1 Introduction

Bluffs are geomorphological landforms with steep subaerial slopes that can rise meters to tens of meters above rivers, lakes, and oceans. Given their height and proximity to waterbodies, bluffs often have desirable aesthetics, making them attractive areas for recreational activities and housing development (Phillips 1999; Phillips 2022). Despite the allure of a bluff waterfront, these landforms offer a false sense of security to homeowners and builders because of the perceived safety of the bluff’s elevated land (Bellis et al. 1975). However, this perception of vertical separation from the hazard does not consider erosional processes such as undercutting and slumping, which act to cut the cliff back and destabilize it over time (Bellis et al. 1975). This is true in places such as North Carolina, where bluff shorelines are typically made of soft sediments, such as sand, silt, and clay, that are highly susceptible to erosion (Davidson-Arnott & Ollerhead 1995). The morphologies, or shape and character, of bluffs are dominated by erosional processes resulting from high wave energy (Davidson-Arnott 2010) and other mechanisms (Wilcock et al. 1998). This poses a hazard management challenge because the conditions that cause bluff erosion are anticipated to become more prevalent as storm frequency and intensity, sea-level rise, and human development increase and accelerate through the end of the century (Emery & Kuhn 1982). Indeed, future forecasts indicate increased risks of coastal inundation from storms caused by rising sea levels, stronger hurricane winds and heavier rains, and a rise in Category 4–5 hurricanes despite an overall decrease in tropical storms and hurricanes (NOAA 2023). Therefore, effective hazard management strategies must be built on a fundamental understanding of the underlying processes of bluff erosion (Lawrence 1994; Williams et al. 2018; Foyle 2023).

Figure 1 provides a baseline understanding of and standard reference on bluff morphology and wave-induced erosional processes and is inspired by USACE (1966), Riggs and Ames (2003) and Davidson-Arnott et al. (2010). While acknowledging the significance of subaerial processes in shaping cliff faces, such as climate factors, groundwater hydrology, and slope dynamics, our study concentrates on wave-induced erosion. For comprehensive coverage of subaerial processes, refer to Davidson-Arnott (2010). To illustrate bluff morphology, the top panel (Fig. 1A) shows a bare bluff profile (i.e., no debris at the toe), which is what residents typically observe. Two critical transition zones bound the bluff face: the crest and toe. The bluff’s crest is the edge at the top of the bluff where the slope changes abruptly between the bluff upland and bluff face. The bluff’s toe, situated at the base, marks a transition between the bluff and backshore zones to areas such as the shore platform and beach, which are visible to residents. In this area, where the slope changes from the bluff face, waves erode sediment and debris, a process that residents may witness. It is important to note that the bluff toe can also directly border the water body in cases where beach deposits are absent (see the background in the image provided in Fig. 1). During storm surges or sea-level rise, elevated water levels and marine weather conditions create wave action that propagates across the shoreface, intersecting with the bluff’s toe and face, which residents may observe. This hydrodynamic interaction causes undercutting, leading to slope failure along the bluff face (Sunamura 2015; Fig. 1B). These processes and interactions are referred to as bluff erosion and recession, phenomena that property owners are likely to notice. Bluff erosion is typically measured as the volume of material removed between the bluff crest and toe, the three-dimensional measurement along the bluff face (Davidson-Arnott 2010). Horizontal retreat or recession of the bluff face landward is measured as the distance from a reference point along the bluff face (Davidson-Arnott 2010). Building upon our understanding of bluff morphology and wave-induced erosional processes outlined in Fig. 1, our study employs distinct measurements while recognizing the interconnected nature of these processes. Therefore, we may use the terms interchangeably, recognizing that bluff recession leads to material erosion along the bluff face. Since bluffs have no recovery period to balance erosion after storms, the hazards caused by bluff erosion pose a significant risk to life and property (Fig. 1C).

Fig. 1
figure 1

Bluff erosional processes as a direct result of wave action. A This panel shows a typical bare bluff profile where a) is the upland and bluff front, b) the crest is where there is an abrupt change in downslope, c) the face is bounded by the crest and toe where we see slope failure, and d) the toe is where wave action causes erosion. The shore platform is often a narrow sandy beach. In the figure, Mean High Water is the average high-water height observed, while Mean Low Water is the average low-water height observed. B This panel shows wave action at the bluff toe, causing failure along the face. Storm surge is an unexpected rise of stormwater above the predicted tide. C This profile shows post-storm erosion of the bluff face and encroachment of the Mean High Water shoreline inland. D A ground-level photo was captured of a bluff composed of clay and sand in Craven County, North Carolina

Full size image

The risks related to bluff erosion hazards are often analyzed objectively through technical assessments of past bluff recession and erosion (e.g., Lee et al. 2001; Hall et al. 2002; Walkden & Hall 2005; Young et al. 2011; Islam et al. 2020; Swirad & Young 2021; Foyle 2023). These assessments establish a baseline understanding of historical bluff recession and erosion rates, patterns, and factors, facilitating comparison with residents' perceptions. Empirical data on bluff retreat rates and volume loss can validate residents' observations of shoreline changes. Furthermore, discrepancies between resident perceptions and technical assessments can identify knowledge gaps, guiding targeted efforts to enhance community understanding of bluff erosion processes. While technical assessments provide valuable insights into bluff erosion hazards, integrating residents' views and actions with these assessments can offer a more holistic approach to understanding coastal risk perception.

