Documenting barrier morphology and coastal setting is a vital first step to understanding shoreline evolution. Airborne lidar uses scanning laser altimetry as a survey method of obtaining topographic information for coastal dunes and intertidal areas above the low-water mark (Fig. 2). Aircraft-mounted sensors combine global positioning systems (GPSs) and laser range finders to remotely map the surface of the Earth over areas tens to hundreds of kilometres in extent, with a horizontal resolution of 1 m or less and a vertical accuracy of 0.10 to 0.15 m. Detailed information about the elevation of the land surface and vegetation is acquired by emitting laser pulses, which reflect off objects and produce a backscatter recorded by the sensor. In addition to a “travel time” for each pulse and subsequent return signal, an intensity of reflectance is also often measured and used to identify vegetation canopy versus ground surfaces. Drones equipped with lidar are being explored as a lower-cost option to acquire coastal data, but it is still expensive and requires experience to use (including a pilot license in some airspace) (Klemas, 2015). This section does not discuss the complex details of how to collect or process lidar, but rather optimally utilizing professionally acquired and processed data.

Traditionally, air photographs, satellite images, and topographic profiles have been used to assess coastal systems as well as plan fieldwork. The advent of platforms like Google Maps, Google Earth, NASA Worldview, and NASA Word Wind, which provide free imagery collected over time, bolstered the amount of data available (Fig. 2). Lidar penetrates the vegetation that often obscured details of the morphology in aerial imagery and removes this obstruction during processing. Digital terrain models created form lidar data refine the morphology, detecting subtle dune topography. This dataset can be used to extract topographic profiles and calculate the volume of barrier sediment supplied above mean sea level (Dougherty et al., 2015, 2012; Oliver et al., 2014). The classic prograded barrier system located near Moruya, Australia, offers an example of the detail and lateral extent mapped in lidar (Fig. 1a) compared to the original two-dimensional topographic profile (Fig. 1b). The lidar captures the uniform shoreline progradation represented by the series of beach–foredune ridges (yellow with high crest in red, Fig. 1a) as well as interactions from inlets, tidal creeks, and open ocean (green and off-white colours, Fig. 1a). Detailed barrier morphology derived from lidar can be used to (1) target areas modified by natural and human processes to understand their impact or (2) avoid them to isolate the influence of storms, sea level, and sediment supply versus accommodation space.

The display or rendering chosen to analyse and present lidar data can impact interpretations. Since coastal systems are relatively low-lying features, the elevation scale range and colour scheme chosen should at least define the barrier from intertidal areas (done using cool and warm colours in Fig. 1a). In more complicated systems the display should be such that important changes in the surrounding geologic setting or within the dune morphology are easily discernible (Fig. 2). Once the lidar is optimally rendered, these remotely sensed data needs to be ground-truthed to detect any errors in data acquisition or processing deficiencies (Gutierrez et al., 2001). This can be done in the field by checking the elevations using traditional survey equipment such as levels and total stations or real-time kinematic (RTK) GPS. It is acknowledged that lidar is not available for large parts of the world and other technologies for mapping morphology exist that may be easier or less expensive (e.g. drone-based “structure from motion”; Christopher Hein, personal communication, 14 December 2019). When using another comparable method to acquire high-resolution, large-spatial-extent topographic data, it is equally important to ground-truth and render them properly.

