Somogyi Múzeumok Közleményei 18. (Kaposvár, 2008)

FEKETE CSANÁD: Predictive archaeological modelling in Somogy county

3-6%, 6-9%, 9-12%, 12-15% and above 15%. Most of the sites are located on the slope that are less than 12%. None of the sites can be found on the slope above 12% with their whole territory and there are only 7 sites that are partly situated there. At the end I digitized some elements of the soil map. Because of the scale of 250 metres to the centimetre, the soil map does not fit into the topographical map with larger scale. I only digitized the polygons of the alluvial deposit that typical of the inundation area, which surprisingly fit into the 70-metre-bufferzone measured from the watercourses.

12 (Figure 5.) The data of the soil map was only used in part for the model because of the already mentioned inaccuracy of the map and partly because the sites do not seem to prefer only one or some soil type. Less than 2/3 of the sites can be found on the same soil environment. This result does not make the whole soil map suitable to be integrated into the model, hence its conclusions are detailed in the next chapter.

13 The final result map indicates the areas that has southern-eastern aspect, that are within 300 metres, but out of 70 metres, measured from the waterflows and that can be found on the slope less than 12%. (Figure 6.) The probability model, based on the archaeological sites of the sample area and the physical paramètres of the geographical environment, covers the whole territory of 14 and partly 33 sites. 2 out of the 47 modelled sites are hardly recognizable from their parts affected by the model. Whereby there are 45 sites (out of 60) on the sample area that identifiable fit into the predicted territory with archaeological interest. This means 75% probability. (Table 4.) The conclusions of resctricting the areas with archaeological interest Some regularity can be drawn from the analysis of the location of the archaeological sites. Much more sites can be found on the surface with southern-eastern aspect, which means that humans preferred the windless, sunny surfaces to settle down.

14 This fact can be explained by the dominant wind direction of eastern Transdanubia

15 and the N-S course of the valleys of Somogy county Most of the sites, located on northern-western aspect terrain have finds from more ages, but almost all of them has Middle Age surface finds (Árpádian Period, late Middle Ages). These sites appear to group in a comparatively confined area, which could possess favourable settling conditions despite of its aspect. 12 The process of the soil formation was prevented by the periodically repeating floods and the their sediment. 13 The statistical distribution of the paramètres can be seen in Table 1. 14 Outer-Somogy has the favourable condition, related to the character of the terrain, that its long, sloping surfaces with southern aspect are windless. 15 One of the most significant feature of Hungary's wind climate is that the dominant wind directions are affected by the Alps and the Carpathians. The dominant wind courses of western and middle Hungary are northern and northeastern. In: PÉCZELY 1979, 263. Determining the distance from the watercourses needed a strongly compromised solution. Neither the 70-metre-buffer that reflects to the humid, marshy region, nor the 300-metre-buffer that reflects to the maximum distance from the watercourses, cannot consider adequate answers. The first one seems to be more appropriate because its distance equals to the uninhabited zone, marked by the alluvial deposit of the floodplain. It is more a hard task to determine the zone where is no site out of. Some of the sites are, partly or entirely, out of the 300-metre-zone, still this is the best possible solution to narrow the inhabited area. There is no point in considering a distance less than 300 metres, because of the increasing loss of the sites that would have been suffered. I created a map with 400-metre-buffer, which produced a hardly better result, but there were significantly larger areas without sites too. A few of the sites that are out of the 300-metre-zone (Bodrog 8., 9., 10.) might have slipped out of the predicted area because of the defective hydrographical data. According to the observation of fieldwalking there used to be a spring in this territory. Its former existence can be proved by a group of willow that hides in a wet depression. This feature is also marked on the topographical map. Analysing the releif map confirms this conclusion: a small and deep basin connects to the valley of Pogány-völgyi-víz from the NE. It is an important common denominator of these 3 sites (and one more that situates NE direction from them) that all of them possess remains of metallurgical dross, which is the unambiguous evidence of iron metallurgy. The 4 sites of the sample area, which are away from watercourses have only scattered finds, mostly from the Middle Ages. These sites could be dwelled for short period and used only as temporary accomodation, so the proximity of the watersupply was not needed for their development. Between the slope of the terrain and the inhabitable areas have a strong connection. After some pouring rain mud runs down on the surface, whose slope above 12%. This is the area that is unsuitable for settlements because the sides of the valleys, which rise above the more or less flat floodplain of the streams, are the most sloping. There are areas that are fit for settling in every respect, but no sites can be found on them presumeably because of the significant slope. The soil map classifies the soil types according to their chemical reaction, water accomodation and water content attributes, so did I. (Table 3.) As for the reaction attribute, 34 sites of the sample area can be found on slightly acidic soil, 16 on neutral soil, 1 on strongly acid, calciferous soil, while 6 on the alluvial deposit of the floodplain. The chemical reaction of the soil has a strong effect on plants and it determines the variety of them. The neutral reaction offers the best options for the nutrition income of plants and the microorganisms living in the soil.

16 The chemical reaction of STEFANOVITS 1975, 100-101.