Received: 20 November 2016 Revised: 17 April 2017 Accepted: 19 April 2017 DOI: 10.1002/gea.21633 RESEARCH ARTICLE Long-term retreat rates of Israel’s Mediterranean sea cliffs inferred from reconstruction of eroded archaeological sites Ofra Barkai 1,2,3 Oded Katz 1 Amit Mushkin 1 Beverly N. Goodman-Tchernov 3 1 Geological Survey of Israel, Jerusalem, Israel 2 Leon Recanati Institute for Maritime Studies, University of Haifa, Mt. Carmel, Israel 3 Charney School of Marine Sciences Strauss Department of Marine Geosciences University of Haifa, 199 Aba Khoushy Boulevard Mt. Carmel 31905, Israel Correspondence Beverly N. Goodman-Tchernov, Strauss Department of Marine Geosciences, Charney School of Marine Sciences, University of Haifa, 199 Aba Khoushy Boulevard Mt. Carmel 31905, Israel. Email: bgoodman@univ.haifa.ac.il Scientific editing by Christophe Morhange Funding Information Ministry of National Infrastructures, Energy and Water Resources, Israel contract 212-17-014 Abstract Erosion and inland retreat of coastal cliffs present one of the most dynamic earth-surface pro- cesses presently challenging coastlines. The rate of cliff retreat is central to coastal planning and protection of shoreline infrastructure. Until now, the majority of retreat rate estimates have been based on aerial photos from the past century, and therefore do not provide multicentury esti- mates of retreat rates. Here, we studied Bronze Age to Crusader archeological sites (∼3700 years) located on Israel’s coastal cliff and used their spatial relation to the cliff to estimate the long-term centurial-millennial retreat rates of the cliff. To accomplish this, original layouts of partially eroded archaeological structures were reconstructed and compared to the modern coastal cliff line. The eroded parts of the studied structures and their original age constrained the maximum timing of the retreat. The resulting retreat rates are significantly lower than those previously calculated using observations of around 100 years. The archeological data also record the episodicity of the cliff failure events. The research highlights both the issue of the loss of valuable archaeological cultural resources, and simultaneously the usefulness of eroding coastal archaeological features to resolve questions of modern significance. 1 INTRODUCTION Inland retreat of coastal cliffs is an ongoing natural erosion process that challenges coastal communities and infrastructure worldwide (Field, Barros, Mach, & Mastrandrea, 2014; Solomon, 2007; Sunamura, 1983; Young, Guza, Flick, O’Reilly, & Gutierrez, 2009). These chal- lenges are further highlighted by the potential acceleration of coastal erosion in light of climatic changes and projected sea-level rise (Bray & Hoke, 1997; Zhang, Douglas, & Leatherman, 2004). The fundamen- tal metrics typically used to quantify this erosion process are time- averaged cliff retreat rates, which are also commonly used to guide planning and risk management decision making. Accordingly, deter- mination of short-term (annual to decadal) versus longer term “back- ground” sea cliff retreat rates comprises a key aspect in our under- standing and management of coastal cliff erosion worldwide (e.g., Solomon, 2007). The short-term erosion of sea cliffs is readily documented through repeat observations such as those obtained by aerial photos, satel- lite images, and field/airborne surveys. However, as such data are typically not available for longer timescales, background rates of sea cliff erosion are characteristically inferred from geological, archeo- logical, and/or historical data. In this study, we focus on determining the background retreat rates of Israel’s Mediterranean sea cliff using archeological observations. Taking advantage of the unique regional abundance of historical documentation and preservation of cliff top archeological structures, we constrain the retreat rate of this sea cliff since ∼1700 B.C. The outcome of this study is a data set of long-term background retreat rates along Israel’s coastal cliff. This new data set is unique both locally and globally. Locally, it enables a better understanding of background cliff retreat rates that is required to detect current and/or future changes in sea cliff retreat rates in response to climatic changes, projected sea level rise and intensified anthropogenic effects (e.g., Klein & Zviely, 2001). More broadly, this new data set enables robust comparison between short and long-term retreat rates, which is critical to the understanding of how rates of an episodic natural system can be measured (Epifânio, Zêzere, & Neves, 2013; Fernandes & Read, 1997; Katz, & Mushkin, 2014; Sesli, Karsli, Colkesen, & Akyol, 2009). 2 BACKGROUND About one-fourth of Israel’s ∼180 km long Mediterranean coastline consists of up to 50 m high cliffs that are undergoing active erosion and retreat (Figs. 1 and 2). The cliffs are carved into a sequence of coast parallel, Nilotic, eolianite ridges that are the lithified remains of Geoarchaeology. 2017;1–14. c  2017 Wiley Periodicals, Inc. 1 wileyonlinelibrary.com/journal/gea