Augmenting air photos or satellite images with lidar provides a more complete understanding of the geologic setting to contextualize and understand coastal evolution as well as plan fieldwork. A Google Earth image of the Rangitaiki Plains in New Zealand displays a filled coastal embayment that has a prominent series of foredune ridges behind the present-day shoreline (Fig. 2). The lidar data in Begg and Mouslopoulou (2010) show that the infilling did not occur by uniform shoreline progradation, like at Moruya (Fig. 1), but is a rather complex evolution influenced by tectonic and riverine processes. This lidar imaged the modern prograded barrier island that formed after the area experienced ∼5 m of subsidence between 2.1 and 1.72 kyr ago (Begg and Mouslopoulou, 2010: circled in black in Fig. 2d). The lidar data also identified remnants of prograded foredune ridge sequences preserved in the eastern section of the embayment (white circles in Fig. 2d). The detail revealed that the easily erodible beach and dune sands along the seaward side of these prograded sequences appear to have been modified. However, their landward extent does not appear eroded, especially the oldest two sequences that display the same natural transition to back-barrier deposits identified in the modern barrier island (documented by cores in Pullar and Selby, 1971). To test the hypothesis that these features formed similarly to the modern analogue, resulting in a unique set of prograded barrier islands, the lidar data were used to determine the best location to collect GPR transects (grey lines in Fig. 2d). Toggling between overlain lidar and Google Earth images provided pre-field reconnaissance of obstacles (trees, houses, etc.) to consider logistics when targeting each specific profile. Given how rapidly and drastically coastal landscape changes, selecting the Google Earth image dated closest to when the lidar was collected is instrumental to providing good correlation in the overlay. It is also optimal to publish the lidar data augmented with aerial imagery when possible. This is useful for the reader to analyse barrier morphology in relation to shallow subaerial offshore, inlet, estuary sediment deposits, and/or human modification that is sometimes not captured in the lidar.

Once the surface morphology is analysed, the next step to determine how a barrier formed is to study the history preserved in the shallow subsurface. The lidar data should be used to make informed decisions on where best to acquire detailed stratigraphy using geophysics. Ground-penetrating radar (GPR) is a high-resolution geophysical technique can image dune, beach, and nearshore facies with decimetre resolution over kilometres of coast (e.g. Buynevich et al., 2009; Barboza et al., 2011). To achieve subsurface imaging, GPR emits short pulses of electromagnetic energy (microwave radiation) into the ground (Jol et al., 1996). These transmitted high-frequency radio waves are sensitive to the electrical conduction properties of the material being penetrated (dielectric permittivity) and differences in permittivities cause them to be reflected, refracted, or scattered back to the surface. A receiving antenna records variations in the return signal, detecting changes in material properties of subsurface structures and facies by travel time within the waveform. Individual waveforms display changes within the subsurface by recording a wave-amplitude spike at a stratigraphic boundary surface. Collecting GPR along a transect line stacks individual wave traces laterally such that low wave-amplitude signals represent homogenous sediments, and increase in amplitude is associated with greater contrast in sediment characteristics (e.g. change in water content, mineralogy, grain size, sorting, etc.). The variation in waveform detects changes that occur at stratigraphic boundaries, as peaks of high amplitude merge to form strong reflection surfaces. It also detects more subtle changes within the facies, with lower-amplitude peaks forming medium to weak reflections (Fig. 3).

Of the three GOaL techniques, GPR is the most easily accessible and affordable method for coastal geologists to collect and process data independently. The ability to buy or rent a GPR increased as their operation became more user-friendly (e.g. from completely analogue systems with a stylus recorder to partially digital systems using DOS on a control unit or laptop computer, and now some are complete with digital antennas using Bluetooth communication run through simple Windows interfaces on tablets). Currently there are many brands and configurations of different ages in use as well as a variety of software packages that can be utilized to process their data. It is not within the scope of this article to discuss all the differences in components, set-up configurations, settings, processing steps, and terminology. This article advocates for neither a particular unit, antenna, and software nor specific settings or a certain set of processing steps, as multiple variations produce similar high-resolution images of barrier stratigraphy when used correctly. Ultimately the type of equipment used for a certain project likely depends on what is available to the researcher. Novice users should utilize the extensive literature that exists on GPR and its use in coastal settings (e.g. Bristow and Jol, 2003; Buynevich et al., 2009). In addition to acquiring standard knowledge of GPR and the basics of processing, it is useful to research previous publications that use the same equipment that is available to the scientist for specifics. It is also important to reiterate that when starting out it is best to collaborate or consult with someone who has experience with GPR, not just for acquisition and processing, but also especially for interpreting the data. For use in GOaL, it is expected that there is a level of competency in GPR data collection, basic processing, and interpretation.

