The survey was conducted in the mountainous Upper Galilee, in a region that had not previously been surveyed in the Israel Archaeological Survey Maps, and in areas where ancient sites had not been recorded. The objectives of the survey were to practice systematic archaeological documentation in areas for which there was no prior information on sites, and to test an archaeological predictive model prepared for this region (Wachtel 2018; Wachtel et al. 2018). Archaeological predictive modeling assesses the probability of finding new archaeological sites in areas that have not previously been surveyed, based on statistical analysis that weighs up and analyzes the locations of recognized settlement sites and their environmental attributes. The model helps screen the entire area for ‘environmental niches’ that are suitable for settlement, and where there is a relatively high probability of locating ancient sites. The student surveyors were divided into four teams of four to five members, each headed by an experienced surveyor. In the four-day survey, seven area, in which the discovery of new sites based on the archaeological predictive model was considered highly probable, were examined (Fig. 1: Nos. 1–7). The spatial samples covered areas ranging between tens to hundreds of dunams and contained no registered antiquity sites and no sites documented in research literature or previous surveys (Frankel et al. 2001).
 
The two complementary surveying methods implemented in the survey were collection-units and points-density surveying. The collection-units survey was conducted by walking slowly across a define area at regular distances of 10–15 m between surveyors, looking for surface finds such as pottery or flint items. When such finds were discovered, with a minimum threshold of three finds within a 50 m transect, a collection unit was opened and systematically surveyed. The area of the collection units ranged from one to five dunams, depending on the type of landscape and the distribution of the finds. Each collection unit was marked on a 1:2500 scale aerial photograph and its boundaries were measured precisely with a portable GPS device. Each unit was recorded on a survey sheet, including the unit’s name, area characteristics and degree of surface visibility, as well as the collection time, number of surveyors and description of the ancient remains and finds. The basic reference unit in the survey is not a ‘site’, but a ‘collection unit’. Each ‘site’ can be made up of one or more collection units, depending on the distribution of the surface finds in the field. This method, recently used in a survey of settlement sites in the mountainous Upper Galilee (Sabar 2017; Wachtel 2018), was employed in the current survey.
The points-density survey was conducted in November 2018 at sample points in an area surveyed in Nahal Sekhvi in Biriya Forest. The survey examined sample circles (radius 1.5 m) spaced c. 20 m apart. In each sample circle, all the pottery and flint items were counted, and diagnostic finds were collected. This method provided a spatial analysis of finds density, which included locating the main areas of human activity and their margins.
The results of the two survey methods were analyzed using geographical-computational and archaeological tools. The environmental data, the location of the collection units and sample circles, the areas of the sites and the density of the finds were analyzed with a GIS tool, using ArcGIS Pro software, in the Spatial Archaeology Laboratory at the Hebrew University of Jerusalem’s Institute of Archaeology. The collected items were sorted and classified by their typo-technological characteristics, following standard practice.
The survey results are presented here according to the areas examined, which represent different environments in the mountainous Galilee region. The model points to a high probability of finding ancient remains in most of the area sampled, apart from Mount Safsuf.
 
Biriyya Forest (map ref. 244685–5455/764560–5700; Fig. 1: No.1). The survey area (c. 200 dunams) is a rocky, moderately hilly area in Biriyya Forest, delimited by the upper wadi of Nahal Sekhvi in the east and the deep Nahal ‘Amud gorge in the west (Fig. 2). The area is planted with pine trees interspersed with low-growing herbaceous and shrubby vegetation, with remnants of sparse Mediterranean maquis; the general ground surface visibility was defined as moderate.
In December 2017, five collection units (total area c. 20 dunams) were defined in the northern part of the area, in non-contiguous areas on the forest margins. They yielded scattered, sparsely dispersed potsherds, mostly of Kefar Hananya Ware vessels dating from the Roman and Early Byzantine periods, as well as stone-clearance heaps and quarrying marks, with no clear architectural remains. The area lies 2.5–3.0 km away from the Roman and Byzantine sites at Qadita and Meron. The remains seem to indicate an agricultural area based on pockets of land farmed between the rocks, where a small settlement may have existed, possibly a farmstead or a seasonal site.
In the southeastern part of the area, near the wadi bed of Nahal Sekhvi to the west, a previously unknown Late Chalcolithic settlement was discovered (map ref. 245420/764850; Fig. 3). In the first survey, two collection units were examined (total area c. 10 dunams), although it was clear that the site extends beyond these units. In the collection units, walls tops were documented, and dozens of Late Chalcolithic sherds were collected, including rims, bases, triangular handles, and rope decoration, as well as flint tools, including adzes and chisels, and stone artifacts. The pottery assemblage consists mostly of dark-colored Golan Ware, due to the basaltic element in the clay, although lighter colored pottery was also collected. The Golan Ware sherds date the site generally to the second half of the fifth–early fourth millennium BCE. The Chalcolithic site was named Nahal Sekhvi.
In November 2018, a density survey was conducted in sample circles in an area covering c. 90 dunams around the collection units surveyed in 2017, during which a full count of finds was made from a sample of 64 circles (Figs. 4, 5). The density maps show a high concentration of finds in the area of the collection units, which probably mark the center of the site, although a high density of finds was also documented to the north, south and east of these units. The settlement’s built-up area probably covered an area of 40–50 dunams, although it may not have been entirely built-up.
 
