Dvaraka — Krishna's Sunken Capital

The genuine, mainstream-recognised archaeology at Dwarka began on land in 1963 under H.D. Sankalia, and moved offshore from 1981 under marine archaeologist Dr. S.R. Rao of India's National Institute of Oceanography, who had previously led the excavation of the Harappan port city of Lothal. Between 1983 and 1990, Rao's team, diving in the shallow waters just off the modern town, recovered stone anchors — some triangular, some rectangular, drilled with holes for rope in a style consistent with Late Bronze Age maritime technology found elsewhere around the Arabian Sea — along with wall foundations, bastions, pillars and pottery, including a distinctive lustrous red ware.

Rao and his team dated these structures to roughly 1500–1400 BCE and argued they represented genuine submerged remains of an ancient port settlement, offering, in his own assessment, evidence supporting Dwarka's historical antiquity as described in ancient texts. Thermoluminescence testing of recovered pottery has produced dates in a broadly similar range.

A separate and considerably more disputed claim emerged in 2001–2002, when a National Institute of Ocean Technology survey conducting pollution studies in the Gulf of Khambhat — roughly 120 miles from Dwarka itself — reported sonar readings interpreted by some as geometric, grid-like structures on the seabed at a depth suggesting an age of many thousands of years, far older than any conventional chronology for Indian urbanism. Supporters framed this as evidence of a submerged city potentially even older than the Indus Valley civilisation.

This "Gulf of Khambhat Cultural Complex" claim has been sharply contested by mainstream archaeologists, for reasons distinct from the Dwarka excavations proper: the disputed formations were identified via sonar and recovered largely through dredging rather than controlled excavation, making it difficult to reliably associate any dated material with the structures themselves; a piece of wood dredged from the site and radiocarbon-dated to roughly 7500 BCE has been widely rejected as evidence, since dredged material cannot be securely tied to the seafloor formations; and critics have argued the "structures" may be geofacts — naturally occurring formations that only resemble human construction — rather than genuine artifacts.