ABSTRACT

Finkl, C.W. and Makowski, C., 2020. Lateral extrapolation of coastal catenary sequences using the Biophysical Cross-shore Classification System (BCCS) to create shore-parallel situational zonation mapping units. Journal of Coastal Research, 36(3), 457–471. Coconut Creek (Florida), ISSN 0749-0208.

The Biophysical Cross-shore Classification System (BCCS), which uses transects to assess shore-normal ecological and geomorphological successions from offshore to onshore within a coastal belt (Finkl and Makowski, 2020a), also provides a basis for extrapolating cross-shore catenas into shore-parallel units. This paper shows that three-dimensional transects can be parameterized in terms of alongshore breadth as well as cross-shore width and depth below or elevation above sea level. The codification of cross-shore environments and habitats in the framework of the BCCS provides an interpretative basis for determining the lateral extent of transect units (archetypes and sub archetypes) by lateral extrapolation to polygonal mapping units. The method discussed here is akin to geological cores or cross-sections that are used to laterally extrapolate units within vertical sequences. Repetitive successions of archetypes, based on cross-shore ecological interpretation of satellite imagery, results in a common master sequence referred to as a Dominant Catenary Sequence (DCS). The DCS is composed of generic archetypes, such as Barrier, Beach, Beach Ridge, Cliff, Coral Reef, Delta, Dune, Flat, Ice, Lagoon, Mountain, Rock, Till (Glacial Material), Upland, and Wetland. The more detailed Coastal Ecological Sequence (CES) of a coastal belt, which is defined by a discrete codification sequence built up from the DCS, is formulated by cognitive geovisual analytics to link the dominant catena with a numbered shore-parallel shape distinction and subscripted sub archetypes to refine the sequential composite archetypes in a DCS. Once the DCS- or CES-labeled transect has been plotted on a satellite image, the identified shore-normal units can be extrapolated into shore-parallel polygons by traditional (i.e. visual, cognitive) image interpretive and mapping techniques to show the spatial extent of classified archetypes and sub archetypes.

INTRODUCTION

Coastal classification, in many ways, has remained a bit of an enigma because no universal system has yet been devised. This state of affairs exists because of the complexity of coastal belts that often confounds attempts to rationalize complex geomorphological-ecological setups into comprehensive classification systems. As a consequence, most coastal classifications are special purpose in order to comprehend a rather limited scope of phenomena that are manageable in some useful and informative way (Makowski and Finkl, 2016). Most special purpose classifications are thus useful in their own way and serve the needs of the appropriate research communities in the biological, geological, engineering, and applied sciences as well as management and sociology (Fairbridge, 2004; Finkl, 2004; Kelletat, 1989, 1995; Kelletat, Scheffers, and May, 2013; Klemas et al., 1993; Makowski, 2014; Makowski, Finkl, and Vollmer, 2015, 2016, 2017; Short, 2006; Short and Woodroffe, 2009; Woodroffe, 2002). In a similar vein, the Biophysical Cross-shore Classification System (BCCS) has been proposed (Finkl and Makowski, 2020a,b) as another example of a special purpose coastal classification that focuses on cross-shore identification of geomorphological-ecological natural systems as opposed to more common approaches that feature alongshore characterization of shorelines and coastlines. Shore-parallel classifications are useful in many ways and serve as a viable means of identifying some specified property of the shore. These approaches, however, are limited by the fact that they do not penetrate very far inland, and consequently many important aspects of a coastal belt are neglected to the extent that prominent features of the coast are ignored.

An example of such a case might occur in the designation of shores that are characterized as sandy beaches. Whereas it may be true that the shore contains a sandy beach, it may be equally true that natural features (e.g., strandplains, cliffs, mangroves, wetlands) lying behind the beach may better characterize the overall nature of the coastal belt. In an effort to better comprehend the overall character of the coast beyond designation of shore types, Finkl and Makowski (2020a,b) proposed the idea of a coastal belt that incorporates the inclusion of inland extents of coastal features in a cross-shore classification system. Identification of specific types of coastal belts is the result of a three-dimensional (3D) classification that includes length, width, and elevation (including bathymetry) via cross-shore transects that contain catenary associations of archetypes and sub archetypes.

The BCCS is not intended as a replacement for existing coastal classifications, especially since these prior efforts well serve their intended purposes. Rather, the cross-shore approach of the BCCS is a different tack that tries to bring the concept of coastal width to shore-parallel classificatory efforts. The advantage of such an approach is that it provides a larger overview of what constitutes a coastal scene and brings together linkages between coastal environments (Finkl and Makowski, 2020a,b).

Purpose

The intent of this paper is to show how cross-shore catenas of archetypes and sub archetypes can be laterally extended or extrapolated to characterize an entire satellite scene that represents part of a more extensive coastal belt. The goal of extension of geomorphological-ecological units in terms of archetypes and sub archetypes is to show that BCCS cross-sections have additional utility in a more complete description of coastal belts.

