Knowledge of exchange flow and its variations is important, because it determines the distribution of salt, nutrient, pollutant, and suspended sediment in estuaries. Although exchange flow has been studied extensively, their variations over the spring-neap tidal cycle remain unclear. Many studies have indicated that the exchange flow is weaker during spring tides than neap tides, but others have reported that it might be stronger during spring tides. Therefore, we investigated the spring-neap variation of an exchange flow and its cause based on intensive observational data and an analytical model applied to the Sumjin River estuary (SRE), Korea. The observations revealed that exchange flow increased during the spring tide but decreased during the neap tide. The horizontal salinity gradient increased about seven times more during spring tide than during neap tide in the middle of the estuary. The analytical model implied that the exchange flow in the SRE is mainly driven by the horizontal salinity gradient. The large horizontal salinity gradient and tidal current amplitude ratio in the SRE between the spring and neap tides suggests that the horizontal pressure gradient driven by the salinity gradient overwhelms vertical mixing in driving exchange flow in the SRE.

The horizontal salinity gradient has been reported to play a crucial role in fortnightly variability of estuarine exchange flow in short estuaries. However, spatiotemporal variations in the salinity gradient and exchange flow have not been examined over an entire short estuary, as only data observed only at specific points was available. We analyzed the variation in salinity gradient along the entire Sumjin River estuary and its effect on the exchange flow over fortnightly tidal cycles based on observations and numerical model experiments. The salinity gradient and exchange flow were in different phases between the lower and upper estuaries by 6–7 days. The maximum salinity gradient periodically reciprocated along the channel as a result of salt flux changes determined by vertical mixing. The stronger exchange flow (> 0.04 m s-1) changed location from mouth to head of estuary while the tidal range decreased, resulting from variability of the salinity gradient. The horizontal salinity gradient is large enough to overwhelm the vertical mixing effect on the exchange flow. The spatiotemporal changes of strong exchange flow correspond well with the horizontal Richardson number value (> 3). This study suggests that for the health of estuarine ecosystems, it is important to determine the spatiotemporal variation in exchange flow throughout the estuary.