Shoreline retreat and beach nourishment are projected to increase in Southern California

The study addresses accelerating shoreline retreat and rising coastal erosion along semi-arid, low-lying urban coasts, focusing on Southern California's Gulf of Santa Catalina. Sea-level rise, increasing aridity, altered precipitation patterns, and extensive urban growth have intensified sediment imbalances, reducing natural sediment supply to beaches and driving erosion. The research posits that this region is representative of global semi-arid sandy coasts undergoing similar hydroclimatic fluctuations and anthropogenic pressures. The primary objective is to quantify and forecast shoreline retreat rates and to estimate the sand nourishment volumes and associated costs required to mitigate retreat through 2050, 2075, and 2100. By combining multi-decadal photogrammetric shoreline analyses with DSAS-derived change metrics, SLR-induced retreat estimates, and land-use/population dynamics, the study aims to inform sustainable coastal management and adaptation strategies under increasing urbanization and climate-driven extremes.

Prior work links shoreline retreat on sandy beaches to both sea-level rise and long-term sediment budget imbalances. Leatherman et al. and Zhang et al. emphasized the combined roles of SLR and sediment deficits in driving erosion. The Global Scale Assessment Model (GSAM) provided global insights for ambient sandy beaches, but arid and semi-arid systems with complex sedimentation (e.g., Gulf of Santa Catalina) were underrepresented due to data limitations. Southern California has documented retreat during strong El Niño events (e.g., 2015–2016), highlighting storm-driven erosion vulnerability. Beach nourishment has been widely applied as a soft adaptation measure in the US, Netherlands, Australia, China, and the Mediterranean, with recognized economic co-benefits but also ecological concerns (turbidity, habitat alteration) and uncertainties over long-term impacts. The Bruun rule is extensively used to approximate SLR-driven beach profile adjustment and shoreline recession, though its assumptions and applicability have been debated; validations exist under specific morphological and hydrodynamic conditions along US and Mediterranean coasts. Regional studies report reduced fluvial sand supply to Southern California beaches due to damming and shoreline armoring, exacerbating sediment deficits and erosion. Collectively, the literature underscores the need to integrate human-induced sediment supply changes with SLR to assess and forecast shoreline evolution in semi-arid urban coasts.

Study area and validation: The study area spans the 150 km Gulf of Santa Catalina (Southern California) across the San Pedro (≈70 km) and Oceanside (≈80 km) littoral cells. Two validation sites with comparable semi-arid conditions and urbanization were analyzed: Corona del Mar State Beach (CA, USA) and Hammamet North Beach (Gulf of Hammamet, Tunisia). Both feature pocket and embayed sandy beaches backed by cliffs and are influenced by reduced fluvial sediment supply due to dams and coastal engineering. Datasets: Over 100 photogrammetric datasets were used, including aerial photography (1992, 2005, 2018) and orbital imagery (Landsat 1–5 MSS, Landsat 5 TM, Landsat 8 OLI, Sentinel-2B MSI). Shoreline positions were extracted for 1992, 2005, and 2018. High-resolution Global Surface Water Explorer (GSWE) time series (1984–2018) informed water occurrence and shoreline change. Land-use classifications used Landsat scenes from 1985 and 2015. Shoreline change analysis (DSAS): The Digital Shoreline Analysis System generated 1600 cross-shore transects at 50 m spacing (135 km in SC; 500 m in validation sites). DSAS computed Net Shoreline Movement (NSM), End Point Rate (EPR), and Linear Regression Rate (LRR) for 1992–2005 and 2005–2018, producing observed shoreline retreat rates (SRR_observed). Field validation occurred in 2018 (HAM) and 2022 (Corona del Mar, Huntington Beach). The DSAS-derived index of agreement for predictions was 0.9 over 1992–2018 calibration. SLR-induced retreat (Bruun rule): SLR-driven SRR (SRR_SLR) was computed via SRR = (l / tanβ) × SLR, using nearshore slopes (tanβ) from global datasets (Athanasiou et al.) and tide-gauge SLR trends and uncertainties (San Diego, Mediterranean). This 2D equilibrium approach assumes quasi-constant wave climate and does not incorporate alongshore sediment exchanges. Comparison of DSAS and Bruun results isolates the coastal-erosion (sediment imbalance) component: %SRR(coastal erosion) = (SRR_observed − SRR_SLR) / SRR_observed × 100. GSWE-derived shoreline metrics: GSWE annual water occurrence was processed in Google Earth Engine and ArcGIS Pro to estimate shoreline movement rates (with ~6% error) for comparison against DSAS and Bruun outputs. Land–water interface and land-use classification: Coastlines were delineated using a water index (Green–NIR)/(Green+NIR) to differentiate land/water. Terrain classes (continental water, bare soils, vegetation, urban spaces, sandy deposits) were produced using ISODATA unsupervised methods and refined with supervised Maximum Likelihood classification where needed. Accuracy was assessed via confusion matrices, achieving Kappa ≈0.89–0.93 (1985) and 0.94–0.95 (2015); overall accuracy ≈96–98%. Reference data included USGS maps, Google Earth imagery, municipal land-use plans, and GPS field surveys. Forecasting model linking SRR to population density: A conceptual, linear model relates SRR trends to Population Density growth Rate (PDR), using LRR-based extrapolation from 1992–2018 to project SRR and PDR to 2050, 2075, and 2100 under three scenarios (minimum, average, maximum), with error propagation from DSAS co-registration and EPR uncertainties. Nourishment volume and cost estimation: Annual nourishment volumes per km were estimated from projected SRR for Corona del Mar and scaled to littoral cells. Costs used a baseline unit sand price of $20/m³ (2018) with an assumed average 3% annual increase (implying ≈$45/m³ by 2050). Forecast horizons emphasized 2050 (≈32 years) with 2075 provided illustratively. Historical nourishment volumes and costs for Southern California programs contextualized results. All monetary values for future years are expressed in 2018 USD. Data sources and processing details are provided in Table 5 and the OSF repository.