Goal

Crucial to this goal is the elucidation of image analysis in terms of cognitive mapping and visual analysis. For the simple task at hand, interpretation of coastal environments can be achieved without reference to complex machine classifications of satellite imagery. Human perception and interpretation are often undervalued or unappreciated but inevitable components in remote sensing image analysis. Human intervention is requisite for visual image interpretation when the interpreter manually performs the analysis (e.g., Campbell and Wynne, 2011; Finkl and Makowski, 2019a,b; Snook et al., 1987; Wang et al., 2015). Although image processing has become more and more automated (Costa, 2019), human screening and interpretation remain indispensable at certain stages (León-Pérez, Hernández, and Armstrong, 2019), especially in cases where identification of geomorphological-ecologic units is a simple task. One particular example is where the operator plays a crucial role in the development of reference maps for coastal belts, as in the case of interpreting shore-parallel units from cross-shore transects. This technique is often accomplished by visual interpretation of an image by an operator (e.g., Gardin et al., 2010; Rasid and Pramanik, 1990; Schowengerdt, 1983).

METHODS

The methodological approach reported here involves lateral expansion of cross-shore transects to produce sketch or reference maps where spatial distributions of archetypes and sub archetypes are displayed on top of satellite images in the form of polygonal layers. That is, the satellite image functions as a base map upon which polygons delineate the spatial extent of units demarcated in the cross-shore transects. Hand digitization of units seen on the satellite image can be accomplished in a number of ways using various geo mapping software programs such as ArcGIS, AutoCAD, Google Earth Pro, or any number of image analysis programs such as GIMP, PhotoShop, or even presentation software such as PowerPoint, etc. The choice of digitizing software is almost unlimited, but the actual method of manual heads-up or onscreen digitization is not important to this study; interpreters should use whatever procedure they are comfortable with. Because the delineation of archetype and sub archetype boundaries is a simple task, any line-drawing software will be sufficient to outline polygons on the satellite image.

The basic steps that were deployed here involved selecting the desired satellite scene using Google Earth Pro, importing the image into GIMP (a cross-platform image editor available for GNU/Linux, OS X, Windows, and more operating systems) to increase the scene resolution, and then exporting the high-resolution image to storage for use in a line-drawing software program. In this case, the image was imported into PowerPoint for labeling and construction of polygons, but these tasks could have just as well been accomplished in GIMP. No matter what software is used to accomplish these tasks, the main point is to relate and show the lateral extent of the units shown in the cross-shore transect. Delineation of polygon boundaries, labeling, and use of translucent color tints help to define the spatial context of the satellite image interpretation.

Derivation of useful spatial information from the satellite images is the task of image interpretation, which includes visual detection of the geomorphological-ecological units indicated in the cross-section. The identification procedure is used as the first step of image interpretation to identify vegetation, topography, soils, lithology (rock types), water bodies, etc. The higher the spatial/spectral resolution of an image, the more detail that can be derived from the image; this is why it is important to select the highest resolution possible in Google Earth Pro and then enhance the image in another program. The delineation procedure is then used to outline the recognized targets for mapping purposes. Identification and delineation combined together were then used to map the extent of archetypes and sub archetypes in the satellite scene. When the whole coastal belt scene is processed by these two procedures, the image is classified using the BCCS transect as an initiator for map construction.

The success of the map construction phase depends on the ability of the human analyst/interpreter to extract pertinent information by visual inspection (cognition) of an image composed from the image data. That is, the image interpretation depends on the analyst effectively exploiting the spatial, spectral, and temporal elements present in the composed image product. Information spatially, for example, is present in the qualities of shape, size, orientation, and texture. Coastlines, shorelines, river systems, topography, rock outcrops, ecotonal boundaries between archetypes, and so on are usually readily identified by their spatial disposition. Spectral clues are also used based on the analyst's foreknowledge of, and experience with, the spectral reflectance characteristics of typical ground cover types and knowledge of how those characteristics are sampled by the sensor on the satellite used to acquire the image data (e.g., Das et al., 1997; Fuller et al., 1987; Gardin et al., 2010; Richards and Jia, 1999).

Even though the satellite data from Google Earth Pro is available in digital form that can be spatially quantized into pixels and radiometrically quantified into discrete brightness levels, the several computerized approaches that are possible for extracting information are not necessary as methodological procedures to visually analyze the scene. Visual interpretation, cognition, and manual collation of map information are all that is required to complete the task of laterally extrapolating the cross-section units. Although there is a general tendency for researchers to attempt various aspects of computer image processing, as described for example by Lasaponara and Masini (2011), it is emphasized here that human screening and interpretation procedures remain indispensable for certain tasks such as those identified in the delineation of coastal archetypes and archetypes.