Considering that the public is directly responsible for their own shoreline protection in many cases, their knowledge and perceptions must be considered and used to understand why perception of risk may deviate from technical facts (Slovic 1987; Navarro et al. 2021). For instance, the closer we build to the bluff crest, the greater the risk of danger or property loss (Davidson-Arnott 2010; Fraser et al. 2017; Young 2018; Foyle 2023). Recent work has shown that a person’s knowledge of erosion through direct observation influences how they may or may not respond to the hazard (Arkhurst, 2022). When people remember what they have observed, they gain a more realistic perception of the risks and provide valuable feedback when managing hazards (Slovic 1986; Kasperson et al. 1988; Pace & Montz 2014). For instance, residents in North Carolina have observed decades of bluff erosion (Riggs 2001; Phillips 1999; Phillips 2022), which has likely strongly influenced their perceptions of the magnitude of the risk and whether protective action is necessary (Slovic 1987; Krewski et al. 2011; Pace & Montz 2014). However, one complication is that personal memory can become distorted over time, and they may mistakenly recall the risk as a low concern (Pace & Montz 2014; Ropeik 2022). In addition to an individual’s awareness of the hazard, other complex factors influence their perceptions, such as social variables, gender, age, and length of residence (Peacock et al. 2005; Dash & Gladwin 2007; Lindall & Hwang 2008; Pace & Montz 2014; Chadenas et al. 2023). Understanding the public's role in shoreline protection is crucial, particularly in regions like North Carolina, where coastal policies intersect with community perceptions.

All coastal policies adopted at the county level in North Carolina must conform to the generalized Coastal Area Management Act (CAMA Act) of 1974. In CAMA, bluff shorelines are considered areas of environmental concern and require a permit to build erosion control structures (N.C. Rule 15A NCAC 07H 0.1100). Based on guidance provided by the North Carolina Division of Coastal Management (DCM), which enforces CAMA regulations, suitable approaches for managing bluff shoreline erosion may include 1) land planning and regulation (e.g., setback rules; building permits) and 2) physical control structures (e.g., bulkhead; riprap revetment; seawall) (DCM 2023). As a result of CAMA regulations, many coastal counties, including Craven County, the study area for this research, have implemented initiatives to protect property owners with bluff shorelines, for instance, the use of erosion control structures in shoreline areas (Craven County 2009). Previous studies in North Carolina have effectively examined bluff shoreline change rates (e.g., Cowart et al. 2011; Eulie et al. 2017) and erosional processes (Phillips, 1999; 2022). However, these studies primarily focus on technical aspects and may not fully consider the valuable insights provided by individuals' experiences and perceptions of bluff erosion. Therefore, it is important to bridge this knowledge gap by first conducting technical assessments to evaluate historical bluff recession and erosion. These assessments can then be compared to residents' observations, understanding shoreline changes while validating residents' perceptions.

This study examines the connection between individuals’ experiences with bluff erosion and coastal risk perception in eastern North Carolina. We surveyed residents and professionals to understand their experiences and perceptions of bluff erosion processes and the factors influencing their decision to use control structures. Through the development of a GIS workflow, we analyze erosion and recession in the Neuse River Estuary, leveraging Light Detection and Ranging (LiDAR) Digital Elevation Models (DEMs) and techniques such as horizontal change assessments using the endpoint method (e.g., Fletcher et al. 2003; Cowart et al., 2011; Currin et al. 2015) and volumetric analyses using DEM of Difference approach (Wheaton et al., 2010; Sirianni et al. 2022). Moreover, by connecting shoreline changes and erosion patterns with property owners' experiences, we refined our dataset to discern the impact of individual actions on bluff erosion perceptions. This research holds significance as it examines public and professional risk perceptions and shares critical insights into decision-making processes influenced by understandings (or misconceptions) regarding bluff erosion.

2 Materials and methods

2.1 Physical setting

The North Carolina coastal region comprises ~ 23 open-ocean-facing barrier islands that cover about 320 km of shoreline known as the Outer Banks and over 4800 km of estuarine shoreline known as the Inner Banks (Crawford 2007; Gharagozlou et al. 2020; Fig. 2). It is home to the second-largest estuarine system in the United States, the Albemarle-Pamlico Estuarine System (https://apnep.nc.gov/). The major sub-estuary of this system is the Neuse River Estuary (from now on referred to as NRE), a river valley drowned by post-glacial sea-level rise flooding across the lower elevations of North Carolina’s coastal plain (Riggs 2001). The NRE is a wave-dominated estuarine system where waves > 1 m are infrequent and tidal ranges are less than 0.25 m (Phillips 1999). Unlike tide dominated coasts, wave-dominated coasts evolve through erosion, transport, and deposition of sediment associated with waves and wave-dominated currents (Davidson-Arnott, 2011). Wind plays a crucial role in this system, with wind tides significantly impacting shoreline changes within the NRE (Phillips 2022, and 1999). For instance, elevated water levels along the southern shoreline are anticipated during northwest to northeast wind events, while higher levels on the north shore are expected during southerly events. In understanding the dynamic interplay of wind, waves, and tides in the NRE, it is crucial to trace the estuary's path as it transitions into the Pamlico Sound. This transition marks a pivotal point where the NRE, though predominantly wave-dominated, encounters three major inlets – Beaufort, Ocracoke, and Oregon – connecting it to the Atlantic Ocean.

Fig. 2
figure 2

The research focused on the Neuse River Estuary in North Carolina, covering approximately 16 km of bluffs and 26.5 km of high sediment banks along the north and south shores

Full size image

The NRE transitions into the Pamlico Sound at its ~ 12.5 km wide mouth, which connects with the Atlantic Ocean through the three major inlets (see Fig. 2). In most cases, the Outer Banks block the NRE from direct contact with wave energy from the open ocean unless intense storm events disrupt normal regimes (Dalrymple et al. 1992). An example is inland breaching, where storm surges or other intense wave action breach barrier islands like the Outer Banks, allowing direct contact between the NRE and wave energy from the open ocean. Therefore, during intense storm events such as hurricanes, storm surge can propagate through inlets, affecting the NRE and exposing the areas further up the estuary to elevated wave energies that cause shoreline erosion (Clunies et al. 2017). This connection is depicted in Fig. 2, illustrating the orientation of the NRE and the major inlets that allow the interaction between the estuary and oceanic wave forces.