- Observed retreat (1992–2018): Southern California sandy beaches show average retreat near -1 m/yr, with EPR values ranging roughly from -0.75 to -1.24 m/yr at hotspots. Validation sites show severe erosion: Corona del Mar NSM ≈ -35 m; Hammamet North NSM ≈ -90 m over 1992–2018. In HAM, EPR ranges -4.49 to -0.21 m/yr, averaging about -2.35 m/yr. - Semi-arid benchmark rates: Combining the study area and validation sites yields a current semi-arid average shoreline retreat of about -1.45 m/yr. - Attribution of retreat: DSAS-observed SRR exceeds Bruun SLR-only estimates by 12–19% in Southern California and 43–75% in Hammamet (1992–2018), indicating substantial contributions from sediment imbalance due to human and hydroclimatic drivers. Example (2018): Corona del Mar SRR_SLR ≈ -0.67 m/yr vs DSAS ≈ -0.83 m/yr (human-driven ≈ -0.16 m/yr, 19%); HAM SRR_SLR ≈ -1.4 m/yr vs DSAS ≈ -2.45 m/yr (human-driven ≈ -1.05 m/yr, 43%). GSWE-derived retreat rates (-0.76 m/yr in SC; -1.92 m/yr in HAM) are consistent with Bruun within processing errors. - Drivers: Aridity and land-use change have reduced fluvial sediment supply (e.g., damming, flood control, cliff armoring, harbor jetties). Precipitation declines since 1980–2018 were ~25% (Southern California) and ~56% (Hammamet). In SC, damming has obstructed about 47% of natural fluvial sand yield (~1.5 million m³/yr), and cliff armoring further reduces supply (~26,750 m³/yr). Urban areas expanded markedly at the expense of vegetation (1985–2015), corroborated by high-accuracy land-use classifications. - Projections of retreat: Average total retreat rate across semi-arid coasts is projected to increase from ~-1.45 m/yr (present) to -2.12 m/yr (2050) and -3.18 m/yr (2100). Site-specific scenario projections (Table 2) show Corona del Mar average scenario: -1.09 m/yr (2050), -1.85 m/yr (2100); Hammamet average scenario: -3.16 m/yr (2050), -4.5 m/yr (2100). - Nourishment volumes and costs: Present nourishment need is ~1223 m³/yr/km to hold equilibrium at Corona del Mar; study-wide forecasts suggest this could triple to ~3669 m³/yr/km by 2050. For Corona del Mar scenario 2, volume doubles to ~2446 m³/yr/km by 2050, with annual cost rising from ~$24,460 to ~$125,973 per km (2018 USD). Costs could increase about fivefold by 2050 given unit sand price escalation. For the San Pedro littoral cell, current annual nourishment is ~305,822 m³/yr (at ~-0.5 m/yr), with projected 2050 expenditures around $16 million/yr (lower-bound estimate, excluding additional structural/logistical needs). - Comparative context: Projected NSM for semi-arid urban beaches by 2100 exceeds GSAM values for ambient sandy beaches, implying semi-arid coasts may retreat faster than globally averaged ambient beaches.