Visual analysis procedures were applied to four satellite scenes in different latitudinal and climatic zones to obtain the results shown in paired diptych figures with one image scene that is not annotated (no cross-shore transect) and the corresponding annotated scene with a cross-shore transect and extrapolated polygons. There are two examples from low latitudes (tropical and subtropical climatic zones) and one each from middle and subpolar latitudes to illustrate the flexibility of the BCCS for a range of latitudinal zones and different types of coastal belt climates. The first image scene (Figure 1) occurs on the tropical central coast of the Northern Red Sea Region of Eritrea (16°32′03″ N, 39°09′34″ E) about 120 km NE of Asmara, the capital city that is situated about 70 km inland from the coast. The climate of this tropical desert coastal region is classified as BWh (Hot Tropical Desert Climate) in the Köppen-Geiger climate classification system (e.g., Peel, Finlayson, and McMahon, 2007). The second image scene (Figure 2) to which these procedures were applied occurs on the subtropical monsoon SE coast of Madagascar in the vicinity of Ambanihazo and Antsotso Avarata (24°35′58″ S, 47°14′50″ E). The climate of this coastal belt is classified as Am (Tropical Monsoon Climate which is also occasionally known as a Tropical Wet Climate or a Tropical Monsoon and Trade-Wind Littoral Climate). The third example features mapping procedures based on the BCCS cross-shore transect extrapolation in a middle latitude coastal belt with an abrasion platform (51°26′12″ N, 3°35′39″ W) about 400 m wide. This oblique image scene (Figure 3) occurs on the SW coast of Wales in the Vale of Glamorgan facing the northern shore of the Bristol Channel. The warm temperate climate here is classified as Cfb (Marine West Coast Climate). The high latitude cold region example is from the west coast of subpolar Hudson Bay and shows beach ridges (58°08′26″ N, 92°54′46″ W) in Wapusk National Park about 45 km south of Churchill in NE Manitoba, Canada (Figure 4). The boreal climate of this coastal belt is classified as Dfc (Subarctic or Subpolar Climate). Application of visual analytics to these four examples is presented in what follows as a result in terms of coastal belt imagery that is annotated with a BCCS cross-shore transect and graphics showing the lateral extension of archetypes and sub archetypes.

RESULTS

The Eritrean tropical desert coast is shown in Figure 1a as a raw image without any annotations or graphics for comparison with the annotated Figure 1b. At first glance, in gross form the coastal belt is classified by the Dominant Catenary Sequence (DCS) of Coral Reef-Beach-Wetland-Upland (Cr-Be-W-U) archetypes. The Coastal Ecological Sequence (CES) of this northern transect classifies out as 2,4CrcpBecaWma,mr,slUde. As shown in Table 1, this codification translates to curved and embayed tropical compound (combination of fringing and patch) coral reefs backed by a carbonate beach with wetland mangroves, marshes, and salinas that grade into a desert upland. This interpretation is the result of the northern BCCS cross-shore transect shown in Figure 1b, to which could be added the additional sub archetype of desert upland if the transect, as shown, is extended farther inland. Because the spatial distributions of the wetland ecosystems are fairly complicated, they are called out separately from the main cross-shore transect by separate smaller arrows that point to the image pattern mangrove, marsh, and salina. These ecosystems can be visually differentiated on the basis of texture, tone, and pattern in the satellite image (e.g., Wang et al., 2003). Their spatial distributions can thus be extended laterally both north and south of the cross-shore transect, giving a more complete characterization of the coastal belt. The southern transect introduces Lagoon and Flat archetypes in a slightly more complicated catenary sequence because of the presence of an island. This transect shows that archetypes and sub archetypes repeat as part of the cross-shore catena to give the following CES: 2,4CrcpFsaCrfrWmaWslLopWmr,slUde. This result of applying the BCCS looks somewhat complicated, but when compared to the southern transect in the image it comprehends a coastal setting with an offshore island. Consequently, it should be anticipated that some catenas would repeat and others might be introduced to accommodate the complexity that is present in the satellite scene of this coastal belt. This complex situation was deliberately selected to demonstrate the robustness and flexibility of the BCCS to cope with intricate setups in coastal belts. The results of these two transects can be used to define different types of coastal belts or be amalgamated for a broad overlook depending on the needs of research priorities to characterize this coastal belt in general terms of a Coral Reef-Beach-Wetland-Upland (Cr-Be-W-U) archetypical sequence.