The development of a dataset that classifies and catalogues the spatial distribution of landforms and coastal structures throughout the NRE was used in this study to identify bluffs (~ 16 km) and high sediment banks (~ 26.5 km) (Sirianni et al. 2023; Sirianni et al. 2024). As shown in Fig. 2, the NRE’s southern shores in Craven County, which is the focus of the surveys in this study, comprise the greatest number of bluffs and development (Riggs & Aimes 2003). Those areas along the shore that do not contain bluffs and high sediment banks are widespread marshlands, which are also marked on the map in Fig. 2 as “No bluffs/No high banks.” These conditions affect the amount of development that has occurred along the north and south shorelines. The extensive marshlands on the north shore prevent extensive development, although small communities have grown where the land is higher. Development is more widespread along the south shore, particularly in the upper estuary, where the land is well-elevated above the water level, providing what is perceived to be more stable and protected locations. In understanding the existing coastal features and development patterns within the NRE, it becomes important to consider the dynamic vulnerability of these landscapes. This vulnerability comes to the forefront during intense weather events, such as hurricanes, which have had an impact on the identified bluffs, sediment banks, and the overall coastal topography over the last three decades.

Past hurricanes have severely impacted the bluffs and high sediment banks, with thirteen hurricanes passing within 60 nautical miles of the NRE over the last thirty years (between 31 December 1992 to 31 December 2022). In addition to the forward momentum, the counterclockwise rotation of hurricanes in the northern hemisphere produces winds out of the right quadrants that align perpendicularly with the orientation of the NRE’s coastline, thus allowing more wave energy through the inlets to reach the estuary. Storm timing and succession are also critical factors for bluff recession erosion in the NRE. Phillips (1999) found that Hurricane Bertha in July 1996 caused enhanced water flow, which removed much of the woody debris and sediment aprons at the base of bluffs. Two months later, Hurricane Fran generated waves that directly attacked the bare bluffs themselves, causing a bluff retreat of 3–12 m (Phillips 1999). Most recently, in September 2018, Hurricane Florence’s extended period of high-water levels and wave attacks allowed debris removal at the base of the bluffs to occur. As much as 22 m in natural bluff toe recession was observed (Phillips 2022), and a storm surge of 3.3 m was recorded in New Bern (https://www.ncdc.noaa.gov/stormevents/, accessed May 2023). While local reports document striking and accelerating morphologic changes to bluffs in association with hurricanes, the connection between individuals’ experiences with bluff erosion and coastal risk perception in eastern North Carolina has never been formally studied. In this study, we establish a baseline understanding of historical bluff and high sediment bank recession and erosion rates and patterns in the NRE. Additionally, we surveyed residents and professionals on the south shore in Craven County to understand their experiences and perceptions of bluff erosion processes and the factors influencing their decision to use control structures. By aligning erosion patterns with the experiences of property owners on the more vulnerable south shore, our goal is to offer insights into decision-making processes shaped by perceptions—whether accurate or misconceived—regarding erosion.

2.2 Surveys

We surveyed residents and professionals to understand their experiences and perceptions of bluff erosion processes and the factors influencing their decision to use control structures. The target population for the survey comprised two strata: 1) Craven County residents and 2) working professionals. Therefore, two questionnaire instruments were developed to collect data on each stratum (from now on referred to as resident and professional questionnaires, respectively). The complete surveys can be found in Sirianni and Montz (2023).

For the resident questionnaire, we employed a GIS approach to target the entire population of property owners living on or near bluffs. Property addresses were obtained as parcel or cadastral data of property boundaries and attributes from Craven County’s GIS website (https://gis.cravencountync.gov, accessed June 2023). A spatial query was conducted in ArcGIS Pro v.3.0 where the Select Layer By Location (Data Management) tool was used to select parcels based on the following two criteria: 1) their location was within 100 m of the shoreline created by Sirianni et al. (2024), and 2) it geometrically intersected the bluff crest. Because bluff tops transition into high sediment bank tops, these landforms will also be inadvertently included in some parcels based on this criterion. The query results led to 243 bluff property owners who served as our target population for the resident questionnaire. This 28-question instrument included closed (e.g., ‘tick the box’) and open-ended questions for deeper insights. Once the questionnaire was written, a pilot survey was tested with individuals who live near or regularly visit the area studied, and any necessary revisions were made (e.g., unclear or ambiguous questions or instructions). The questionnaire included a cover letter explaining the nature of the survey. Considerations such as the global pandemic of coronavirus disease 2019 (COVID-19), a rural environment, and the age of residents influenced our decision that the most appropriate type of survey technique was a paper mail-out and mail-back questionnaire. We received 82 completed resident questionnaires, providing a 33% response rate. Follow-up was considered but rejected due to the one-third response rate allowing the results to be generalized without additional expense for little gain (Williamson 2002). Questionnaires were analyzed using SPSS v.29 for descriptive statistics and Chi-Square tests.

A shorter different questionnaire targeted a sample of professionals and was designed to elicit information on where they believe bluff erosion to be the greatest and the extent to which it has affected land use decisions. This required fewer questions than those in the resident survey. We used purposive sampling to select Craven County realtors, planners, and local officials. The results were 54 professionals sent emails explaining the nature of the survey with a link to the professional questionnaire. We used Qualtrics online survey software v.2023 to distribute the 10-question instrument (see Sirianni and Montz 2023). After the initial email, three additional follow-up emails were sent. Despite these attempts, there was only a 19% response rate. Because of this, we recognize that the results may not be representative of land use professionals in the County. At the same time, the information provided by the respondents provides some insight into the effects of bluff erosion on land use and values.

