Document Type


Publication Date



Tidal currents -- Flow estimation -- Mathematical models, Storm surges, Tides


Improving methods of assessing risk and designing structures to withstand extreme events and changing sea levels is a vital component of strategies for reducing risk to coastal resources and assets. An obvious approach to improving the statistical robustness of risk assessments is to increase the number, time span, and quality of available water-level data sets, and to assess trends and non-stationarity. In this report we discuss efforts to recover, digitize, and analyze hundreds of station-years of lost-and-forgotten tide data and other water-level measurements that extend back to the early 19th century. To date, more than 6,500 station-years of previously lost or forgotten tide data have been identified, of which more than 1,600 station-years have been recovered and more than 550 station-years digitized. An additional 500+ years of once-a-day river-stage measurements in tidal rivers have been recovered and digitized. Approximately 300,000 documents have been recovered.

In this report we also demonstrate how data recovery can help engineers better characterize their water-level environment and understand long-term trends, resulting in better risk assessment and ultimately more robust design of new infrastructure and adaptation of existing infrastructure. Long records help improve hazard estimates and can help characterize whether statistics are nonstationarity over climate-relevant time scales. Further, long and more complete records exhibit more natural variability and can be used to validate numerical simulations of historical storms (e.g., the 1893 New York hurricane). Historical tide data can help inform how tidal datums such as High Water and Low Water have changed, can be used to hindcast historical river flow, and provide a much richer view of long-term trends in local sea level. Used with numerical models, historical data help explain how anthropogenic activities have changed tide range and river slope, with the largest effects often observed far inland from the coast. Moreover, numerical models combined with tide data can be used to assess whether local engineering or other long-term changes have influenced storm surge risk. In conclusion, improved assessment of the past can help us better plan for the future.

Persistent Identifier