Initial cross-sectional models of prograded barriers display generalized shallow stratigraphy with large-scale subsurface facies boundaries interpolated from drill core data and isochrons extrapolated from 14C age samples (e.g. Bernard et al., 1962; Curray et al., 1969; Thom et al., 1978: Fig. 1b). The electromagnetic properties of sandy barriers are ideal for producing excellent GPR images because of the high resistivity of the sediment opposing the flow of electrical current (Leatherman, 1987). Collecting GPR across entire prograded barriers can extract high-resolution stratigraphic records providing a continuous cross-sectional view of barrier architecture that detects small-scale features and large-scale facies boundaries previously unrecognized in point source core data (e.g. Fitzgerald et al., 1992; Jol et al., 1996; van Heteren et al., 1998). A unique outcrop of a Pleistocene prograded barrier in One Tree Point, New Zealand, illustrates the sensitivity of GPR in detecting stratigraphy (Fig. 3). The geophysical record shows how the heavy mineral beachfaces create the strongest reflections between 2 and 5 m. Medium-strength reflections are detecting the more diffuse heavy mineral concentrations within the dune sequence (0–2 m of depth) and in the cross-bedding preserved as a bar migrated onshore in the nearshore (6–7 m of depth). The weak, reflection-free areas in the dune and nearshore represent homogenous deposits. However, GPR uncovers structure in the fine-grained, well-stored, quartz dune sand at the top that would have been otherwise invisible to the naked eye.

Ground-penetrating radar can detect differences such as compaction and/or water content, allowing stratigraphy to be more obvious in the geophysical records. For example, the prominent reflections between 5 and 6 m identify the transition in sands between beach and nearshore facies that is otherwise not detectible without grain-size analysis (Fig. 3). The ability of GPR to detect individual beachfaces as well as their boundaries with dune and nearshore facies enables them to be mapped throughout a prograded barrier. Mapping the beachfaces through time allows their elevation to be used as a sea-level proxy (e.g. van Heteren et al., 2000; Costas et al., 2016) and their geometry to produce storm records (e.g. Goslin and Clemmensen, 2017; Lindhorst et al., 2008). Because GPR is sensitive to subtle changes in the subsurface, the record must be ground-truthed using cores, augers, or outcrops in order to verify barrier facies and boundaries (e.g. Costas and FitzGerald, 2011; Hein et al., 2013, 2016). Additionally, topographic profiles of the present-day beach and sediment samples from each facies should be collected, preferably capturing both storm and swell geometry and sedimentology. Typically, erosion concentrates storm lag deposits on the steepened upper beachface and/or flattened lower beachface, which causes high-amplitude reflections that are more prominent than the low-amplitude signature of the homogenous berm sands that accrete during intervening swell conditions. Mapping these distinct geophysical signatures throughout the barrier enables storm records to be extracted (Buynevich et al., 2007, 2004; Dougherty et al., 2004). As a whole, the high- to medium-amplitude beachface signatures stand out compared to the weak or reflection-free signals in the dune and nearshore facies (Fig. 4). This contrast allows beachface elevation to be used as a proxy for sea level (e.g. van Heteren et al., 2000; Rodriguez and Meyer, 2006; Dougherty, 2014). While lidar can be used to apply a coarse topographic correction to the GPR data, it is recommended that precise topographic profiles be surveyed in the field and tied directly to the GPR transect, in particular if the aim is to extract sea-level and storm records.

