The siege of Potidaea (480-479 BC) was an unsuccessful Persian attempt to capture the strongly fortified city in the aftermath of Xerxes's retreat from Greece, and is notable for the first historical record of a tsunami.
After the Greek naval victory at Salamis, Xerxes of Persia decided to return to Asia, leaving 300,000 men under his brother-in-law Mardonius in Thessaly to complete the conquest of Greece in the following year. Xerxes marched overland to the Hellespont, accompanied by Artabazus, son of Pharnaces with 60,000 of Mardonius's men (perhaps suggesting that Xerxes didn’t entirely trust the rest of the army). Once Artabazus had safely delivered Xerxes to the Hellespont he turned back and began the return trip to Mardonius.
On his way back Artabazus discovered that the inhabitants of Pallene, the western-most of the three peninsulas of Chalcidice, had revolted against the Persians the moment Xerxes and his army had retreated past them, having earlier provided troops and ships for his expedition. The rebels were defending the city of Potidaea, at the narrow northern end of the peninsula. Artabazus decided to lay siege to Potidaea, suggesting that the Persians weren't suffering from the serious supply problems suggested by Herodotus and other Greek sources.
The siege can be dated to the winter of 480-479 BC. The battle of Salamis took place in late September 480. After the battle Xerxes stayed in Attica for a few days, and then took 45 days to return to the Hellespont, taking us to mid-November 480 at the earlier. The Persians then had to return to Chalcidice, taking us into December at the earliest. The siege lasted for three months, so can probably be dated to December 480-February 479 or January-March 479.
Potidaea was a very difficult city to besiege. The walls stretched across the entire peninsula. Artabazus had no ships, and so was unable to move any part of his army onto Pallene. As a result the Potidaeans were able to receive supplies and reinforcements from the entire peninsula. Artabazus was also besieging the nearby city of Olynthus, which he suspected of supporting the revolt, so his forces were divided. Only after the fall of Olynthus was he able to focus all of his efforts on Potidaea.
One of the most reliable ways to capture an ancient city was treachery within the ranks. Artabazus came close to success in this way, after opening communications with Timoxenus, commander of the contingent from Scione, one of the communities of Pallene. The plot was discovered when their method of communication went wrong. They were sending each other messages wrapped around arrows that were fired at a pre-determined location. One of these arrows hit and wounded a Potidaean bystander. When the people nearby rushed to his aid, the message was discovered attached to the arrow. The message was taken to the council of leaders controlling the defence of the city, who discovered Timoxenus's guilt. The plot was foiled, although Timoxenus's role in it was kept secret for some time in order not to stigmatize the town of Scione.
After three months Herodotus reports an incident that appears to be the first historical account of a tsunami. He describes a very low tide that created shallows along the coast. The Persians decided to try and sent troops around the city using these shallows, but when they were only half way around the water came back, in the biggest tide every seen by the locals. Many Persians couldn't swim, and were drowned, while others were killed by the defenders, who took to their boats to kill the swimmers. The Potidaeans gave the credit for this to Poseidon, god of the sea and of earthquakes. In the aftermath of this disaster Artabazus abandoned the siege and returned to the main Persian camp in Thessaly. Potidaea retained its freedom, and was able to send a contingent of 300 men to the Greek league that fought at Plataea.
Herodotus's account fits very well with a tsunami. First the sea recedes, as in a very low tide, but after a short period it returns suddenly and violently, what used to be known as a tidal wave. Recent research suggests that this part of the Greece is indeed vulnerable to tsunamis, supporting Herodotus's account.