Short Answer
No. Double jeopardy of the Fifth Amendment to the United States Constitution prohibits someone from being prosecuted more than once for the same factually specific crime (i.e. "the same offense"). The Double Jeopardy Clause of the Fifth Amendment provides: “[N]or shall any person be subject for the same offence to be twice put in jeopardy of life or limb.”
The original acquittal was correct, because it was for a crime that did not happen. The later trial is for a different crime that actually did happen at a different time.
In the same way, an acquittal for a DUI committed on Tuesday does not bar prosecution of the same defendant for a different DUI committed on Thursday.
So, double jeopardy does not apply in the scenario presented.
A Similar But Harder Case
The analysis would be trickier if:
A shoots B.
A is tried and acquitted for the murder of B.
B was only presumed dead but was actually alive.
After being acquitted for Bs murder, B dies of the gunshot wound from A in (1).
A is later arrested and tried for Bs murder (a second time).
This is a much trickier question, because A was actually tried and acquitted of a crime that hadn't been completed at the time of trial, but in which A's involvement had terminated at the time.
(To make it even more complex if you are inclined, in a first degree kidnapping case, the death penalty applies unless the defendant can prove the affirmative defense that the victim didn't die, and a first degree kidnapping acquittal would probably not bar a later murder conviction because the offenses have elements sufficiently different from each other, even though the death penalty or life in prison without parole sentence for that offense is based on the conclusive presumption that the victim died in law, rather than in fact.)
Issue Preclusion
This might depend upon the nature of the evidence at trial - if A did not raise the failure of the prosecution to provide corpus delecti (i.e. a dead body) or the misidentification of an alleged victim's body as a defense and instead, for example, argued an alibi defense (e.g. he was in jail at the time of the alleged shooting). This narrow fact pattern would be a particularly close question and I wouldn't be surprised to see a court deviate from the usual precedents and general rules under these circumstances.
There is a substantial body of case law on whether prosecution for a crime with elements A, B, and C bars prosecution for a crime with different elements arising from the same facts and circumstances (e.g. if acquittal of a lesser included offense whose elements must all be proved to convict on the more severe offense provides double jeopardy protection), that wouldn't be directly applicable in my alternative scenario because the offense tried the first time and the second would have exactly the same elements.
This depends upon when a concept analogous to the principle of collateral estoppel (also known as "issue preclusion") in civil cases, in which facts previously litigated can bind a party in a later lawsuit, with or without constitutional double jeopardy dimensions, applies in criminal cases.
As a general rule, there was historically no doctrine of collateral estoppel in criminal cases as noted in a 1967 law review article. But, the U.S. Supreme Court has also adopted a limited version of the principle of collateral estoppel under the guise of the double jeopardy clause in criminal cases as it noted in Yeager (discused below) a few years after that article was written.
[I]n Ashe v. Swenson, 397 U. S. 436 (1970). . . we squarely held that
the Double Jeopardy Clause precludes the Government from relitigating
any issue that was necessarily decided by a jury’s acquittal in a
prior trial.
The hard question is determining what a jury's acquittal "necessarily decided" in particular facts and circumstances, given that a jury verdict in a criminal case typically only determines if the jury convicted, acquitted, or hung on each of the charges presented to it in the indictment and not dismissed before tiral. This analysis requires the court to assume that the jury acted rationally, even if extraneous facts allow us to know that it did not act rationally, and to determine that a prior acquittal was logically inconsistent with a new conviction.
In Yeager v. U.S. (2009), however, the U.S. Supreme Court interpreted Ash v. Swenson narrowly. It held that hung juries on counts should be ignored for constitutional purposes as if that the trial of that count never happened.
When Is A Murder Committted?
There would also be a question of when the crime of murder is committed (i.e. when the acts are taken or only once someone dies).
Generally, murder is not a complete crime until someone dies.
For example, if you were prosecuted for murder before someone died and acquitted because they were alive, you could be prosecuted again for murder later if they died of their injuries. (The more usual case is that someone is tried and convicted of assault, then the victim dies, and they are retried for murder, which is allowed since a conviction for assault is not logically inconsistent with a conviction for murder.)
Similarly, if you were convicted of murder after a fair trial and presented the live body of the victim as newly discovered evidence, there is a good chance that you could have the original conviction vacated.
If, however, you were convicted, and the Court found that you knew that the victim hadn't died at the time of trial, but you did not raise the fact that the victim wasn't dead, it isn't clear if you could have the original conviction vacated because it was a fair trial and you knew evidence sufficient to get yourself acquitted (which you may have refrained from presenting to avoid conviction on a lesser charge like kidnapping or aggravated assault), and the status of an "actual innocence" grounds for vacating a conviction after trial is hotly disputed, conservatives like the late Justice Scalia generally say "no", liberals generally say "yes", moderates like to say "yes" but make it almost impossible to establish except in rare cases like one where a live person walks in when there was a murder conviction for killing that actually living person.
Obviously this doesn't come up all that often because usually prosecutors don't bring murder cases until they find a body and identify it and aren't in a rush to do so because there is usually no statute of limitations for murder.
Conclusion In The Harder Case
My overall conclusion is that double jeopardy would not apply even in the much closer case.
This is because a belief that the victim was not dead could have been a basis for the original verdict notwithstanding a presumption to the contrary. The jury knows that a death is a element of murder and might have acquitted not based upon the reasons presented to it at trial but based upon their own personal view that the failure to present a body left them with reasonable doubt for some reason or another, which is a perfectly plausible scenario. If this happened, an acquittal the first time and conviction the second time wouldn't be inconsistent, even if no one argued about a lack of a body in their trial presentation.
Ironically, if A was acquitted of aggravated assault of B in the first trial, that would be far more likely to bar a conviction when B later dies, than an acquittal from a charge of murdering B, because murder is logically inconsistent with an acquittal of aggravated assault in most circumstances, but a previous murder acquittal based upon lack of proof of a dead body is not inconsistent with proof of a later death.
But, this would still be a close case that could come out either way on the double jeopardy issue. A court could conceivably argue that if the death of B was not contested at trial, that the first murder acquittal would be inconsistent with a murder conviction upon the death of B later on from the events that formed the basis of the first prosecution.
Tactical Considerations
Of course, even if prosecution of A for B's murder the second time was not be barred by double jeopardy, it is still likely that the prosecution would agree to lesser charges or not prosecute because the first acquittal showed it was a weak case, and the fact that A would now also be able to argue that the gunshot was not the proximate cause of B's death (and that it was instead, for example, due to medical malpractice or was a de facto suicide due to refusal of adequate treatment).
Scope Of Analysis
This reasoning would apply anywhere in the U.S. and is not specific to a particular state or territory as double jeopardy is a principle of U.S. Constitutional law that applies directly in federal courts (including the courts of territories and commonwealths) and indirectly through selective incorporation against the states via the 14th Amendment.