, aluminum is a high-energy metal. Atmospheric corrosion is slowed considerably, but not permanently. Steel, likewise, is a high-energy metal, however, it is a porous metal and has a fragile oxidation layer. To prevent corrosion, it requires outside assistance.
Well, if anything Kenton, I am learning in the process. :mrgreen:
I would classify it as “generalized” or “uniform corrosion.” The pictures are not clear enough to confirm “pitting corrosion,” and not enough information to confirm “atmospheric corrosion.”
I’m going along Chuck’s line. Stacked galv. steel will go down hill rather quickly when it gets moisture trapped.
It looks like rusting caused by the zinc coating having been scraped off the edges of the roof panels at some point. It exposes the metal beneath and allows it to rust.
The framing underneath could be cedar or treated lumber which will cause that.
Also debris under the lip over time will start corrosion.
There is a type of rusting that occurs after a time due to the constant wearing away of the protection along edges from expanding and contractions of the metal, the wood beneath it. The rubbing away of the coating exposes the metal. The rusting begins at the edges, then works it way across the surface once the metal is exposed. Don’t know the name of it and am too lazy to go look it up.
[FONT=Formata-Condensed][size=1][FONT=Formata-Condensed][FONT=Formata-Condensed][FONT=Formata-Condensed][FONT=Formata-Condensed][size=2]Perhaps poor installation practice. Iron debris left on the roof can rust and stained a roof. Good installation practice and post installation clean-up is important.[/size][/FONT][/FONT]
[/FONT][/FONT][/size][/FONT]
[size=2][/size]
[size=2]But most likely improper storage and/or transit of panels resulted in damage from trapped moisture. [/size]
Kenton already said that dissimilar metals was not the cause.
Metal Rust…Corrosion occurs when the protective coating or galvanization of the roof wears out and bare metal is exposed. In this case it looks like a material flaw from the manufacturer.
It’s one type of Crevice Corrosion.
This type begins during storage from moisture trapped between stacked panels.
There’s something similar to what you describe when metal roofing is laid over comp shingles with no underlayment between. With thermal cycling of the metal roof, the shingle granules sand the protective coating off the steel and corrosion begins.
I thought I could only see the corrsion along the seem where the fasteners are, and figured there was a metal frame of some sorts underneath. 
Alas i am wrong again, great info Kenton.
Crevice Corrosion;
http://www.corrosion-doctors.org/Forms-crevice/Crevice.htm:)