In order to delineate barrier facies and individual beachfaces it is fundamental to ensure that the amplitude of the waveform peak relates to the contrast within the stratigraphy (e.g. the strongest reflections are the storm-eroded beachfaces and the weakest are homogeneous dune sands). The waveform amplitudes can be adjusted using what is referred to as a gain control. Unlike other basic processing steps, there has been relatively little discussion about gain in the literature; but the fact that incorrectly gained data can impact interpretations warrants attention. The correct application of gain is not just important to accurately represent and interpret barrier stratigraphy, but also critical to the extraction of sea-level and storm records from it. Gain represents the value by which the scaled waveform data are multiplied to get the output data. It is important to adjust the gain according to the core, auger, and/or outcrop data as low gain makes all reflections weak and high gain makes all reflections strong. This lack of contrast makes it hard to distinguish different barrier facies boundaries (used as a sea-level proxy), let alone individual beachfaces (to determine eroded paleo-beachfaces used to construct a storm record). It is also important to keep in mind that individual changes in the subsurface result in double peaks within the waveform, which are presented in the GPR record as prominent coupled lines (demonstrated in Fig. 3 as white and black or black and white, depending on normal or reverse polarity). This means that not all lines on a GPR record represent changes in the subsurface (e.g. Fig. 3). As such, it is not recommended to simply trace every line when interpreting GPR data. Three-dimensional grid modelling can be used to visualize how good gain control can distinguish barrier facies boundaries (Fig. 4a) and isolate storm-eroded beachfaces by interpolating the highest-amplitude reflections among a series of shore-perpendicular transects (Fig. 4b). The use of 3-D models is not necessary for extracting sea-level and storm records, but could be useful in studying shoreline rotation (Harley et al., 2011; Short and Trembanis, 2004) or smaller-scale and more irregular features such as beach cusps (Coco et al., 1999; Masselink et al., 1997).

The final step of the GOaL approach is to apply a chronology to barrier formation using detailed morphostratigraphy. Adding a temporal component to coastal formation is important to understand shoreline evolution over timescales that operate on longer terms than that documented historically. Optically stimulated luminescence (OSL) dating is a geochronology technique that determines the time elapsed since buried sand grains were last exposed to sunlight (e.g. Huntley et al., 1985). Upon burial, ionizing radiation from surrounding sediment (by radioactive decay of U, Th, Rb, and K) and cosmic rays is absorbed by the mineral grains and stored in traps within their crystal lattice. Exposure to sunlight can bleach away light-sensitive luminescence signal and reset the “clock” to zero. This stored radiation dose can also be evicted with light stimulation in the laboratory and the energy of photons being released can be measured. Calculating the age when the grain was last exposed to sunlight is based on quantifying both the radiation dose received by a sample since its zeroing event and the dose rate which it has experienced during the burial period. OSL chronology can provide the resolution necessary to decipher the decadal-, centennial-, and millennial-scale patterns of coastal behaviour necessary to reconstruct sea-level curves, determine storm frequencies, and calculate sediment supply and progradation rates.

Originally, dating coastal barrier formations was dependent on sourcing scarce organic matter (often involving extensive coring) and extrapolating the conventional radiocarbon dates from the nearshore to the surface using isochrons (e.g. Fig. 1a). Since OSL chronology determines the time elapsed since mineral grains were buried, this technique dates when paleo-beachfaces and relict foredunes were forming. Dating coastal systems using OSL has been very successful on a global scale (e.g. Jacobs, 2008). Quartz is both a principle mineral used in luminescence dating and abundant in coastal barriers. Therefore, lidar and GPR can be used to target specific stratigraphic layers in a strategic manner for sampling.

This targeted approach using OSL has been shown to more accurately date beach and dune formation than inferred radiocarbon ages from deep nearshore or offshore organic deposits (e.g. Oliver et al., 2015). Oliver et al. (2015) offer an example comparing radiocarbon and luminescence ages at the Moruya Barrier. Because this study focused on comparing chronologies, lidar and GPR data were not presented in Oliver et al. (2015), but both techniques were used to target specific stratigraphic layers prior to OSL sampling (Fig. 1c). The results revised the long-standing theory, based on radiocarbon dates, that the barrier prograded at two different rates before halting 3000 years ago due to diminished sediment supply (Roy and Thom, 1981). The OSL data revealed that the barrier has prograded at a constant rate throughout the Holocene (Fig. 1c). Nevertheless, the radiocarbon dating of shell deposits within the beach facies has been shown to provide similar ages to OSL dates acquired from associated beach and dune deposits (e.g. Hijma et al., 2017; Murray-Wallace et al., 2002). Therefore, where suitable organic material exists within the barrier sands, radiocarbon dating can be utilized at a fraction of the cost and time of OSL analysis. Accurate 14C dating requires experience to understand the provenance of the organic material and scrutinize the type of shells used, since beach facies often contain reworked shells (e.g. Rodriguez and Meyer, 2006).