The raw satellite image in Figure 2a on the subtropical monsoon SE coast of Madagascar shows a typical barrier coastal belt without any annotations or graphic overlays. It should be used to compare the results of cross-shore catenas of archetypes and sub archetypes in Figure 2b that form the following CES: 7BambBeca,siDuWma,mrLclUgr. The catenary codification (Table 1) for this coastal belt translates to straight mainland barrier (with tombolo) carrying a subtropical mixed carbonate-silica beach with dunes backed by wetland mangroves and marshes with closed lagoons fronting grassy uplands. This overall BCCS of the coastal belt related to the cross-shore transect is shown in Figure 2b, with overlaid mapping notations serving as the lateral extrapolation of archetypes and sub archetypes. The extrapolated spatial distributions of archetypes and sub archetypes are the results of visual analytics and cognitive mapping, where the units are shown in colored polygons, lettering, and arrows that are superimposed on top of the image. The annotations are meant to be compared with the raw image in Figure 2a, which can be used as a guide to the interpretations and extrapolations shown in Figure 2b. The type and sequence of archetypes and sub archetypes shown in Figure 2b are typical of barrier coastal belts in general. That is, the results of visual analytics applied to the SE coast of Madagascar shown here suggest or constitute a generic guide to barrier DCS and CES coastal belts worldwide, regardless of latitudinal zonation.

The unannotated oblique satellite image shown in Figure 3a, from a warm temperate Cfb climate (Marine West Coast Climate) on the SW coast of Wales in the Vale of Glamorgan facing the northern shore of the Bristol Channel, shows an example of a wide abrasion platform that fronts a cliff archetypical coastal belt. Figure 3b shows how this Rock-Beach-Cliff-Upland (R-Be-Cl-U) DCS can be expanded into the following CES: 4,6Rpl,tsBeca,rp,siClse,vc0%Ufo,gr using the codification codes given in Table 1. Translation of the CES catena is easily stated as middle latitude embayed headlands and promontories interspersed by rock platforms backed by mixed carbonate-silicate beach ramparts seaward of bare (∼0% vegetative cover) sedimentary cliff talus and scree deposits surmounted by forest and grassy uplands. The large, red arrow in the figure exemplifies a typical cross-shore transect that might be made in this kind of situation. Arrows, polygons, and lettering indicate the lateral or shore-parallel extension of the cross-shore catenary sequences. The interpretation of the satellite image in Figure 3b should be compared to the unannotated raw image in Figure 3a to discern the shore-parallel extrapolations.

Figure 4a is an unannotated satellite image from the west coast of subpolar Hudson Bay in Wapusk National Park in NE Manitoba, Canada. This high latitude example was included to show the viability of BCCS results in a cold region coastal belt. The annotated example in Figure 4b shows repetitive successions of archetypes and sub archetypes along a barrier coast that fronts strandplains and wetlands. The broad red arrow with BCCS codes with subscripted type shows the cross-shore sequencing of sub archetypes that compose the following codification of this coastal belt: 7BabiFmuBesiDuWmrBrspWmrBrspWmr. This is the complete, long version of the BCCS characterization of this cross-shore sequence; however, ignoring the repeats in the catenary associations, the code would be Ba-F-Be-Du-W-Br or Barrier-Flat-Beach-Dune-Wetland-Beach Ridge (cf. Table 1) archetypical sequencing that translates into: Straight subpolar tidal mudflats fronted by barrier islands and backed by silica beaches, dunes, wetland marshes, and beach ridge strandplains. Figure 4b shows the lateral (shore-parallel) extension of the cross-shore codes to provide a broad characterization of this coastal belt. The result of this effort is the creation of classification mapping units through the lateral extrapolation of cross-shore catenas, thereby allowing for a wider cross-shore swath that characterizes a larger area of this coastal belt.

DISCUSSION

The Biophysical Cross-shore Classification System (BCCS) was devised as a tool for cross-shore characterization of coastal belts worldwide. It provides a means of classifying coasts with a landward component versus a traditional alongshore-only designation such as a sandy or rocky coast. Furthermore, by incorporating nearshore bathymetry and onshore land elevation, the BCCS constitutes a 3D impression of a coastal belt that reached from some distance offshore to inland environments (see Finkl and Makowski, 2020a,b). Based on coastal geomorphology and biophysical ecology, the BCCS was amenable to the elucidation of cross-shore sequences of archetypes and sub archetypes that formed commonly occurring catenas that were, in fact, so ubiquitous and repetitive that they could be used to characterize coastal belts in all latitudinal zones. The cross-shore concatenations of a codification system (Table 1) were designed to be applied as shore-normal transects. Thus, numerous closely spaced transects would provide keen insight to the cross-shore properties of a coastal belt.

Lateral Extrapolation of Cross-shore Catenas

An alternative to the construction of numerous closely spaced transects is raised by the possibility of lateral extrapolation of cross-shore concatenations via the same processes that were used to compile the original transects. In addition, cognitive mapping procedures assisted by visual analytics, including the collation of collateral data, could be easily applied to alongshore extents in such a manner that ecological (environment and habitat) maps could be prepared with little difficulty. The process is scale dependent and limited in the first instance to a single satellite scene. Although not described here, it is obvious that satellite scenes could be conjoined to represent larger areas of coastal belts. In this paper, four satellite images were interpreted to test the effectiveness of lateral extrapolation of cross-shore catenas into mapping units.

