Thanks D.D. that was written under duress haha. Yes, it has to do with the composition of the market at expiry and particularly quarterly
expirations when a big mix of contracts are changing over. That structure is held together by contracts that, as you imply, need to get closed or rolled. The market is then temporarily freed from the bonds created in the interplay between buyers and sellers of those positions. The effects transcend the day, though, as they result in a new set of possibilities going forward. This hedging matrix forms a substantial part of the total value of the largest US index, and because of the notional value and increasing use of derivatives, the influence is probably even more profound than we think.
The broad range of information in everything from retail-oriented news pieces to economics journals is overwhelming. Depending on one's starting point, it can feel like being in chemistry after having ditched algebra. In practical application you don't need much. You know how we get a feel for things over time.
@Dazedandconfused posted some good stuff on that subject recently. As for intraday volatility,
here is a relatively recent European paper on the general subject that has references and I will try to find something more to your question. Aside from the general impact of expirations, I would say that statistical data is going to be general and limited by the fluid nature of the derivatives market even if up to date academic papers offer a basic understanding.