Collection of OSL samples in the field is relatively easy following various methods described in the literature or guidance from someone with experience; however, the processing and analysis of samples requires a scientist trained in luminescence chronology (e.g. Bailey and Arnold, 2006; Huntley et al., 1985; Jacobs, 2008). Therefore, it is not within the scope of this paper to discuss how to process OSL samples (e.g. sample prep and mineral separation) or the complicated intricacies of analysis (e.g. assumptions like water content and burial history or considerations of experimental conditions and statistical models for each sample to be dated). Instead this section focuses on demonstrating the utility of OSL in barrier systems and how it can be optimized by using lidar and GPR to inform researchers on optimal locations for OSL sample collection.

An advantage to OSL dating is that samples can be collected from anywhere in sandy barrier systems in contrast to radiocarbon dating, which relies on organic material often found at depth and thus requires assumptions on the spatial extent of the radiocarbon ages. However, because OSL dating is expensive and time-consuming (relative to radiocarbon), it is advised to strategize sample collection in order to minimize cost while addressing the research questions. The collection and analysis of lidar and GPR provides a detailed understanding of the system stratigraphy, which enables strategic OSL sampling. To ensure accurate sampling of the target facies, it is recommended to also operate the GPR in the field during OSL sample collection, especially if it is collected by coring rather than from an open trench.

When reporting ages in publications, it is important, in particular for younger ages, to indicate the date of OSL measurement (Zhixiong Shen, personal communication, 5 March 2018), since the ages refer to time before OSL measurement. This is also important when presenting both radiocarbon (in calibrated years BP) and OSL dates in the same discussion. While the difference matters little when discussing dates of >10 000 years, it can be significant for the last 1000 years. For these shorter time periods, converting the radiocarbon ages and presenting all dates in years CE is a solution (Christopher Hein, personal communication, 14 December 2018).

Morphostratigraphy from aerial imagery, lidar, and GPR is not only useful in determining where best to collect OSL samples that capture uniform progradation to construct complete paleoenvironmental records (Fig. 1), but also to identify significant shifts in barrier evolution so that they can be dated (Fig. 5). East Beach Barrier in New Zealand demonstrates how surface and subsurface data guided OSL to better understand a recent transition from uniform progradation to a more complex evolution (Fig. 5a; Dougherty, 2011). In order to decipher the timing of this shift, aerial imagery was used to target the changes in morphology and GPR to locate corresponding differences in the underlying stratigraphy (Fig. 5). The OSL samples were measured in 2005. The age of the last relict foredune preserved indicates that the barrier prograded uniformly until at least 1700 years ago (Fig. 5b). After this time, a large dune blowout formed, modifying any previously existing morphology. The distinct shift in both stratigraphy and morphology was dated at ∼1000 years ago (Fig. 5b). This younger age is important to understand the change in evolution within the context of the regional setting, since in the last millennium three major events could have impacted the coastline: (1) the arrival of the Maori people (Wilmshurst et al., 2008), (2) sea level stopped dropping from a mid-Holocene highstand (Dougherty and Dickson, 2012), and/or (3) a large tsunami struck the area (Nichol et al., 2004).