2.3 Mapping past bluff top and shoreline retreat

2.3.1 LiDAR data

This study employed Light Detection and Ranging (LiDAR) Digital Elevation Models (DEMs) to analyze distance change (horizontal migration) and erosion (volumetric change) along the NRE’s bluff and high bank tops. In the Spring of 2014 and 2020, two airborne LiDAR surveys were flown over the NRE when waters were at or below normal levels (available at: https://coast.noaa.gov/dataviewer/; accessed 15 October 2023). The surveys utilized distinct sensors, Leica ALS70HP (2014) and Riegl VQ880GII (2020). Both vendors classified the LiDAR returns as ground used to generate LiDAR DEMs with horizontal resolutions dependent on the nominal pulse spacing (i.e., nominal pulse spacing multiplied by two). The 2014 LiDAR ground returns had a nominal pulse spacing of 0.7 m, where the vendor generated a 1.5 m resolution DEM. On the other hand, the 2020 LiDAR ground returns had a slightly better nominal pulse spacing of 0.5 m, which allowed for a finer 1 m resolution DEM to be created. Alignment of the two DEMs is crucial for GIS overlay, achieved using the Resample (Data Management) tool in ArcGIS Pro v.3.0. We resampled the 2020 DEM to match the cell size and alignment of the 2014 DEM (i.e., both DEMs had a 1.5 m spatial resolution where centers of input cells align). Another requirement is the matching of units and datums. No conversions were necessary for this study because both DEMs were horizontally referenced to NAD 83 (2011) using the State Plane Coordinate System, vertically referenced to the North American Vertical Datum of 1988 (hereinafter NAVD 88) using Geoid 18, and all units were in meters. While no unit or datum conversions were needed, it's essential to account for horizontal errors associated with datums, any previous transformations, and underlying LiDAR DEMs (Cooper et al. 2013; Sirianni et al. 2022).

NOAA’s vertical Datum transformation tool, VDatum, estimates the uncertainties related to datums and the transformations made between them (https://vdatum.noaa.gov/; accessed 15 November 2022). They report the transformation error from the International Terrestrial Reference Frame (ITRFxx) to NAD 83 (2011) is equal to one standard deviation of 2 cm, while the transformation between NAD83 (2011) to NAVD 88 using Geoid 18 is equal to a standard deviation of 5 cm (https://vdatum.noaa.gov/; accessed 15 November 2022). Similarly, the standard deviation associated with NAD 83 (2011) is 2 cm, and 5 cm for NAVD 88 (https://vdatum.noaa.gov/; accessed 15 November 2022). The cumulative uncertainty of the horizontal and vertical datums and the transformations made between them is, therefore 7.6 cm (\(\sqrt{{2}^{2}+{5}^{2}+{2}^{2}+{5}^{2} }=7.6\ cm\)); see the review article by Cooper et al., 2013). In addition, the LiDAR DEM vendors report a vertical Root Mean Square Error (RMSE) of 9 cm and 5 cm for the 2014 and 2020 LiDAR DEMs, respectively. Combining the largest vertical RMSE from the 2014 DEM (9 cm) and the cumulative datums’ uncertainty and transformations made between them (7.6 cm) yields a minimum cumulative vertical uncertainty of 11.4 cm (\(\sqrt{{7.6}^{2}+{9}^{2}}=11.7\ cm\)). Applying the factor of 1.96 gives the vertical linear error at the 95% confidence level (1.96 x 11.7 m = 23 cm). In other words, we are 95% confident that most vertical errors occur within the limits of ±23 cm, provided the uncertainty in the underlying data (FGDC, 1998; Gesch, 2009; Cooper et al., 2013). This vertical linear error of ±23 cm will determine the minimum volumetric change (erosion and accretion) detected in this study. 

In the past, LiDAR DEM vendors did not commonly provide information regarding their products' horizontal accuracy. For instance, only the 2020 DEM included horizontal Root Mean Square Error (RMSE) data, which was reported as 6.4 cm. Observing a notable difference in reported vertical RMSE values between the 2014 and 2020 DEMs (9 cm for 2014 and 5 cm for 2020), we assumed a similar scaling factor for the horizontal RMSEs of both datasets. To enhance the reliability of our calculations, we doubled the reported 2020 DEM's horizontal RMSE of 6.4 cm to 12.8 cm. We considered the transformation error from the International Terrestrial Reference Frame (ITRFxx) to the North American Datum of 1983 (NAD 83 (2011)), which had one standard deviation of 2 cm, along with the datum error associated with NAD83 (2011), also measured at 2 cm (NOAA 2009; Cooper et al. 2013). These values were then combined with the scaled horizontal RMSE of the 2020 DEM, yielding a cumulative horizontal uncertainty of 13.1 cm (\(\sqrt{{2}^{2}+{2}^{2}+ {12.8}^{2} }=13.1\ cm\)). The 95% confidence interval for the cumulative horizontal uncertainty was computed for this study as 22.7 cm (13.1 × 1.7308 = 22.7 cm) following FGDC (1998). Consequently, a minimum critical horizontal threshold of ± 22.7 cm was established. This threshold will determine the minimum horizontal distance change that can be detected in this study.

