Sea-level rise threatens to increase saltwater intrusion in low-lying coastal communities worldwide. In addition, changing patterns of evaporation and precipitation are expected to modify the lengths of saltwater intrusion in estuaries. This study aims to define the length of saltwater intrusion among distinct estuaries with contrasting responses to variations in river discharge and water level. A Markov-chain Monte Carlo model was used as a supervised machine learning of the relationship among saltwater intrusion length, river discharge, water level, wind velocity and pluvial precipitation in three estuaries The supervised learning provided a relationship to project saltwater intrusion for different climatic conditions in the estuaries. The length of saltwater intrusion and river discharge were proportional for the dynamically deep Dee River estuary (UK) and Capibaribe River estuary (Brazil). Increased river discharge in those two estuaries was followed by increased limit of saltwater intrusion. For these two cases, however, it is hypothesized that beyond a threshold of river discharge, both estuaries will show a decreased saltwater intrusion with increased river discharge and transition to dynamically shallow estuaries. The dynamically shallow St. Johns River estuary (USA) had the typically expected response of decreased saltwater intrusion after increases in river discharge.
Read full abstract