Costas et al. (2016) provided an excellent example of how GOaL can be used to reconstruct Holocene sea level from Troia Peninsula, Portugal. Lidar of this complicated spit system highlights the prograded section of the barrier targeted for GPR and OSL collection across the entire barrier (not one but two transects), capturing a complete progradational history (Fig. 6a). The presentation of both raw (not shown) and interpreted GPR data (Fig. 6b) across the entire barrier is ideal for the reader to see the beach and berm elevational markers used as a sea-level proxy. Complete transects are often not collected for logistical reasons, and when collected, they are often published only partly. It is best to collect at least one single transect line that spans the entire barrier to capture a complete Holocene history. It is also very informative to indicate the location of OSL samples on the GPR profile, regardless of whether it is displayed on the entire record or on selected detailed sections. This allows the specific stratigraphic section dated to be identified. In Costas et al. (2016), topographic profiles of the modern beach and cores were used to ground-truth the GPR such that the berm–beach-dune contact could be interpreted as a proxy for sea level (Fig. 6b); this is summarized in Fig. 6c. Results showed good agreement with known sea-level curves in southwest Europe. This study, along with work from North America (van Heteren et al., 2000; Rodriguez and Meyer, 2006; Billy et al., 2015), demonstrates the potential of applying this method to regions where middle to late Holocene records are not as well documented and/or are debated (e.g. Dougherty, 2018b).

Oliver et al. (2017b) used GOaL on two proximal prograded barriers (Wonboyn and Boydtown) along the southeast coast of Australia. GPR data spanning millennia to the present-day berms were collected, and Oliver et al. (2017b) concluded that all of the paleo-beachfaces in the geophysical record were stacked storm profiles with no fair-weather stratigraphy preserved. However, this conclusion likely overestimates the number and impact of storms recorded due to the processing, “interpretation”, and analysis of the GPR data with no cores or augers to ground-truth any beachface reflections. The processed data have a high gain applied (e.g. Fig. 7b) and the “interpreted” transects used for analysis simply have every line traced (e.g. Fig. 7a) instead of annotating interpreted barrier facies and/or individual beachfaces used for analysis (e.g. Fig. 6b). The representation of the GPR data in Oliver et al. (2017b) makes it hard to differentiate beach and dune facies, let alone detect differences within the paleo-beachface stratigraphy (Fig. 7).

Analyses of these GPR-imaged beachfaces by Oliver et al. (2017b) document a similar range in geometries traced in all transects, including Boydtown A shown in Fig. 7a. Given that the geometry of the paleo-beachface stratigraphy is similar to that imaged in the modern berms, which by their nature are constructed in fair-weather conditions, the conclusion that only storm-eroded profiles are preserved is called into question (Dougherty, 2018c). Without cores or augers to ground-truth these beachface reflections, it is impossible to definitively determine if the beach facies consist only of storm lag concentrations (e.g. heavy minerals, coarse grains, and/or shell hash) or if there are intervening quartz-rich layers deposited during fair-weather conditions. This demonstrates the importance of using sedimentological analysis to appropriately adjust the gain of GPR data so that this remotely sensed information can be accurately interpreted and conclusions unequivocally drawn.

An ideal implementation of the GOaL approach to extract a regional storm record from prograded barriers is as follows: (1) use lidar to determine proximal systems that are likely to contain the most complete Holocene record and plan where to acquire continuous transects across each barrier; (2) collect GPR data from the barriers and adjust the gain to highlight the strongest reflections in the beach facies and ground-truth to confirm they represent eroded paleo-beachfaces consisting of storm lag deposits (e.g. Dougherty, 2014, 2018c); and (3) use the GPR data to locate the most prominent eroded paleo-beachface reflections and obtain OSL samples from associated materials (e.g. Buynevich, 2007). Once the processed GOaL data are combined, construct records of all events and compare those of similar timing and magnitude to analyse for recurrence intervals (e.g. Dougherty, 2014, 2018a).

Oliver et al. (2017a) used GOaL to decipher the complex progradation of Seven Mile Barrier in Tasmania, Australia (Fig. 8a). The GOaL dataset was used to conclude that there was a “pause” (which by definition means “temporary stop”) in sediment supply and progradation from ∼6.5 to 3.5 kyr ago and over the past 500 years (Fig. 8a, b). However, gaps in the lidar-derived morphology, GPR stratigraphy, and OSL chronology coincide with these timeframes, raising the question of whether these interpreted hiatuses resulted from a lack of data (Fig. 8d). The breaks in the chronology and stratigraphy may have stemmed from the choice of render for the lidar dataset and how the topographic profiles were extracted from it. The green colour scheme obscures features in the morphology that distinguish changes in the barrier evolution, such as when the largest relict foredune ridges formed as the barrier prograded a minimum of 200 m between ∼6.5 and 3.5 kyr ago (Fig. 8a).