2.3.2 Bluff and high bank top retreat in the Neuse River Estuary

In this study, we use the terms 'bluff erosion' and 'recession' interchangeably, recognizing that bluff recession leads to the erosion of material along the bluff face. However, recession and erosion are calculated differently. Horizontal changes in the position of the bluff and high bank tops were assessed in this study using a standard endpoint method. Using the 2014 and 2020 tops generated for this study as an example, the rate of change is equal to the change in horizontal top positions divided by the change in time between the two tops, with positive values indicating accretion and negative values indicating erosion (Fletcher et al. 2003; Cowart et al., 2011; Currin et al. 2015). We calculated bluff and high bank top change rates at shore-normal transects spaced 1.5 m along the shoreline created by Sirianni et al. (2023, 2024), contingent upon the horizontal resolution of the 1.5 m DEMs. The shoreline transects were generated using the AMBUR (Analyzing Moving Boundaries Using R) package in R v4.3 (Jackson et al. 2012). The double baseline technique, specifically the "ambur.transects" function with the "near" method for transect casting, was adopted to align the transects with non-overlapping curved shoreline segments (Jackson et al. 2012). These transects were then brought into a GIS environment and superimposed with the 2014 and 2020 bluff and high sediment bank tops. To calculate the horizontal bluff and sediment bank recession, horizontal changes were calculated by subtracting the 2014 transect distance from the 2020 transect distance. Using our minimum critical horizontal threshold of 22.7 cm to define the minimum change distance, all changes in horizontal distances ± 22.7 cm were excluded from further analysis (equating to a change rate of ± 4 cm, also removed). The rate of change was calculated as the change in horizontal bluff or sediment bank tops divided by the 6-year interval. The results were a comprehensive table with attributes pertinent to the horizontal changes in NRE spanning from 2014 to 2020. These attributes encompass descriptors such as distance and rate of change in the shoreline, bluff tops, and sediment bank tops, with positive values indicating migration towards the water and negative values indicating recession towards land. The dataset is available for public download (Sirianni and Pettyjohn 2023).

2.3.3 Bluff and high bank erosion in the Neuse River Estuary

In this study, bluff and high sediment bank erosion and deposition volumes are assessed using the DEM of Difference of Difference (DoD) approach, which is used in various morphological change detection studies (e.g., Wheaton et al., 2010; James et al., 2017; Young et al. 2018; Swirad and Young 2021; Sirianni et al. 2022). The DoD involves subtracting the older DEM from a newer DEM using the following equation:

$${\text{DoD}}={DEM}_{2}-{DEM}_{1}$$
(1)

where in our case, DEM2 is the 2020 DEM and DEM1 is the 2014 DEM. Since we are assessing large-scale morphological change, and ground truth measurements are not available to estimate spatially variable DEM error (Wheaton et al., 2010), we subject the spatially uniform DEM error to the vertical linear error of ±23 cm (see Sect. 2.3). The remaining DoD grid cells were multiplied by their respective area (1.5 horizontal resolution squared = 2.25 m2), yielding a raster where each grid cell value represents the volume change between 2014 and 2020. For consistency with the analysis of horizontal shoreline change, volume grid cells are summed along transects, and the summed values are assigned to a unique identifier relative to the 2020 classified shoreline. Specifically, the transect extending from the 2014 shoreline to the 2014 bluff and high sediment bank tops represent the volume in 2014, while those extending to the 2020 tops indicate the volume as of 2020. Volume change was calculated by subtracting the 2020 volume from the 2014 volume. The rate of volumetric change was determined by dividing the sum of volumetric changes along each transect by the 6-year interval. The outcomes yielded a comprehensive table with attributes crucial for understanding volumetric changes in the NRE from 2014 to 2020. These attributes include volume (magnitude) and rate of volumetric change, where positive values indicate accretion, and negative values signify erosion. Descriptive statistics were also extracted from a sample population, with N representing the number of transects along a shoreline category minus 1 (e.g., landform, shoreline modification). The dataset is available for public download (Sirianni and Pettyjohn 2023).

2.3.4 Linking bluff and high bank top recession and erosion with surveyed respondents

In the next phase of our study, we aimed to connect changes in the shoreline, along with retreat and erosion of bluffs and high banks, to the experiences of the majority of the 243 shoreline property owners who were the main focus of our resident questionnaire. We also made sure to protect the anonymity of those participating in the survey. To achieve this, we employed the Intersect (Analysis) tool in a GIS to generate a new layer representing the intersections between these property owners' locations and the areas of bluff and high bank tops. An essential detail from the shoreline data provided by Sirianni et al. (2024) was the identification of areas subjected to human alterations from 2014 to 2020. This distinction is crucial for isolating changes attributable solely to natural processes, such as retreat or erosion intensified by the elevated water levels and surges from Hurricane Florence in 2018, from those resulting from human activities. Accordingly, we excluded any shoreline and corresponding bluff and high bank tops marked as modified during this period, as documented by Sirianni et al. (2024). The outcome was a refined dataset of bluff and high bank tops intersecting with the properties of our target respondents, devoid of any human-induced alterations between 2014 and 2020. This reduced the original 243 parcels to 195 parcels, indicating roughly 20% of the survey respondents (48 out of 243) remodified their shoreline.

3 Results

3.1 Residents’ experiences and perceptions

In our study, we explored the experiences and perspectives of residents regarding bluff erosion and the measures they have taken to mitigate it. The length of time the residents have lived in their homes—referred to as tenure—has the potential to greatly shape their views on bluff erosion risks. Most of the respondents to the resident survey live either on the bluff front or on the beach below, with more than half having lived at their current address for more than 10 years, though a quarter are relative newcomers (Table 1). Even with the differences in location and tenure, 91% responded that they have seen changes in the location of the shoreline during their time living there.

