Progress in Understanding Seaclifff Evolution Using Remote Sensing Techniques and Historical Data Source

Seacliff failure and subsequent retreat impacts people, infrastructure, and public lands along high-relief coastlines around the world. The California coastline has a high abundance of seacliffs and continued chronic and catastrophic failures result in increasing vulnerability to communities in these coastal environments. Due to the episodic nature of most coastal cliff failures, understanding detailed variation in retreat processes and timing has been limited by lack of time series data; shoreline change analyses have historically used aerial photography and satellite imagery as data sources but cliff features are difficult to interpret from heretofore 2-dimensional data sources. The recent explosion of Structure from Motion (SfM) applications and data has exponentially increased our access to the type of 3-diminsional data sets required to more deeply study coastal cliff evolution, including utilizing historical aerial photographs and incorporating several decades of lidar data. This significant increase in data availability will help to vastly improve predictive models for forecasting cliff vulnerabilities to future events and conditions.

This study utilizes digital elevation models (DEMs) derived from aerial imagery using SfM techniques for five time periods between 2005 and 2019, as well as historical lidar data from 1998, to analyze a 1.5 km section of coast south of San Francisco, CA. The dataset is used to quantify multi-decadal change as well as shorter-term (years) change that incorporates a major El Nino season and a year with strong atmospheric rivers (ARs). We found that at all time scales examined, the northern half of the site consistently eroded at higher rates (~1.5 m/yr) as compared to the southern half of the section (~0.5 m/yr). Patterns of change varied during El Nino versus AR seasons, with the wave-dominated El Nino period resulting in lower amounts of retreat than during rainfall-dominated AR periods.