(Thanks anonymous wiki-owner for letting me in!)

First, make sure you get in writing that they will record and release a full unedited version. Or get in writing that you can bring in your own people to record it. Refuse to do the debate otherwise.

You should know about his shtick of observational science vs historical science. It's so stupid, that I won't do a fair job of describing it. As far as I can tell, it's based on the mistaken notion that "true science" can only be done with controlled double-blind experiments in a lab. Thus, you have to throw out all of geology, cosmology, astronomy, ... and so forth, because you cannot do a naive controlled experiment in a lab. I don't know how to approach that. Offhand, I like the metaphor of sitting on a jury in a murder case, and using the evidence to conclude facts about the past. Under Ham's contorted definitions, it's a "historical science" - the bad kind - which concluded that the guy is guilty of murder, but that might reach people that you can get rather solid evidence of past historical events through "historical science".

You should also know about Ham's purposeful misuse of the word "evolution". He defines "evolution" to include all of the following:
- big bang theory and the origin of time-space
- formation of galaxies, stars, and planets, including accretion, stellar nucleosynthesis, and the rest,
- abiogenesis
- and evolution proper
And probably more. I forgot, and I don't care to look it up. IMHO, it's rarely productive to argue over definitions, although it might be useful for the audience to simply state once that the word evolution in academia generally refers only to the process which gave us the modern diversity of life starting from a single ancestor, and note that evolution can be a proper study without explaining where the first life came from, just like astronomy can be a proper study without explaining where the big bang "came from" (or insert a better comparison if you think of it).

If you want to argue the evidence, here are some invaluable resources:

http://www.talkorigins.org/
http://wiki.ironchariots.org/index.php?title=Main_Page
http://rationalwiki.org/wiki/Evidence_against_a_recent_creation

And a shameless self-plug if only because I don't know of a better source which puts the relevant conclusions in obvious and access language.
http://freethoughtblogs.com/aronra/2013/09/17/how-bad-it-is-on-texas-textbook-review-board/#comment-41620
http://freethoughtblogs.com/aronra/2013/09/17/how-bad-it-is-on-texas-textbook-review-board/#comment-41702

If I was there, I might do this. I would do a gish gallop. I would state up front that I'm going to do this massive dump, but explain that I'm not doing that to avoid rebuttal. I'm doing it only to emphasize the sheer overwhelming evidence of many different kinds which point towards an old Earth. I would go line by line from the evidence against a recent creation, explaining how each point is independently confirmed, and how each point alone clearly and indisputably points towards an Earth older than 11,000 years (or older depending on what evidence we have). Maybe 5 minutes just on this gish gallop.

Then, I would spend a little more time on dendrochronology and distant starlight. I like these the best, because they rely on science that is accessible to the high school student. Tree rings are understood by everyone, and require no advanced math or science. Furthermore, the idea of putting together a jig-saw puzzle is understood by everyone. Because of at most one ring per year, you can show tree life is older than 11,000 years. Distant starlight is another good one. Some (young Earth) creationists will like to trot out bullshit excuses like the speed of light was different in the past, and that is why you need to be intimately familiar with SN 1987A. After explaining the problem of distant starlight, explain verbatim that "And no, the speed of light is not different in the past. How do I know this? Because it was measured." And then go into the basic high school geometry details that allow us to know the speed of light in the past thanks to SN 1987A. Maybe 5 or 10 minutes here.

Finally, you have to be prepared for the morality argument. As Matt Dillahunty says, almost every argument or debate over things like this will come down to the morality argument. While he's not liked in some circles for some of his unrelated opinions, I think that Sam Harris is an unparalleled writer and thinker in this area. Familiarize yourself with the Moral Landscape and his related work and speeches.

Here's a few bits of arguments I've cobbled together which are my favorites.

(Basically cribbed from Sam Harris:) If we ought to do anything, if morality means anything, it means that we should avoid the worst possible never-ending misery for everyone. (If you want, couch it in terms of what the audience will understand - if ought means anything, it means that we should avoid the possibility where everyone goes to hell of eternal suffering forever.) Emphasize that it would make no sense for someone to say "Oh, shoot, I had a higher priority, and I accidentally led to the outcome where everyone goes to hell". Nothing should have a higher priority than avoiding the worst possible suffering for everyone. Then forcefully argue that thus morality is about the suffering and well-being of conscious creatures. You might lose some people, but in my very humble opinion, you have to argue that morality is first and foremost about human suffering and well-being, and if the Christian god's dictates disagree, then the Christian god is immoral. Emphasize that it makes no sense to say "Well, shoot. I did what the Christian god said, even though I knew that would condemn everyone to hell."

Invariably, some Christians will disagree. They will argue that obeying the Christian god's dictates is the true "source" of morality. They won't like it, but hypothetically ask what would they do if the Christian god's dictates disagreed with human happiness and well-being. If it did, then the Christian god is evil. Anyone who says that they're willing to blindly follow the Christian god's dictates even though they full well know that it will lead to more human suffering is evil and likely immediately unreachable. Fuck them. In fact, mock them openly for being so evil and uncaring. It might help with those we can reach.

