The DOI matches its Crossref record; 47 of the 47 references it cites were found in a registry or an index; none of the 22 sources its claims cite is checked for support yet.
Using Archaeological Road Data to Evaluate Limits of Topography on Road Location: Roman Roads in the Near East as a Test Case
Open accessPublished version of a preprint
Abstract
The article presents a new methodological approach to understanding (post-dicting) the location of ancient roads based on an inductive topographic analysis of existing road remains. The research is focused on the analysis of a spatially highly precise dataset of Roman roads in the Near East, using selected topographic variables (maximum slope, topographic position index, vector ruggedness measure). The identified limits of the topographic variables are used to create an isotropic (direction-independent) model of ‘natural corridors of movement’ using an adapted ‘from everywhere to everywhere’ (FETE) method. The modelled movement corridors are then overlaid by the known Roman road network in the Near East, and their overlap is calculated. The overlap between the modelled corridors and Roman roads is around ~ 25%, suggesting that the selected topographic variables and slope categories have limited explanatory power in understanding the layout of the Roman road network as a whole. The performance of the isotropic model is compared to four selected slope-based anisotropic functions (Tobler, Naismith, Herzog, Llobera-Sluckin). It is shown that the isotropic model incorporating a limited number of topographic variables better explains the location of Roman roads than the selected functions. It is argued that these results suggest that additional anthropic, cultural variables have a stronger influence on the layout of the Roman road network, and an alternative approach on how to incorporate topographic and anthropic variables is proposed.
Overview
AI snapshot
Written from the paper’s full text.
TL;DR
The paper tests whether topographic variables alone can explain Roman road locations in the Levant by building an isotropic “natural corridors of movement” model from road-derived slope, TPI, and VRML limits. It finds only ~25% overlap between modelled corridors and Roman roads, and argues that anthropic/cultural factors likely had a stronger influence on road layout.
Key findings
- The isotropic models overlapped with Roman roads at ~24.5% for 90th-percentile corridors (Model A: 24.52% of all roads, 26.55% of analysed roads; Model B: 24.29% of all roads, 27.09% of analysed roads), with “certain” segments scoring slightly better at ~27%.
- Isotropic models outperformed four slope-based functions: maximum 90th-percentile overlap for slope-based algorithms was ~14.7–17% (Tobler 17.03%, Naismith 14.74%, Herzog 16.54%, Llobera-Sluckin 17.04%), i.e. 7–10% less than the isotropic models.
- Topographic analysis found 98.4% of road length below 20° slope, and 93.68% of road length (3,451.3 km) fit all three criteria: slope <20°, TPI between -80.71 and +91.17, and VRML <0.002332.
- The author concludes that the selected topographic variables have limited explanatory power for the Roman road network as a whole, and that additional anthropic, cultural variables likely drove road layout.
Methods
- Used the Itiner-e Roman road dataset for the Levant: 3,877.8 km of “certain” roads, with 118.4 km removed due to modern disturbance, leaving 3,759.4 km analysed; roads were split into 3,891 segments of maximum 1000 m.
- Calculated maximum slope, topographic position index (TPI), and vector ruggedness measure local (VRML) from FABDEM (~34 m), then calibrated conductivity surfaces using NPDI against test-case roads around Jerusalem–Lydda–Emmaus.
- Modelled natural corridors with an adapted “from everywhere to everywhere” (FETE) method: 100 random points per run, 50 runs, 495,000 LCPs, aggregated in a 1000-m neighbourhood; compared against Tobler, Naismith, Herzog, and Llobera-Sluckin slope-based functions using simplified FETEs (50 source points, 10 iterations, 24,500 LCPs each).
Limitations
- Stated: the selected topographic variables perform poorly in regions of low topographic variability (e.g., coastal plain, Transjordanian plateau), where LCPs run straight and parallel; the N–S shape of the computational region biases corridors toward N–S orientation; edge effects affect north-eastern and southern extremes; and low spatial accuracy or missing data in parts of the road dataset may reduce overlap.
- Stated: the 1000-m neighbourhood used to aggregate LCPs may miss corridors that run parallel to roads but outside that buffer, and part of the measured overlap is caused by accidental perpendicular crossings.