Mount Safsuf (Namura Farm; map ref. 239564–714/767555–685; Fig. 1: No. 2). In December 2017, an area (c. 55 dunams) was surveyed on a gentle slope on the northeastern edge of Mount Meron, southwest of a broad saddle lying between Mount Meron and Mount Safsuf, opposite the Namura Farm (Fig. 6). The survey area is bordered by deciduous orchards, dirt tracks, and ravines with dense Mediterranean maquis vegetation. The ground surface visibility is low, apart from an area in the center, where the ground surface visibility is fairly good as it is grazing land with herbaceous vegetation. In the center, where the slope is more moderate, five collection units were opened (total area c. 10 dunams). The finds in these collection units were sparsely dispersed on the surface and consist predominantly of Roman and early Byzantine Kefar Hananya Ware, with a few sporadic Ottoman potsherds. The architectural remains surveyed are mostly low terrace walls and stone field walls, whilst some of the wall tops observed may not be agricultural walls. A rock-hewn winepress was documented higher up the slope, beyond the survey borders. The remains in the area are similar to those observed in the northern part of the Biriyya Forest and they may be associated with a permanent settlement, such as an isolated or seasonal farmstead lying in the heart of agricultural land.
 
Mountainous ridge north of Hiram Junction (map ref. 237880–8220/770600–920; Fig. 1: No. 3). The area extends over a moderate northward-facing slope of a rounded hilltop (836 asl; c. 70 dunams) east of Kibbutz Sasa and north of the Hiram Junction (Fig. 7). North of the area, the slope drops steeply down to the Nahal Tapuhim basin, and the area’s southern limits are defined by orchards. The area lies c. 350 m west of a registered Iron Age I site, designated Hiram Junction site (Frankel et al. 2001:37, Site 291; Wachtel 2018:143–144). The ground is covered with Mediterranean maquis with scrubland vegetation and ground surface visibility is poor. Seven contiguous collection units were defined (total area 44 dunams). The surface finds were sparsely scattered and consisted mainly of Kefar Hananya Ware potsherds. Agricultural terrace walls and stone-clearance heaps were also documented. The remains are similar to those surveyed in the Biriyya Forest and Mount Safsuf, and probably attest to farmlands worked in the Roman or early Byzantine period where there may have been a small settlement or a seasonal presence related to the nearby site at Sasa (Wachtel 2018:148–153; Frankel et al. 2001:37, Site 290).
 
Mount Sasa (map ref. 236430–500/770865–1210; Fig. 1: No. 4). Two adjacent survey transects extending from the foot of the southern slope of Mount Sasa, along Road No. 89, to its summit, were surveyed (total area 80 dunams; Fig. 8). The slope is naturally terraced with a few agricultural plots, and a seasonal pool (Sasa Nature reserve) collects at the base of its southeastern side. Due to poor ground surface visibility resulting from the dense Mediterranean vegetation, the survey was limited to woodland clearings.
Four separate collection units were covered in the two transects (total area c. 10 dunams). Several flint items collected near the Sasa Reservoir include bifacials dating from the Neolithic–Chalcolithic continuum; similar items were previously documented in the area by the Upper Galilee Survey team (Frankel et al. 2001:36, Site 289). A few Kefar Hananya Ware and glazed medieval potsherds were collected, and several wall tops were documented. Several flint items were recovered from the northwestern collection unit (map ref. 236450/771200). East of the Mount Sasa summit, a limekiln with several large, roughly dressed stones lying sideways beside it, was documented (it is marked as a cistern on topographical maps; map ref. 236772/771027; Fig. 8, marked with green dot). The pottery sherds include several Kefar Hananya rims and Ottoman Rashaya el-Fukhar pottery. Although the combined remains do not form a clear picture, it appears that the area on the fringes of Sasa’s farmlands had various uses in different periods. The many flint items near Sasa pool nature reserve may indicate the presence of a small prehistoric site, but it is not possible to determine whether it was a settlement, or a knapping site for the production of bifacial tools.
 
