**1. Introduction**

Coastal sediment dynamics is always highly complex, and, thus, it provides a lot of issues for geoscience investigations. An interest in nomenclature development for large clasts that started near the end of the 20th century [1] and attention to coastal hazards facilitated by the Indian Ocean Tsunami of 2004 [2] have shaped a new international research direction, namely megaclast studies [3]. Megaclasts of ocean and sea coasts have been investigated most intensively. These studies have been conducted in so different places of the world as Baja California in Mexico [4,5], North Eleuthera of the Bahamas [6,7], and Rabat in Morocco [8]. Although the works have tended to focus on only some regions like the Mediterranean (e.g., [9–13]), the global evidence of large clast accumulations has become huge already, and it continues growing. This evidence requires extension and generalization for further conceptual treatment. Previous reviews of rocky shores [14] and megaclasts [3] confirm such research is promising. Undoubtedly, coastal megaclast deposits of Quaternary age are of special importance because the evidence of such deposits from the earlier geological periods remains scarce [3,15].

Coastal megaclast deposits attract significant attention of modern researchers for two aspects. The first aspect is the development of large clast nomenclature. Somewhat coherent and somewhat alternative proposals were made by Blair and McPherson [1], Bruno and Ruban [16], Blott and Pye [17], and Terry and Goff [18]. Additionally, Cox et al. [10] offered a novel approach for measurements of roundness of megaclasts, and, thus, the noted nomenclature development should emphasize not only on grain-size categories, but also on various morphological parameters. The second aspect of coastal megaclast deposits is linked to genetic investigations and, particularly, interpretations of the evidence of past storms and tsunamis (storm-versus-tsunami origin of megaclasts is a popular and hotly-debated topic in the modern sedimentology) [4–13,19–28]. Irrespective of which of these aspects and particular opinions to follow, finding new localities coastal megaclast deposits is of utmost importance. The knowledge from the 'classical' localities like those studied in the Bahamas [6,7] and western Ireland [29–35] need extension and refinement with the information from many other localities of the world. Although megaclast studies are urgent and relatively cheap, the circle of the involved researchers remains too restricted to expect documentation of even a triple of all megaclast localities. Moreover, some of the latter can be situated in remote places travelling to where faces serious difficulties in regard to researchers time, safety, and expenses.

The dilemma of the high demand for the really global knowledge of megaclasts contrasting the geographical restriction of their research can be addressed with application of the modern remote sensing techniques. Particularly, the Google Earth Engine [36,37] seems to be promising due to by definition a big size of megaclasts that makes these well visible on satellite images. The extraterrestrial investigations of megaclasts have proven efficacy of similar approaches [16]. The objective of the present paper is to demonstrate the application of the Google Earth Engine for finding coastal megaclast deposits on the basis of several representative examples. This experience permits also to highlight perspectives and restrictions of this approach. In the other words, the focus of this paper is methodological, and it concerns chiefly virtual identification of localities, not comprehensive sedimentological description of deposits represented there. In this paper, it is also undertaken to summarize briefly the available information on megaclasts in some major regions of the Earth in order to demonstrate how finding new localities can fill geographical gaps in the knowledge of megaclast distribution along coasts.

#### **2. Methodology**

Large clasts are detrital sedimentary particles larger than 256 mm in size, according to the standard Udden–Wenthworth classification, or larger than 100 mm, according to the alternative classification proposed by Bruno and Ruban [16]. Boulders are large clasts with the size ranging between 0.10 and 1 m, and megaclasts are detrital sedimentary particles larger than 1 m in size [16]. However, it should be noted that different researchers proposed the different lower limit for this category of particles (= the different upper limit of boulders) [1,17,18]. Depending on their size, megaclasts can be subdivided into blocks (1–10 m), megablocks (10–100 m), and superblocks (>100 m) [16]. Coastal boulder deposits are distributed along coasts of seas, oceans, and great lakes and include a significant amount of large clasts (true boulders and megaclasts); these deposits associate often with rocky shores and reflect influences of storms and tsunamis and the relevant clast transport (inland, above high-water mark, and even to the cliff-top position) [13,20,27]. Synonymous terms are boulderite coined by Dewey and Ryan [21], boulder beach [19,38–40], and boulder field [8,41]. Finally, coastal megaclast deposits are coastal boulder deposits dominated by clasts larger than 1 m in size or, at least, bearing recognizable accumulations of such clasts.

The Google Earth Engine is a software that offers satellite images of different scales for the planetary surface and allows their analysis; it also incorporates well-justified cartographical basis, GIS technologies, and some other information, including photos provided by the users. This instrument can be applied successfully for solution of various geoscience tasks [36,37]. For instance, it has been used efficiently in landslide susceptibility mapping [42,43]. The most elementary function of the Google Earth Engine is visual surveying of the Earth surface at an appropriate resolution. Taking into

account that megaclasts are >1 m in size, the resolution of the available satellite images permits finding them in almost all regions and describing them preliminarily when image resolution is especially high. A significant advantage of this approach is a no-cost, quick, and geographically unrestricted search for megaclast localities, including those in difficult-to-access places. Anyway, the efficacy of the Google Earth Engine in the search for coastal megaclast deposits needs testing, and the latter was addressed in the present study. The Google Earth Pro 7.3.2.5776 version (free software) of the Google Earth Engine was employed for this purpose. Images of the maximum appropriate resolution are preferred. In some cases, resolution can be increased a bit else, but this leads to smooth contours of natural objects, i.e., to poor-visibility of megaclasts.

In order to test the application of the Google Earth Engine to coastal megaclast deposits, four localities of the latter were considered. These were found as a result of tentative visual coastline surveying with the Google Earth Engine in those areas where the occurrence of megaclasts seems to be possible, but has not been reported earlier. The localities were Hondarribia in northern Spain (the Biscay Bay coast), the Ponza Island of Italy (Tyrrhenian Sea), the Wetar Island in eastern Indonesia (Banda Sea), and the Humboldt o Coredo Bay and vicinities near the border between Colombia and Panama (eastern Pacific coast; Figure 1). These localities represent different geographical domains, namely Atlantic Europe, the Mediterranean, Southeast Asia, and Central America. For each locality, two plots are selected provisionally for more representative approach testing. When possible, these plots were selected so to demonstrate local differences of coastal megaclast deposits. Although this paper focused on finding localities, preliminary qualitative descriptions of the studied coastal megaclast deposits were provided, as this study explored the very possibility of their 'visibility' with the Google Earth Engine. The descriptions avoid details relevant to remote sensing techniques, but emphasize on the general character of the deposits, which is the primary interest of sedimentologists. In other words, the descriptions were addressed to sedimentologists, not specialists in remote sensing. Importantly, these localities were selected to demonstrate the utility of satellite images with a different resolution.

**Figure 1.** New localities of coastal megaclast deposits considered in this paper.

### **3. Results**

Four small case studies were undertaken in order to demonstrate the efficacy of the virtual examination of coastal megaclast deposits with the Google Earth Engine. Two examples deal with satellite images of exceptionally high resolution, and two other examples deal with satellite images of appropriate but somewhat limited resolution. In each case, it is also attempted to stress the relative importance of the new evidence to the regional knowledge of coastal megaclast deposits. For this purpose, the basic published information was summarized.
