The type of fork we now are studying has two ingredients: a loose piece and a mating threat. You can begin your search by looking for either ingredient, and here looking for the second one—the mating threat—might be easier. There are no loose Black pieces, but White does have a mate threat of the type now familiar to us: his bishop attacks h7; his queen would threaten mate if it were aimed the same way, and this can be accomplished with the easy Qe4. That move also attacks the d4 bishop; too bad it isn’t loose. Or is it? Actually it's attacked once (by the White knight) and protected once (by the Black queen). So if White attacks the bishop again with his queen, he wins it. Normally, of course, Black would respond to such a threat by moving the bishop or increasing its protection. The point of the fork is that it deprives Black of time to do those things.
This problem is structurally about the same as those in the previous set, but with the wrinkle we saw earlier in this chapter: a piece attacked once and defended once—and by a piece rather than a pawn—can make as good a target as a loose piece. Consider it semi-loose; it's underdefended. The purpose of this batch of studies is to drive home the idea that pieces in such a position make fine targets. Naturally you could have started your thought process here by noticing that the bishop was vulnerable in this way and looking for ways to attack it while simultaneously threatening mate.