- Evident: the study is limited to the Roman Levant, with the southern Levant overrepresented due to research bias; slope-based comparisons used simplified FETEs due to computational constraints, which may affect direct comparability with the full isotropic models.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
0 of 22 cited sources checked · 0 supported · 0 flagged
- Findingown data
The overlap between the modelled natural corridors of movement and the Roman road network is around 24.5% for the 90th percentile corridors.
“The overlap with Roman roads is not very high at around ~ 24.5% for the 90 percentile corridors, although the 'certain' road segments score slightly better at ~ 27%.”
- Also stated in: § Abstract · plain
“The overlap between the modelled corridors and Roman roads is around ~ 25%”
Quote in the paperSubstantial claim - Also stated in: § Abstract · plain
- Findingown data
The isotropic model incorporating a limited number of topographic variables better explains the location of Roman roads than the selected slope-based anisotropic functions.
“It is shown that the isotropic model incorporating a limited number of topographic variables better explains the location of Roman roads than the selected functions.”
- Also stated in: § Natural Corridors of Movement · plain
“The overlap with Roman roads is weaker for all four algorithms, with maximum overlap in 90 percentile frequency category between ~ 14.7 and 17%, so 7-10% less than in the isotropic models.”
Quote in the paperSubstantial claim - Also stated in: § Natural Corridors of Movement · plain
- Findingown data
The 'certain' Roman road segments have an overlap with modelled corridors of about 27%.
“although the 'certain' road segments score slightly better at ~ 27%”
Quote in the paperNot a substantial claim - Findingown data
Combination of slope categories and topographic variables better explains locations of Roman roads than slope only.
“The fact that these variables had an influence on layouts of Roman roads is corroborated by comparison of isotropic models with the slope-based functions which showed that coincidence of Roman roads with isotropic models is higher, and combination of slope categories and topographic variables better explain locations of Roman roads than slope only.”
Quote in the paperSubstantial claim- The authors state this as an assumption: Roman roads are valid proxies reflecting past human movement and its limitations.
- Findingown data
FABDEM and EDTM are comparable both with 99 percentile slope value around 22°.
“FABDEM and EDTM are comparable both with 99 percentile value around 22°.”
Quote in the paperSubstantial claim
- Findingown data
Coincidence of Roman roads with isotropic models is higher than with slope-based functions.
“The fact that these variables had an influence on layouts of Roman roads is corroborated by comparison of isotropic models with the slope-based functions which showed that coincidence of Roman roads with isotropic models is higher, and combination of slope categories and topographic variables better explain locations of Roman roads than slope only.”
Quote in the paperSubstantial claim- The authors state this as an assumption: Roman roads are valid proxies reflecting past human movement and its limitations.
- Findingreasoning
The overlap between modelled corridors and the Roman roads in the isotropic models A and B is only ~25%.
“Nevertheless, the overlap between modelled corridors and the Roman roads in the isotropic models A and B is only ~ 25%.”
Quote in the paperSubstantial claim - Findingown data
A corridor runs around the southern tip of the Dead Sea.
“Finally, a corridor runs around the southern tip of the Dead Sea.”
Quote in the paperNot a substantial claim - Findingown data
A couple of highland roads (e.g., Bethoron road, 'watershed road' between Gophna and Hebron) are captured by the model.
“and couple of highland roads (e.g., Bethoron road, 'watershed road' between Gophna and Hebron, etc. Figure 9B).”
Quote in the paperNot a substantial claim - Findingown data
Approximately 98.4% of all road length is located below 20° slope.
“Approximately 98.4% of all road length is located below 20° slope but note the long tail with extreme values up to 39.9°.”
Quote in the paperSubstantial claim - Findingown data
Parallel north-south corridors in the piedmont of the Judaean and Samaritan highlands in the coastal plain and on the Transjordanian plateau, east of the valleys, are also found in common with the isotropic models.
“Parallel north-south corridors in the piedmont of the Judaean and Samaritan highlands in the coastal plain and on the Transjordanian plateau, east of the valleys, are also found in common with the isotropic models.”