To demonstrate how choosing a render with greater contrast can accentuate barrier features, lidar data with 5 m resolution (freely accessible from Geoscience Australia at http://www.ga.gov.au/elvis/, last access: June 2017) were augmented with a Google Earth image (Fig. 8c). This highlights not only the height of these larger relict ridges, but also reveals that these features bifurcate to the east (indicating greater progradation in this part of the barrier, possibly ∼1 km) and extend laterally to the west, filling an abrupt increase in accommodation space (Fig. 8c). The discontinuous topographic profiles presented in Oliver et al. (2017a) not only mask the increased size of these prominent relict foredune ridges, but also omit the anomalously large foredune that formed along the southeast half of the barrier over the past 500 years (Fig. 8b, d). While no GPR data were collected for this foredune or the one in the north, there is evidence of transgression as the large 60-year-old foredune is unconformably deposited on top of the 1400-year-old low-lying foredune (Fig. 8c and d).

Modifying the display of the lidar dataset and extracting continuous topographic profiles from it identified gaps in the data which challenge the Oliver et al. (2017a) conclusion that sediment supply and progradation paused or stopped temporarily twice during the formation of Seven Mile Barrier. During ∼6.5 and 3.5 kyr ago progradation did not stop, but rather appears to have slowed as the barrier built into varying accommodation space. Progradation has now likely ceased permanently as the barrier seems to have transitioned to transgression in the last 500 years and it is unlikely to resume prograding during accelerating sea-level rise. The existence and size of foredune features deposited during both timeframes indicates that there was a continuous supply of sediment to build them. Additional GPR and OSL data are necessary to confirm these interpretations (Dougherty, 2018b). In the meantime, these serve as examples of how despite obtaining a GOaL dataset, significant features or gaps in data can be overlooked and lead to questionable interpretations. Where it is not feasible to collect parts of the dataset, this absence of data should be acknowledged, addressed, and considered when discussing interpretations or conclusions as well as the level of confidence with which they are asserted.

Applying the methodology presented in this paper can optimize the GOaL dataset at Seven Mile. This would not only fill the gap in knowledge with respect to barrier formation and sediment supply, but could also produce a Holocene sea-level curve and storm record to complete a GOaL hat-trick. An ideal implementation of the GOaL approach at Seven Mile would be as follows: (1) use lidar to identify a transect spanning the entire Holocene record that captures shifts in evolution (western profile in Fig. 8c, d) and utilize areal imagery to locate the nearby road and airstrip that both provide access across the entire barrier. (2) Collect a continuous shore-normal GPR profile spanning the barrier, with additional data acquired specifically to document the larger foredunes that represent shifts in the barrier evolution. (3) Ground-truth the GPR and lidar data using cores, augers, outcrop mapping along the eroded backside of the barrier, sediment analysis, and topographic profiles. (4) Utilize the lidar and GPR data to plan OSL sample locations targeting the most prominent storm-eroded beachfaces throughout the barrier and bounding stratigraphy of the larger foredunes to date their formation. (5) Integrate the GOaL dataset for analysis after all the components have been processed and rendered. (6) Use the digital elevation model from combined lidar and OSL data to calculate the volume of barrier sand above mean sea level, for different areas, to determine sediment budget over time (e.g. Dougherty et al., 2015, Dougherty, 2018b). (7) Combine GPR and OSL data to reconstruct a record of storms (e.g. Buynevich et al., 2007; Dougherty, 2014). (8) Use the combined GPR and OSL data to construct a sea-level curve (e.g. van Heteren et al., 2000; Billy et al., 2015; Costas et al., 2016). Finally, (9) evaluate barrier formation to determine the nature of shifts in evolution through time and consider them with respect to any changes identified in sediment supply relative to accommodation space, sea level, and/or storm erosion. In the future, completing GOaL hat-tricks on some of the many prograded barriers around Tasmania and mainland Australia will decipher local from regional signatures.