Table 1 Location by Tenure of Residential Survey Respondents
Full size table

Given a list of choices for the changes they have seen, with the option to check all, more than half indicated loss of land (62%) and bluff retreat (51%), followed closely by loss of beach (45%) and damage to structures (37%) (Fig. 3, A). The vast majority attributed the changes to storm surge and waves (85%), with the next most checked cause being continuous slow erosion over time (28%) (Fig. 3, B). A quarter of respondents checked both storm surge and waves and slow erosion over time. A number of respondents cited hurricanes, and many specifically noted Hurricane Florence, as the biggest cause for the changes they have seen.

Fig. 3
figure 3

A illustrates the observed changes as reported by participants (N = 82, multiple responses permitted). B presents the causes of changes as identified by the residents (N = 82, multiple responses permitted). C outlines the actions taken by respondents in response to these changes (N = 82, multiple responses permitted). Finally, D features photographs of various actions taken, as documented in 2020 by Sirianni et al. (2024)

Full size image

While 53% of respondents indicated that they have witnessed extensive changes in the shoreline (and 32% moderate changes), only 35% are extremely concerned with another 55% somewhat concerned. Yet, these experiences and perceptions have led most to take actions to address the problems. Looking at just those on the beach or bluff front, there is a significant Chi Square association (p = 0.003) with 93% of those on the bluff front taking action compared to 64% on the beach. Similarly, comparing those who have been at their current address for more than 10 years to those with 10 years or less, a significant Chi Square association (p = 0.001) was found with 96% of the more tenured residents having taken action compared to 68% of the newer residents. These are not surprising findings given 1) the speed at which an undercut bluff can continue to retreat and therefore threaten property and 2) the fact that those living there longer will have had more time to witness the changes. This is reflected in the actions taken, the two most common being structural responses, specifically building bulkheads and rip rap walls (Fig. 3, C-D).

There is a range of issues about which respondents are concerned, with more than 70% of them checking more than one issue. More than 80% are concerned about changes in the future with 95% being either somewhat (64%) or extremely concerned (31%). This does not vary significantly by location or tenure. Because of these anticipated changes, almost 60% believe they may have to take protective action in the future, with the most cited option being building bulkheads, followed by rip rap walls, beach nourishment, and adding a new sloped yard. At the same time, the respondents are split on their concern about climate change affecting their property with 49% saying they are concerned and 48% saying they are not.

Most (81%) said they are not considering moving given their concern about shoreline retreat, and 85% said they would rebuild if their home is significantly damaged. One who said they might not rebuild added a comment: “Due to roof loss during Florence, had to rebuild home. I'm 72 – not sure I could do it again.” Another who would rebuild noted: “Have lived here all my life, not leaving now; would rebuild but not in the same location.”

While there is recognition that protective action will be needed in the future, there is no agreement on who should take that action. Most (78%) noted that the town or county has not done anything to protect the shoreline, and 52% think that they should. Yet experiences with help for protective action are mixed as exemplified by the following two responses when asked what the government has done: “Grant money helped replace a steep slope” and “Nothing.” Many have undertaken projects at their own expense, as illustrated by: “We have an 8' seawall that we built about 15 years ago,” and “Built retention wall to hold backyard.” In other cases, assistance was provided: “Federal government brought in fill dirt/ Craven County. Not sure who paid for it.” Further, there is a perceived inequity in assistance, “They should reimburse me for doing to my yard what they did to neighbors' yards.” These and other comments reflect a strong divergence among respondents relating to what should be done and who should take responsibility. They range from a fully subsidized approach (“Install riprap to prevent erosion”) to a shared responsibility (“Subsidize proper shoreline protection; assist with homeowner efforts”) to hands off (“It's private property”). These differences in both experiences and perspectives can pose even greater complications for shoreline management than do the physical processes themselves.

3.2 Professionals’ experience and perceptions

As previously mentioned, there was a 19% response rate to a Qualtrics survey sent to professionals in Craven County. Realtors accounted for most respondents (70%), which is not surprising because there are more realtors than planners, developers or local officials. Respondents were asked to mark on a map where they think bluff erosion is the greatest. The heat map collected from this question shows the locations respondents selected on a satellite image of the NRE (Fig. 4). The color indicates the frequency of respondents’ selections, with warm colors (e.g., red) representing a high number of responses and cool colors (e.g., blue) representing a low number of responses. In some cases, the respondents selected locations that do not consist of bluff shorelines. In fact, one respondent acknowledged choosing non-bluff areas and commented: “Most of the pins I dropped are losses to marsh land. Some are developed.” This statement suggests that erosion is not seen just as a bluff issue, though that is where the losses are more evident.

Fig. 4
figure 4

Heat map where respondents were asked, “Please click on the map (up to 10 locations) where bluff erosion is the greatest.”

Full size image

Most of the professionals view the risk of erosion to be at least moderate and are concerned about it, with only 10% indicating a little risk and 20% a little concern (Fig. 5). Those who provided comments on the risk are among the most concerned, as evidenced by “Some of those places [near the pins placed] have lost as much or more than 100 yards of shoreline depth in my life” and “We have seen properties devalued and valuable resources washed away.”

When asked how much erosion has affected land use decisions, property values, property sales, and development potential, 70% reported that property values have been affected at least a moderate amount. Further, 60% indicated that land use decisions and development potential have been affected at least a moderate amount. Respondents were divided about the impact on property sales, with 50% choosing none at all or a little for the effect and 50% choosing a moderate amount or more.

As was the case with the residents, opinions on what should be done to address the problem vary, with some calling for hardening of the at-risk areas (“Bulkheading is the answer to stop the erosion in some locations. Rip-rap in others”) and others calling for living shorelines (“More research into living shorelines”). Views of responsibility also vary, with comments such as “The state should come in and repair” and what is needed is “a moratorium on allowing septic systems and building within 100 feet of the east side of the Neuse River” (a local land use decision).