The methodology shown in this paper was to laterally extend the cross-shore archetypes and sub archetypes in an alongshore direction. As a practical means, the satellite images could be annotated in various ways to convey the alongshore extension of the so identified cross-shore units by using lettering, graphic symbolization, and polygons. The methodology is the same as applied in the construction of geological cross-sections from well logs, for example, except that instead of a vertical depth profile the cross-shore transect is horizontal. It is the same methodological principles used in a different orientation. Another example can be drawn from benthic ecological assessment methods, such as the Benthic Ecological Assessment for Marginal Reefs (BEAMR, Makowski and Keyes, 2011; Makowski et al., 2009), which uses in situ cross-shore transects to laterally survey the marine benthos. The advantage of lateral extrapolation of cross-shore BCCS units is that an entire satellite scene can be ecologically mapped to show dominant environments and habitats. The spatial distribution of the units laterally spread from the cross-shore transect thus permit a visual perspective of the coastal belt that could not be obtained from the transects alone.

This kind of effort, however, is not without difficulties. The procedure assumes a basic level of expertise on the part of the coastal researcher to interpret essential features shown in the satellite scene. Visual interpretation of satellite scenes uses the same analytical techniques that are applied in aerial photo interpretation. Familiarity of the elements of image interpretation thus involve perception of location, size, shape, shadow, tone/color, texture, pattern, height/depth, and site/situation/association. Application of these parameters facilitates the distributive process going from cross-section to lateral extrapolation, with the end result being the classification or characterization of larger swaths of coastal satellite scenes that would otherwise not be possible from just the cross-shore transects. Lack of ability on the part of the researcher to perform basic image interpretation would thus thwart any attempt to construct a cross-shore transect using the BCCS, and, likewise, lateral extrapolation of units would not be possible. Whereas this could be a potential drawback that may confound some attempts to apply the BCCS, it is anticipated that most current coastal researchers have been exposed to some basic principles and practices of image interpretation.

Creation of Shore-parallel Mapping Units

The creation of shore-parallel mapping units is dependent on what can be graphically shown at the scale of the satellite scene and thus the ability of the interpreter to discern the desired features, whether they be archetypes or sub archetypes. The determination is scale dependent to a large degree and also depends on the purpose of characterization for the coastal belt. Scale of the coastal belt mainly depends on the purpose of the project and what is desired to be shown or illustrated. Such as the case, image scene scale is actually determined by the geomorphological-ecological features to be studied or mapped.

The coastal belt along the tropical Eritrean coast (Figure 1) was deliberately selected as an example of the kind of complexity that can occur on an otherwise rather simplistic coastal setting where desert ecosystems come to the shore. In the example of this coastal belt, complex setups with coastal wetlands in an otherwise arid climatic regime occur. Figure 1a shows what may appear to be a somewhat confusing array of geomorphological-ecological systems, but when the image scene is transected cross-shore, it becomes a relatively easy procedure to identify the major environments making up the scene. The apparent complexity is simplified by the identification of archetype and sub archetype catenary sequences via two demonstration transects. The northern transect provides the more general characterization of the coastal belt with the following concatenation: 2,4CrcpBecaWma,mr,slUde. This is a fair characterization of the coastal belt without intervening Flat and Lagoon archetypes, as shown in the southern concatenation CES of 2,4CrcpFsaCrfrWma,slLopWmr,slUde. These transects are provided to show the type of complexity that can occur over a short distance in the same coastal belt, and, ultimately, the researcher has to decide which transect best fulfills the needs of the study. The choice is determined by the level of detail that is required and whether one wishes to characterize the mainland shore of the coastal belt or emphasize the presence of islands, flats, and lagoons. For a broader view that provides general characterization of small coastal belts with wetlands in this tropical desert environment, it would be prudent to zoom out of this image scene and determine the frequency of occurrence of this type of coastal belt. Advantages and disadvantages of such action would depend on the purpose of the research in the study. For example, details of the Wetland archetype (i.e. mangrove, marsh, and salina sub archetypes) spatial distributions are too complicated to polygonise at this scale, so they are simply shown graphically by annotated arrows. Another feature in coastal belt settings is that cross-shore archetypes and sub archetypes tend to repeat in a catenary fashion, as shown in Figure 1 where there is an offshore island.