Some Christians will say that it is impossible that the Christian god's dictates disagree with human happiness and well-being. For them, ask "How do you know that? Are you an invaluable judge of morality? I thought you were fallible. Because you are fallible, you could be wrong about morality. You could be wrong, and morality could be about causing suffering in order to please god's capricious nature. God is infallible, and you are fallible, by your own admittance, right? Either you could be convinced by your god that torturing a baby purely for god's amusement is a good thing, or you admit that human happiness and well-being is the true "source" of morality."

Finally, some Christians are really hard on for a "basis" for morality. For those people, it never seems to work, but in my very humble opinion the only possible resource is as follows. You need to do a refresher on basic epistemology. When you ask someone "How do you know that?", they answer in terms of other held beliefs. They answer with a justification. A justification is an argument with premises of other held beliefs. You can model this with math graph theory, with the nodes of the graph modeling beliefs, and the edges modeling justifications. First, human knowledge is finite, and thus any purported infinite graph of justifications can be discarded out of hand. Similarly, and cycle in the directed graph is circular reasoning, and hopefully it can be agreed to discard that out of hand. From basic logic and math only, it necessarily follows that any non-empty belief graph will invariably have beliefs without justification. We call these beliefs "axioms". The belief graph forms an axiomatic belief system.

With that in place, ask them "Why do what god says? Because it's good? Why do what's good? Because it's definitional that 'good' is that which we ought to do?". Then point out the massive equivocation fallacy here. You cannot both define the word "good" to be that which is god's nature, and that which we ought to do. It can be one or the other. I know this is a mostly hopeless exercise in my experience, but it is the only correct way. Also, if you can get them to agree that we should define "good" as that which we ought to do, ask them how do they know that god is good. God could be evil. Or merely indifferent or uncaring.

At this point, the only things approaching a coherent argument I've ever heard are 1- variants of the ontological argument, and 2- because Jesus told us that he cares about human suffering and wants us all to get into heaven to avoid hell. 2 is factually false, for all of the usual historical reasons. 1 is interesting. The ontological argument is an attempt to escape Hume's is-ought distinction. It's an attempt to argue by pure logic that god is good. I hate it - I hate it so much, but if the argument reaches here, you have to address it, because it is the core of everything, but oh god is this tedious to do. It's the worst kind of word games.

What I'm trying to get at is - with Hume's is-ought distinction in place, there is no possible justification for being a decent human being who cares about the happiness of others. (Meh, for the sociopaths, you can show that this is an effective strategy for personal happiness in most situations, but this is not what most people mean by "good".) Ask them why they do what god says? Because it's good? Why do good? Why? They choose to do good. it's a completely unjustifiable choice. It's exactly the same choice I make. I
choose to do good. There is no basis. There is no justification. It's part of being a decent human being to want to do good by your fellow human being. It's having empathy. It's giving a damn. The existence or non-existence of a god does not change this in any way.

Finally, I've heard it said that "on atheism", it all ends in the eventual and unavoidable heat death of the universe, where none of my actions have tangible consequences, and thus nothing matters now. What nihilistic crap. I might like a longer life, but I'll cherish what I have. For example, when you go to a movie theatre to see a movie, you know the movie is going to end, and you know that it likely has no impact on your life afterwards. By your own argument, it doesn't matter. Yet you enjoy the movie. You enjoy the time you spend and how you spend it, even though it all gets to the same place in the end. We atheists agree. We see this life as a movie which will end, and we endeavor to make the best of it that we can. Also see above about "giving a damn" about your fellow human beings.

PS: One last thing. You need to be able to detect Pascal's Wager -like reasoning, and you need to call people on it every time it is mentioned. One of the reasons that Pascal's Wager fails is because it has the implicit premise that the Christian god is more likely than the god of Rigel 7 who made the aliens of Rigel 7 in its image. I can name one equally plausible god hypothesis for every star in the observable universe. (Oh - that star doesn't have a planet? God made the planet. Problem solved.) Each god hypothesis refers to the god which made the aliens of that planet in its image with a soul, and we humans are merely a cosmic accident, or perhaps just a minor part of god's plan to teach his chosen people a lesson in a million years time when we happen to meet in space. With a sample space that stupidly high, any and all talk about gods from first-cause arguments, or design arguments, or such, fall flatly on their face. It is by arrogance that they assume the design is for them, and not for the aliens of Rigel 7, and that is why it's a Pascal's Wager in disguise. Call them on it. Ask if they are deists, or if they are Christians. If they are Christians, then the first-cause arguments, the design arguments, and so forth, are complete non-sequiturs w.r.t. Jesus Christ.

PPS: Of course, I strongly suggest that you should not talk about (biological) evolution. it's too complex. Focus on the age of the Earth until you get consensus on that. The evidence is overwhelming, and far more understandable to the a layman.