3.3 Retreat and erosion in the Neuse river Estuary between 2014 and 2020

Figure 5 illustrates the median change rate in the shoreline (Fig. 5, A) and the tops of bluffs and high sediment banks (Fig. 5, B) within the Neuse River Estuary (NRE) from 2014 to 2020, a period that encompasses Hurricane Florence's impact in 2018. The figure employs a standardized color scheme to denote various changes: red indicates areas where the shoreline or the tops of bluffs and high sediment banks have retreated landward, indicating erosion. In contrast, blue signifies areas of accretion that have shifted waterward. Furthermore, Fig. 5 features bar charts for both the estuary's north and south shores, providing a quantitative analysis of the changes in the shoreline (Fig. 5, A) and bluff and high sediment bank tops (Fig. 5, B). These charts effectively highlight the differences in recession and growth (accretion) across the shores. Notably, the south shore, with its greater number of bluffs and high banks, shows a more predominant blue coloration representing accretion, suggesting high volumes of sand from eroding bluffs helped build the beach below (Fig. 5, A). Upon examining the changes in bluff and high bank tops as shown in Fig. 5, B), it is evident that these areas are undergoing retreat and erosion. From 2014 to 2020, the bluffs and high sediment banks in the Neuse River Estuary saw an overall net volume loss of about -1.74 m3 per year. Specifically, the north shore experienced a net loss of -0.29 m3 per year, whereas the south shore experienced a significantly higher net loss of -2.25 m3 per year during the same period. These results indicate that the areas on the south shore, particularly in Craven County, are more vulnerable to retreat and erosion.

Fig. 5
figure 5

A) The figure displays alterations in shoreline locations adjacent to bluff and high sediment bank tops, where areas depicted in red are those that have moved inland, indicating retreat, and areas in blue represent accretion or shifts towards the water. B details the specific changes observed in the bluff and high sediment bank tops

Full size image

3.4 Retreat and erosion relative to areas targeted in the survey

We generated distinct maps delineating the refined dataset depicting bluff and high bank tops intersecting with the properties of our target respondents, without any human-induced alterations between 2014 and 2020 (195 of the 243 parcels, see Sect. 2.3). We used a standardized color scheme to represent each attribute, as illustrated in Fig. 6. In Fig. 6(A), regions featuring a combination of bulkhead and riprap are depicted in yellow, bulkhead-only areas in orange, riprap areas in purple, and other organic features in light blue (e.g., natural bluff shorelines). Figure 6(B) highlights retreat zones in red and accretion zones in blue. Furthermore, accompanying the maps in Fig. 6 are tables providing summaries of shoreline length, retreat, and these classifications.

Among the 195 surveyed parcels that remained unmodified, 2.8 km of bluff and 2.6 km of high sediment banks front the shoreline, constituting approximately 52% and 48% of the total shoreline length, respectively (see Fig. 6 C). This inclusion arises from the inherent characteristics of the data, where bluff tops transition into high sediment bank tops, inadvertently encompassing these landforms within the parcels based on the criterion set in Sect. 2.3. The predominant modification observed along the shoreline was riprap (1.85 km), followed by bulkhead (1.37 km), with a combination of both totaling 0.64 km. Out of the total shoreline length of 5.4 km, the modified sections constitute approximately 71% of the shoreline, while 1.7 km (approximately 29%) remains categorized as other or organic (e.g., natural bluff or high sediment bank).

Figure 6 (D) reveals that bluff tops are eroding at a higher median rate of -0.59 m per year, compared to -0.19 m per year for high sediment banks. Structures combining bulkhead and riprap were the most effective against erosion from 2014 to 2020, showing a median rate of -0.14 m per year. Bulkheads had a median erosion rate of -0.29 m per year, whereas riprap alone experienced the fastest erosion rate at -0.42 m per year among modified shorelines. Excluding areas with human alterations, the dataset shows an overall net volume decrease of approximately -1.89 cubic meters per year, slightly more than the overall net loss for the Neuse River Estuary, which was -1.74 cubic meters per year (refer to Sect. 3.3). These findings suggest that, despite efforts to modify shorelines, the south shore remains prone to retreat and erosion.

Fig. 6
figure 6

This figure presents = maps showing the distribution of landform alterations and retreat. Each map is accompanied by a table summarizing their interrelationships. Section A illustrates modifications, while section B depicts retreat for the bluff and high sediment bank tops. Section C highlights the connection between landforms and modified shorelines, and section D demonstrates the relationship between modified bluff and high sediment banks and their median rate of erosion between 2014 and 2020

Full size image

4 Discussion

4.1 Addressing Bluff erosion challenges in the Neuse River Estuary

The issue of bluff erosion presents a significant challenge, marked by a range of experiences and opinions among residents and land use professionals surveyed in this study. Despite the diverse perspectives on severity and solutions, there is a unified concern underlying the need to take action to address the problem. Surveys conducted among residents reveal a consensus where 90% have observed alterations to the bluffs, primarily attributing these changes to storm surges and waves. This aligns with our initial objective to understand the community’s awareness and perceptions of bluff erosion. The quantitative bluff retreat rates from 2014 to 2020 further emphasize the vulnerability of the south shore, aligning with our goal of providing a spatial analysis of erosion.