Figure 2 shows a typical barrier coastal belt, in this case on the subtropical monsoon SE coast of Madagascar. A clear unimpeded view of this mainland barrier coastal belt may be seen in Figure 2a and should be used for comparison with the annotated version in Figure 2b. The concatenation presented in the large shoreward facing red arrow typifies the catena for this type of coastal belt, whether tropical or extratropical. The scale of the image does not readily allow the construction of polygons for laterally extrapolated units that tend to have alongshore extent rather than cross-shore spread, as in the case of Beach and Dune archetypes. As a consequence, the Beach and Dune archetypes are shown in Figure 2b as colored lines to define their spatial distributions. The estimated width of the mainland barrier is indicated by a narrow double-ended yellow arrow northward of the red BCCS transect arrow. The barrier is characterized by wetland mangroves and marshes and are so identified in green linage and lettering. The Wetland archetype is then terminated landward by the Lagoon archetype, which is marked by red lines to show a closed lagoon system laterally extended from the cross-shore BCCS transect. Because these ecological units are so clearly visible and of broad extent, it was possible to show their spatial distributions as polygons. Had these archetypes been of narrower cross-shore extent, it might not have been possible to construct polygons to show their areal distributions. As shown in the large, red BCCS cross-shore transect arrow, the mainland barrier and lagoons of this coastal belt are backed by an Upland archetype in the form of grasslands. Because of the complex patterns in the texture, tone, and pattern of the image, the Upland archetype could have been subdivided into other units besides grasslands; however, for the purposes of this study, the intent was to show the general nature and extent of the coastal belt hinterland. Polygons were not created, and only lettering was used to show the lateral extrapolation of the cross-shore grassy upland designation. Figure 2b shows that a variety of graphic techniques may be used to illustrate the results of laterally extrapolating the archetypes and sub archetypes from the cross-shore BCCS transect that, in effect, classified the biophysical ecology of this coastal belt.

Figure 3 on the SW coast of Wales shows unannotated and annotated views of a wide abrasion platform that fronts a cliff archetypical coastal scene. Figure 3a should be compared with Figure 3b to discern how the cross-shore sub archetypes were extrapolated laterally along the coastal belt. The abrasion platform fronting the coastal cliffs is large and easily identified on the basis of texture, tone, and pattern to form the polygon on the south side of the cross-shore transect. The Cliff and Upland archetypes are easily identified on the basis of topography and image properties. The cliffs and mixed forest-grassland uplands were simply denoted by lettering because the lateral extension from the concatenated arrow was obvious. Not so obvious, however, was the distinction between the beach rampart and the rock talus at the foot of the sea cliffs. These features are represented as polygons extrapolated on the SE side of the large, red cross-shore BCCS transect arrow. Although shown as polygons, this is only a rough approximation and illustrates potential lateral mapping extensions of sub archetypes identified from the cross-shore catena. It is also important to remember that limits to image interpretation occur in certain circumstances, and it can become difficult to differentiate similar patterns. Such an example can be found in Figure 3, where the geological materials are the same and yet only subtle differences in topographic form to differentiate rock platforms from beach ramparts from rock talus occur.

Figure 4 from the west coast of subpolar Hudson Bay in NE Manitoba, Canada, shows a complex barrier island system with wetland and strandplain sequences. The cross-shore concatenation that typifies this type of coastal belt catena is extrapolated using polygons and lettering to indicate the spatial distributions of sub archetypes that extend laterally alongshore. The barrier island sub archetype (indicated by lettering in the upper righthand corner of the image scene) occurs offshore and is backed shoreward by shallow mudflats, which are shown by the Fmu sub archetype polygon extending parabathically north of the red cross-shore BCCS transect arrow. The units in Figure 4b should be compared to the image patterns in Figure 4a for confirmation of the interpretation. The Beach and Dune archetypes cannot be easily differentiated by graphics because of the large scale of the image. They are rather identified as a cross-shore transect couplet but cannot be mapped in an effective manner at this scale. Other biophysical features in the cross-shore BCCS transect (i.e. wetland marshes and beach ridge strandplains) can be readily classified as north-south trending patterns emerge. The repeating cross-shore archetypical concatenations can then be extrapolated laterally from the cross-shore BCCS transect arrow. Overall, the BCCS characterization of this coastal belt provides an easy way to understand cross-shore classification, which can then be extrapolated in a shore-parallel direction to further define the ecological properties of this coastal region.

CONCLUSION

The Biophysical Cross-shore Classification System (BCCS) was proposed as a possible alternative means of coastal classification based on the use of satellite images (Finkl and Makowski, 2020a,b). Instead of characterizing alongshore properties in terms of a first-occurring feature, the BCCS posited a cognitive approach that involves the recognition of cross-shore geomorphological-ecological units moving landward from the sea to the shore. Commonly occurring units worldwide, from equatorial to polar regions, were referred to as archetypes and included features such as Barrier, Beach, Beach Ridge, Cliff, Coral Reef, Delta, Dune, Flat, Ice, Lagoon, Mountain, Rock, Till, Upland, and Wetland (Table 1). Subdivisions of these broad naturally occurring archetypes are then referred to as sub archetypes (Table 1). This study demonstrated that the cross-shore BCCS transects used to classify coastal belts, as observed in satellite images, could be extended laterally in a shore-parallel direction by extrapolation of cross-shore catenas (concatenations of archetypes and sub archetypes). Results of the procedure were graphically illustrated as mapping units in figures from low latitude, middle latitude, and subpolar zones from the lateral expansion of the cross-shore transect (Figures 14). Although the extrapolation can become limited by the dimensions of Google Earth Pro's satellite scene boundaries, images can be combined to form longer shore-parallel swaths of a particular coastal belt. Overall, the lateral expansion of BCCS cross-shore transects through the extrapolation of classified biophysical features, thus provides a means of mapping coastal belts using interpretive techniques for satellite images.