The findings confirm that shoreline retreat on semi-arid, urbanized sandy coasts is accelerating and is strongly influenced by human-driven sediment imbalances superimposed on sea-level rise. By explicitly deconvolving SLR-driven and erosion-driven components, the study shows that land-use change, damming, and coastal engineering have substantially reduced sediment supply, amplifying retreat beyond that predicted by SLR-only models. The Gulf of Santa Catalina, hypothesized as representative of semi-arid sandy coasts, exhibits dynamics that align with other regions experiencing aridification and rapid urban growth (e.g., North Africa), supporting the broader relevance of the results. Economically, the required scale and frequency of nourishment to maintain beach equilibrium will rise substantially under projected retreat rates, with unit costs expected to escalate. This will challenge coastal communities' budgets and logistics, particularly in developing semi-arid nations where resources are constrained. Environmentally, increased nourishment may have cumulative ecosystem impacts that remain insufficiently quantified. The integration of DSAS observations, Bruun-based SLR attribution, GSWE water dynamics, and land-use change provides a replicable framework for other semi-arid coasts to anticipate retreat and plan adaptation. The results underscore the need for comprehensive sediment management strategies, reassessment of hard structures that disrupt longshore drift, and proactive planning for combined climate and urbanization pressures.

Monitoring of shoreline positions (1992–2018) using orbital photogrammetry and DSAS in Southern California reveals significant sandy beach retreat rates of about -0.75 to -1.24 m/yr (≈-1 m/yr on average), with similar or higher rates at a semi-arid validation site in Tunisia (-0.21 to -4.49 m/yr; ≈-2.35 m/yr average). The current semi-arid benchmark across both sites is ~-1.45 m/yr, far exceeding the GSAM global average for ambient sandy beaches. Forecasts indicate that population-driven land-use change and sediment imbalance will increase average total retreat rates from ~-1.45 m/yr today to -2.12 m/yr by 2050 and -3.18 m/yr by 2100, compounding SLR effects. Consequently, intensive and costly nourishment will increasingly be necessary. Estimated yearly nourishment demand could rise from ~1223 to ~3669 m³/yr/km by 2050, with associated per-kilometer costs increasing from ~$24,460 to ~$125,973 (2018 USD), imposing growing economic and logistical burdens. Similar acceleration in retreat is likely across semi-arid coasts globally, with particularly severe implications for developing nations. Future research should quantify the impacts of extreme hydroclimatic events on retreat projections, refine socio-economic and cost models, and assess long-term ecological consequences of recurrent nourishment to inform sustainable coastal management.

- Model assumptions: Forecasts rely on linear regression of LRR and linear relationships between SRR and population density growth; actual trajectories may be non-linear, especially under changing storminess and wave climates. - Bruun rule constraints: The 2D equilibrium approach assumes quasi-constant wave climate, gentle slopes, limited longshore gradients, and minimal human impact. It does not capture alongshore sediment exchanges, complex pocket beach dynamics, or morphological feedbacks, potentially underestimating or misattributing retreat components. - Scenario and time-horizon uncertainty: Longer-term projections (e.g., 2075, 2100) have larger uncertainties; 2075 values are illustrative. The study indicates 2050 projections as more reliable. - Cost estimates: Unit sand price escalation is simplified to a constant 3%/yr and reported in 2018 USD. Actual costs vary with market dynamics, logistics, environmental compliance, and may spike with industrial demand; structural measures to retain nourished sand are not costed. - Data limitations: Publicly available datasets may omit some semi-arid systems; GSAM coverage of semi-arid beaches is limited. Ecological impacts of repeated nourishment are acknowledged but not quantified here. Subsurface processes (e.g., localized subsidence from tectonics/groundwater) need further geophysical validation to refine attribution. - Lower-bound framing: Authors note projected retreat rates and costs should be considered lower limits; hydroclimatic extremes (e.g., strong El Niño, medicanes, megadrought) could increase retreat beyond estimates.