The community's preferences for addressing bluff erosion lean towards structural measures such as bulkheads and riprap walls. This preference aligns with existing protection strategies detailed in previous research (Sirianni et al. 2024). Supporting this approach, Phillips (1999) recommends the use of low bulkheads positioned away from the beachfront, coupled with the stabilization of bluff slopes, as illustrated in Fig. 3. The Craven County CAMA Plan (Craven County 2009) also endorses the implementation of erosion control structures and the strategic relocation of at-risk structures. Following Hurricane Florence, a significant focus has been placed on slope stabilization, leading to the initiation of 42 projects. These projects are financed through the Emergency Watershed Protection Program by the North Carolina Natural Resources Conservation Service, which contributes 75% of the funding, with the remaining 25% provided by the NC Department of Agriculture (Herring 2021). The reliance on hardened structures, while popular, introduces several environmental and ecological challenges, including biodiversity loss and reduced ecosystem services (e.g., Riggs and Ames 2003; Gittman et al. 2014). In addition to these conventional methods, there is an emerging interest in the development of living shorelines as an ecologically sustainable option.

Living shorelines consisting of marsh and oyster reefs are commonly used in North Carolina’s estuaries and sounds along intertidal areas to reduce shoreline erosion (Gittman et al. 2014; Gittman et al. 2015; Smith et al. 2017; Eulie and Polk, 2018). However, the practicality of applying living shorelines to bluff and high bank shorelines warrants additional investigation. For instance, in Maine, living shoreline techniques in coastal bluff environments consist of creating vegetated step-like terraces that become less salt-tolerant with increasing elevation above the water (Maine Department of Agriculture, Conservation & Forestry 2021). While this idea is innovative, more research is needed to understand how and if similar techniques could be used in North Carolina’s NRE to combat wave-induced erosion and subaerial processes, if feasible.

Protecting property from bluff erosion and recession is both an individual and collective concern. For example, individuals are concerned with things such as maintaining the value of their property, while the collective is concerned with things such as the property tax revenue that is associated with high property values. While bluff erosion is a recognized concern amongst the population, there does not appear to be a cohesive management plan to address the problem in North Carolina. Instead, the approach is response driven rather than proactive and conducted on a parcel-by-parcel, landowner-by-landowner basis. This is not a criticism of county planning but rather a function of the complexities of the problem. In fact, property owners may act to protect their own property without necessarily considering broader implications (Siman et al. 2022). This realization aligns with our objective of understanding the complex nature of bluff erosion challenges while emphasizing the need for holistic management. There is a pressing need for holistic management strategies that account for the geomorphological, political, and economic intricacies of the region. This realization emphasizes the importance of considering broader implications beyond individual property protection.

4.2 Study limitations

As with any survey research, there are limitations that may compromise the ability to generalize the findings. Surveys do not provide as much depth to responses as do interviews, but there are two advantages to surveys: 1) the ability to get a larger number of responses within budgetary constraints and 2) the opportunity to undertake statistical analyses. Yet, an important consideration with surveys is response rates. Rates between 30 and 40% (33% for the resident survey) have been found to be typical and acceptable for mail out mail back surveys. At the same time, the 19% response rate from professionals is particularly low, putting the issue of reliability of the results in question. While these response rates may pose challenges, various strategies were employed to enhance participation, such as personalized invitations, reminders, and assurances of confidentiality. Nonetheless, the insights gained from the obtained responses provide a foundation for future research, and the steps taken to mitigate limitations contribute to the overall reliability of the survey results.

Despite the valuable insights gained from our erosion assessment and survey responses, it is essential to acknowledge certain limitations in our study. For example, the evaluation of bluff retreat rates from 2014 to 2020 heavily relies on GIS analysis using LiDAR DEMs. While these techniques are widely accepted in geomorphological studies (Fletcher et al. 2003; Cowart et al., 2011; Wheaton et al., 2010), variations in data quality, temporal resolution, or other unaccounted factors might influence the accuracy of the results. For example, changes in erosion occurring between individual storms between 2014 to 2020 and beyond this timeframe may not be fully captured, potentially limiting the comprehensiveness of our findings. Additionally, the synthesis of survey responses with spatial analyses faces challenges due to protecting the anonymity of those participating in the survey. Despite these challenges, our study adopts a systematic and transparent methodology to minimize biases and enhance the reliability of the findings. Ongoing efforts to validate the spatial analyses and survey results through additional field studies and stakeholder engagements are useful for further increasing the robustness of our conclusions and recommendations.

5 Conclusions

This study examined bluff and high sediment bank erosion, coastal risk perception, and community responses in eastern North Carolina’s Neuse River Estuary. Our aim was to understand any connections between individuals’ experiences with bluff erosion and their coastal risk perceptions. To achieve this, we surveyed residents and professionals to understand their experiences and perceptions of bluff erosion processes and the factors influencing their decision to use control structures. Employing a GIS workflow with LiDAR Digital Elevation Models, we carefully assessed erosion and retreat. The resulting refined dataset connected erosion patterns with property owners' experiences, highlighting the impact of individual actions on bluff erosion perceptions. The key conclusions of this study are as follows:

  • Recent events, as highlighted by Hurricane Florence, have shaped residents’ experiences and the Neuse River Estuary’s morphology.

  • Long-term residents (> 10 years) exhibit a preference for structural solutions (bulkheads, rip-rap), accompanied by a recent surge in the adoption of sloped yards.

  • Advocacy for exploring the synergy between structural measures and nature-based solutions, emerges as a sustainable approach.

  • There is a need to investigate the efficacy of combining structural measures with nature-based solutions for enhanced resilience.

  • Targeted risk communication strategies are essential to convey technical findings effectively to the broader community.

  • Ongoing monitoring programs to track changes in erosion dynamics and assess the effectiveness of implemented measures should be established.

Navigating the complexity of environmental processes and human responses, this study not only advances our understanding of bluff erosion but also helps to lay the foundation for informed, community-centric strategies in the dynamic coastal landscapes of eastern North Carolina.