LITERATURE CITED

LITERATURE CITED
Campbell,
J.B.
and
Wynne,
R.H.,
2011
.
Introduction to Remote Sensing.
New York
:
Guilford Press
,
667
p.
Costa,
B.,
2019
.
Multispectral acoustic backscatter: How useful is it for marine habitat mapping and management?
Journal of Coastal Research
,
35
(
5
),
1062
1079
.
Das,
S.;
Behera,
S.C.;
Kar,
A.;
Narendra,
P.,
and
Guha,
S.,
1997
.
Hydrogeomorphological mapping in ground water exploration using remotely sensed data—A case study in Keonjhar district, Orissa.
Journal of the Indian Society of Remote Sensing
,
25
(
4
),
247
259
.
Fairbridge,
R.W.,
2004
.
Classification of coasts.
Journal of Coastal Research
,
20
(
1
),
155
165
.
Finkl,
C.W.,
2004
.
Coastal classification: Systematic approaches to consider in the development of a comprehensive scheme.
Journal of Coastal Research
,
20
(
1
),
166
213
.
Finkl,
C.W.
and
Makowski,
C.,
2019a
.
Classification of coastal seafloor geomorphological features.
In:
Finkl,
C.W.
and
Makowski,
C.
(eds.),
Encyclopedia of Coastal Science, Volumes 1 and 2,
2nd edition.
Encyclopedia of Earth Sciences Series (EESS)
,
Cham, Switzerland
:
Springer Nature,
pp.
540
549
.
Finkl,
C.W.
and
Makowski,
C.,
2019b
.
Nearshore geomorphological mapping.
In:
Finkl,
C.W.
and
Makowski,
C.
(eds.),
Encyclopedia of Coastal Science, Volumes 1 and 2
, 2nd edition.
Encyclopedia of Earth Sciences Series (EESS)
,
Cham, Switzerland
:
Springer Nature,
pp.
1243
1265
.
Finkl,
C.W.
and
Makowski,
C.,
2020a
.
Latitudinal and situational zonation of shore-normal catenary sequences observed from satellite images using the Biophysical Cross-shore Classification System (BCCS).
Journal of Coastal Research
,
36
(
2
),
1
14
.
Finkl,
C.W.
and
Makowski,
C.,
2020b
.
The Biophysical Cross-shore Classification System (BCCS): Defining coastal ecological sequences with catena codification to classify cross-shore successions based on interpretation of satellite imagery.
Journal of Coastal Research
,
36
(
1
),
1
29
.
Fuller,
R.M.;
Parsell,
R.J.;
Oliver,
M.,
and
Wyatt,
G.,
1989
.
Visual and computer classifications of remotely-sensed images. A case study of grasslands in Cambridgeshire.
International Journal of Remote Sensing
,
10
(
1
),
193
210
.
Gardin,
S.;
van Laere,
S.M.J.;
van Coillie,
F.M.B.;
Anseel,
F.;
Duyck,
W.;
de Wulf,
R.R.,
and
Verbeke,
L.P.C.,
2011
.
Remote sensing meets psychology: A concept for operator performance assessment.
Remote Sensing Letters
,
2
(
3
),
251
257
.
Kelletat,
D.,
1989
.
The question of “zonality” in coastal geomorphology - with tentative application along the East Coast of the USA.
Journal of Coastal Research
,
5
(
2
),
329
344
.
Kelletat,
D.
(ed.),
1995
.
Atlas of Coastal Geomorphology and Zonality
.
Journal of Coastal Research
, Special Issue No. 13
,
286
p.
Kelletat,
D.H.;
Scheffers,
A.M.,
and
May,
S.M.,
2013
.
Coastal environments from polar regions to the tropics: a geographer's zonality perspective.
Geological Society
,
London
,
Special Publications,
388
, pp.
33
57
.
Klemas,
V.V.;
Dobson,
J.E.;
Ferguson,
R.L.,
and
Haddad,
K.D.,
1993
.
Coastal Land Cover Classification System for the NOAA Coast-watch Change Analysis Project.
Journal of Coastal Research
,
9
(
3
),
862
872
.
Lasaponara,
R.
and
Masini,
N.,
2011
.
Remote sensing in archaeology: From visual data interpretation to digital data manipulation.
In:
Lasaponara,
R.
and
Masini,
N.
(eds.),
Satellite Remote Sensing. Remote Sensing and Digital Image Processing, Volume 16.
Dordrecht
:
Springer
, pp.
3
16
.
León-Pérez,
M.C.;
Hernández,
W.J.,
and
Armstrong,
R.A.,
2019
.
