If there are two rockey landmasses connected only by a low lying narrow neck or isthmus of sandy soil, a tsunami striking the low sandy neck might possibly push all the sand out of the strait between the two land masses and distribute it over a wider area of sea floor.
Thus it might lower the ground level in the strait enough that it is below low tide level and even too deep to wade across at low tide.
The possibility of that does not depend on the relative or absolute sizes of the two landmasses but on there being only one connection between them which is above sea level, a connection which is both long enough to put each land mass out of sight of the other and narrow enough and low enough for the tsunami to move enough of the sand away and destroy the above sea level part of the connection.
Possibly the two land masses are a continent which is slowly splitting apart, or two continents slowly coming together. What had happened or will happen over geological eras is of very little importance to characters, even if they live in a society which knows about plate techtonics.
So the strait between the two land masses should be hundreds of miles wide in most places, but much narrower in one place. And by coincidence rivers in the two land masses flow down from heights carrying silt and have their deltas almost opposite each other. And the silt deposited in the deltas forms sandbanks above water level which get larger and larger until they join and there is one narrow path across them fron one land mass to the other.
And I suppose that a sufficiently large tsunami could scour away enough sand and silt that the connection would be lowered to at least ten feet below sea level, for hundreds or thousands of years until the two rivers build up their deltas again.
Hog Island was a mile long barrier island south of Rockaway, New York City which formed about 1863 and grew large enough to have a number of businesses and buildings on it. Most of Hog Island disappeared in a hurricane on Augyst 23, 1893, and it was damaged more in a storm in 1896. Hog island is believed to have disappeared entirely in 1902.
I believe that L. Sprague de Camp in Lost Continents: The Atlantis Theme in History, Science, and Literature discusses historic events which may have inspired Plato's account of Atlantis. I believe de Camp mentioned an island split in half by a tsunami, and a city that sank beneath the sea.
Helike was an importan city in ancient Greece that founded colonise in Asia Minor and Italy.
The city was destroyed in 373 BC, two years before the Battle of Leuctra, during a winter night. Several events were construed in retrospect as having warned of the disaster: some "immense columns of flame" appeared, and five days previously, all animals and vermin fled the city, going toward Keryneia.6 The city and a space of 12 stadia below it sank into the earth and were covered over by the sea. All the inhabitants perished without a trace, and the city was obscured from view except for a few building fragments projecting from the sea. Ten Spartan ships anchored in the harbour were dragged down with it. An attempt involving 2,000 men to recover bodies was unsuccessful. Aigion took possession of its territory.
It is believed that an earthquake produced a tsunami which flooded Helike and then receeded, but didn't receed all the way to the former shore because the rock layer beneath Helike had been titled and lowered by the earthquake. So Helike remained underwater, though in shallow enough water that for centuries tourists sailed over it to admire the ruins.
Atalanti is a small island in Greece.
Anciently, the island was known as Atalanta or Atalante (Ancient Greek: Ἀταλάντη). It was noted by ancient geographers and historians as a small island off Locris, in the Opuntian Gulf, said to have been torn asunder from the mainland by an earthquake. In the first year of the Peloponnesian War this previously uninhabited island was fortified by the Athenians to prevent Locrian pirates attacking Euboea.1 In the sixth year of the war a part of the Athenian works was destroyed by the sea, with half the ships on the beach destroyed. Thucydides reports that following an earthquake, the sea receded from the shore before returning in a huge wave.2 Citing similar events at Peparethus and Orobiae, he suggests that earthquakes and such "sea events" are linked—we now know that such tsunami are in fact caused by earthquakes. In 421 BCE, the Peace of Nicias returned Atalanta to Sparta.3 Aside from Thucydides, the island is noted by Strabo,4 Diodorus,5 Pausanias,6 Livy, Pliny the Elder, Seneca, and Stephanus of Byzantium.
Port Royal, Jamaica, was the largest city in the caribbean until destroyed in 1692.
On 7 June 1692, a devastating earthquake hit the city causing most of its northern section to be lost – and with it many of the town's houses and other buildings. Many of the forts were destroyed, as well; Fort Charles survived, but Forts James and Carlisle sank into the sea, Fort Rupert became a large region of water, and great damage was done to an area known as Morgan's Line.3
Although the earthquake hit the entire island of Jamaica, the citizens of Port Royal were at a greater risk of death due to the perilous sand, falling buildings, and the tsunami that followed. Though the local authorities tried to remove or sink all of the corpses from the water, they were unsuccessful; some simply got away from them, while others were trapped in places that were inaccessible. Improper housing, a lack of medicine or clean water, and the fact that most of the survivors were homeless led to many people dying of malignant fevers. The earthquake and tsunami killed between 1,000 and 3,000 people combined, nearly half the city's population. Disease ran rampant in the next several months, claiming an estimated 2,000 additional lives.
Some types of soil, especially when saturated with water, such as the sand in a low sandy isthumus connecting two land masses, liquify during earthquakes and flow like mud. Thus the sandy isthmus could slind into the deeper water to eaither side during an earthquake and might not need any tsunami to scour it away.
But those examples may be on a much smaller scale than necessary to sink enough land that the low land masses are out of sight of each other.
How high are two land masses where the isthmus will sink? There is a distance X that is the maximum distance that a height Y can be seen from another height Z. I don't know how to calculate it, but there are examples of mountains which can be seen from other mountains hundreds of kilometers or miles away.
On the other hand, from places only slightly above sea level, the horizon appears much closer.
I have often stood on the boardwalk of Cape May, New jersey, and seen the lights of Lewes, Delaware across Delaware Bay at night. In the daytime I have clearly seen buildings, trees, and hills on the opposite shore thorugh binoculars. And sometimes I thought that with unaided vision I glimpsed tiny bumps on the horizon in the direction of Lewes, But I couldn't be certain.
I strongly suspect that Delaware would be visible to the unaided eye from the tops of tall buildings in New Jersey, such as the lighthouse at Cape May Point, and vice versa.
Apparently the direct distance across Delaware Bay from cape May to Lewes is 16.75 miles or 26.95 kilometers.
And people who have spent time at other straits around the world can say whether one shore is visible from the other shore and at what distance.
So you need to submerge a length of low, narrow isthmus between the two land masses which is long enough that the two land masses are far enough away not be be visible from the nearest points of their shores, and also make sure that the hills and mountains on those land masses are separated by enough space that peaks on one land mass can not be seen from peaks on the other land mass. And you will need to find the formula for calculating visibility at a distance.
So you need to design the geology of your two land masses so that a giant earthquake and tsunami can separate them.