Spur east of Tel Rosh (map ref. 232100–470/771665–830; Fig. 1: No. 5). An area (c. 100 dunams) was surveyed on a saddle and a moderately sloping spur c. 500 m east of Tel Rosh. Ground surface visibility is good, probably resulting from land clearance and ground preparations in recent decades. Nine collection units (total area c. 50 dunams; Fig. 9) yielded flint items and a few fragments of worn non-diagnostic pottery. Most of the finds were retrieved on and near the saddle, and they include flint blades and a fragment of an adze dated to the Neolithic–Chalcolithic continuum. The finds in the field attest to limited human activity in various periods.
 
Western Peqi‘in Valley (map ref. 228360–660/766365–7500; Fig. 1: No. 6). The survey area covers the western part of the Peqi‘in valley, near Route 864 and directly east of Moshav Hosen. The valley is planted with olive groves, and field crops are farmed on the banks of Nahal Peqi‘in; the ground surface visibility is fairly good. A slope bordering on the north side of Road 864 and to the west of Horbat Sirim was also surveyed.
The 19 collection units covered (total area c. 50 dunams; Fig. 10) yielded an abundance of varied potsherds and a few flint items, most of them worn and not datable. The finds were probably not retrieved in their original location, but were washed in from sites on the edge of the Peqi‘in Valley, such as Horbat Gayis, Horbat ‘Eved and the site of ‘En Tiriya (Aharoni 1957; Wachtel 2018:189–190; Frankel et al. 2001:33, Sites 238–240) and deposited in the alluvium of Nahal Peqi‘in. On the northern slope above the valley, agricultural terraces and stone-clearance heaps were found, as well as a sparse scattering of non-datable potsherds.
 
Kisra-Sumi‘a (map ref. 229780–950/765170–420; Fig. 1: No. 7). The survey area (c. 50 dunams) extends across a hill to the east of the village of Kisra-Sumi‘a, bounded by dirt tracks and olive groves that belong to nearby villages. The hillsides contain exposed rocky outcrops as well as orchards and sparse woodland; in general, the ground surface visibility is medium.
The six collection units surveyed (total area 26 dunams; Fig. 11) yielded Middle Bronze Age, Iron Age (?), Hellenistic, Roman–Byzantine and Ottoman pottery, as well as flint items—including an Early Bronze Canaanite blade—and a single basalt item. Several wall foundations, installations and rock-cuttings were also found. A cluster of walls documented on top of the hill were built of fieldstones and preserved for one course (map ref. 229895/765270).
 

All the areas surveyed in Upper Galilee yielded finds indicating human activity at spots where such activity had not previously been documented. However, significant remains of only one settlement site—the Late Chalcolithic site in the Biriyya Forest—were discovered. This site predates the Bronze and Iron Age settlement sites on which the prediction model was based (Wachtel et al. 2018). A small settlement may also have existed on the outskirts of the village of Kisra-Sumi‘a in these periods. Regarding the pottery, the albeit sparse presence of Roman–early Byzantine Kefar Hananya Ware is noteworthy in a variety of environments and areas, most of which are related to agricultural activity. These finds may attest to isolated farms or seasonally farmed smallholdings belonging to nearby villages—a previously less acknowledged settlement pattern in this period.

The Late Chalcolithic site in Nahal Sekhvi, which covers tens of dunams, is a good illustration of the ability of a systematic, prediction model-based survey to discover significant, yet hitherto unrecognized, sites with a minimal investment of time and resources. Since the total area surveyed in the seven spatial samples is less than one square kilometer, the discovery of a large settlement site in such a limited area is a significant achievement. The site may be one of the largest Chalcolithic sites in eastern Upper Galilee, joining a few contemporary sites, located mainly on the fringes of the basaltic Dalton Plateau (Shalem 2003; Frankel et al. 2001). An interesting feature of this site is, that although it is located on calcareous rock about 4 km from the nearest basalt outcrops (in the Dalton Plateau), the dominant pottery, at least the surface pottery, is made from a basaltic fabric designated Golan Ware, due to its predominance at sites located in the basaltic regions of the Golan (Epstein 1998). Whilst the penetration of this ware into the heart of the mountainous Galilee has already been observed (Shalem 2003), the Nahal Sekhvi site is a prime example of this process, with all the cultural implications it entails.