Characterization and distribution of seagrass habitats in a Caribbean nature reserve using high-resolution satellite imagery and field sampling.
Journal of Coastal Research
,
35
(
5
),
937
947
.
Makowski,
C.,
2014
.
Development and Application of a New Comprehensive Image-Based Classification Scheme for Coastal and Benthic Environments along the Southeast Florida Continental Shelf.
Boca Raton, Florida
:
Doctoral Dissertation
,
303
p.
Makowski,
C.
and
Finkl,
C.W.,
2016
.
History of modern seafloor mapping.
In:
Finkl,
C.W.
and
Makowski,
C.
(eds.),
Seafloor Mapping along Continental Shelves: Research and Techniques for Visualizing Benthic Environments. Coastal Research Library (CRL), Volume 13
,
Dordrecht, The Netherlands
:
Springer International Publishing
, pp.
1
47
.
Makowski,
C.;
Finkl,
C.W.,
and
Vollmer,
H.M.,
2015
.
Geospatially Integrated Seafloor Classification Scheme (G-ISCS): A new method for cognitively interpreting benthic biogeomorphological features.
Journal of Coastal Research
,
31
(
2
),
488
504
.
Makowski,
C.;
Finkl,
C.W.,
and
Vollmer,
H.M.,
2016
.
Classification of continental shelves in terms of geospatially integrated physio-graphic realms and morphodynamic zones.
Journal of Coastal Research
,
32
(
1
),
1
34
.
Makowski,
C.;
Finkl,
C.W.,
and
Vollmer,
H.M.,
2017
.
Geoform and landform cassification of continental shelves using geospatially integrated IKONOS satellite imagery.
Journal of Coastal Research
,
33
(
1
),
1
22
.
Makowski,
C.
and
Keyes,
P.,
2011
.
Using the Benthic Ecological Assessment for Marginal Reefs (BEAMR) method to quantify nearshore reef conditions in the southeast Gulf of Mexico.
Journal of Coastal Research
,
27
(
3
),
428
440
.
Makowski,
C.;
Prekel,
S.E.;
Lybolt,
M.J.,
and
Baron,
R.M.,
2009
.
The Benthic Ecological Assessment for Marginal Reefs (BEAMR) method.
Journal of Coastal Research
,
25
(
2
),
515
522
.
Peel,
M.C.;
Finlayson,
B.L.;,
and
McMahon,
T.A.,
2007
.
Updated world map of the Köppen-Geiger climate classification.
Hydrology and Earth System Science
,
11
,
1633
1644
.
doi:10.5194/hess-11-1633-2007
Rasid,
H.
and
Pramanik,
M.A.K.,
1990
.
Visual interpretation of satellite imagery for monitoring floods in Bangladesh.
Environmental Management
,
14
(
6
),
815
821
.
Richards,
J.A.
and
Jia,
X.,
1999
.
The interpretation of digital image data.
In:
Remote Sensing Digital Image Analysis.
Berlin
:
Springer
, pp.
75
88
.
Schowengerdt,
R.A.,
1983
.
Techniques for Image Processing and Classification in Remote Sensing.
New York
:
Academic
,
248
p.
Short,
A.D.,
2006
.
Australian beach systems—nature and distribution.
Journal of Coastal Research
,
22
(
1
),
11
27
.
Short,
A.D.
and
Woodroffe
,
2009
.
The Coast of Australia.
Melbourne
:
Cambridge University Press
,
288
p.
Snook,
P.W.;
Winterberger,
K.C.;
Merritt,
N.E.,
and
Czaplewski,
R.L.,
1987
.
Comparison between digital and manual interpretation of high altitude aerial photographs.
Photogrammetric Engineering and Remote Sensing
,
53
(
5
),
531
534
.
Wang,
M.;,
Ahmadia,
G.N.;
Chollett,
I.;
Huang,
C.;
Fox,
H.;
Wijonarno,
A.,
and
Madden,
M.,
2015
.
Delineating biophysical environments of the Sunda Banda Seascape, Indonesia.
International Journal of Environmental Research and Public Health
,
12
(
2
),
1069
1082
.
doi:10.3390/ijerph120201069
Wang,
Y.;
Bonynge,
G.;
Nugranad,
J.;
Traber,
M.;
Ngusaru,
A.;
Tobey,
J.;
Hale,
L.;
Bowen,
R.B.,
and
Makota,
V.B.,
2003
.
Remote sensing of mangrove change along the Tanzania coast.
Marine Geodesy
,
26
(
1–2
),
35
48
.
Woodroffe,
C.D.,
2002
.
Coasts: Form, Process and Evolution.
Cambridge
:
Cambridge University Press
,
623
p.