Quote in the paperNot a substantial claim - Findingown data
The model of natural corridors built on analysed topographic limits showed that Roman roads do not overlap with the corridors.
“The model of the natural corridors of movement built on the analysed topographic limits of the Roman roads showed that the Roman roads do not overlap with the corridors (i.e., with places in the landscape that are more conductive to channel movement).”
Quote in the paperSubstantial claim- The authors state this as an assumption: Roman roads are valid proxies reflecting past human movement and its limitations.
- Findingown data
There is a long tail of slope values with extreme values up to 39.9°.
“Approximately 98.4% of all road length is located below 20° slope but note the long tail with extreme values up to 39.9°.”
Quote in the paperSubstantial claim - Findingown data
Only ca.0.45% of road lengths fall outside the range of TPI min -80.71 to TPI max 91.17.
“When taking the minimum value of TPI max (-80.71) and maximum value of TPI min (91.17) as a range of values where Roman roads are typically located, we notice that only ca.0.45% of road lengths fall outside this range.”
Quote in the paperSubstantial claim - Findingnone given
Among the places located on the major corridors are several key Roman sites, such as Hippos, Pella, Philadelphia-Amman, Madaba, Characmoba-Kerak, Petra, Antipatris, Lydda, Jerusalem, Mampsis and Oboda, with many more minor settlements and forts.
“Among the places located on the major corridors are several key Roman sites, such as Hippos on the crossroad of corridors through the Jordan Valley and the Golan Heights, Pella, Philadelphia-Amman, Madaba, Characmoba-Kerak, Petra in Transjordan and Antipatris, Lydda, Jerusalem in Judaea, and Mampsis and Oboda on the Incense route in the Negev, with many more minor settlements and forts.”
Quote in the paperNot a substantial claim - Findingown data
Calculation of VRML was heavily influenced by an edge effect, where all road segments located within 1 raster cell from the sea show extremely high values.
“Calculation of VRML was heavily influenced by an edge effect, where all road segments located within 1 raster cell from the sea (represented by null values) show extremely high values.”
Quote in the paperSubstantial claim - Findingown data
Fewer corridors are found in the hilly areas across the case study region for the slope-based algorithms.
“However, fewer corridors are found in the hilly areas across the case study region.”
Quote in the paperNot a substantial claim - Findingown data
One of the LCPs produced by a model more sensitive to slope categories, critical slope, and topography very closely follows the Bethoron road between Jerusalem and Lydda.
“However, one of the LCPs produced by a model more sensitive to slope categories, critical slope, and topography in fact very closely follows so-called Bethoron road between Jerusalem and Lydda (Fig. 3D) that was not included in the initial test.”
Quote in the paperSubstantial claim - Findingown data
There are disconnections between some Roman sites and modelled corridors.
“There are also disconnections between some Roman sites and modelled corridors.”
Quote in the paperSubstantial claim - Findingnone given
A major corridor runs through Ginae-Jenin and Narbata.
“Instead, major corridor runs through Ginae-Jenin and Narbata.”
Quote in the paperNot a substantial claim - Findingown data
There is no major corridor in Meggido-Legio area.
“For example, there is no major corridor in Meggido-Legio area, which is historically attested as major crossroad of international routes (Isaac & Roll, 1982).”
Quote in the paperNot a substantial claim - Findingnone given
Model B does not significantly differ from model A in where the major corridors are located.
“The model B does not significantly differ from model A in where the major corridors are located.”
Quote in the paperNot a substantial claim - Findingreasoning
Total 93.68% of the road length (3,451.3 km, not including segments affected by the VRML edge effect) fits all three criteria (slope < 20°, TPI -80.71-+91.17, VRML < 0.002332).
“We may note that total 93.68% of the road length (3,451.3km, not including segments affected by the VRML edge effect) fits all three criteria (slope < 20°, TPI -80.71-+ 91.17, VRML < 0.002332).”
Quote in the paperSubstantial claim - Findingreasoning
Edge effect influenced mainly north-eastern and southern extremes of the region resulting in poor overlap with Roman roads.
“Edge effect influenced mainly north-eastern and southern extremes of the region resulting in poor overlap with Roman roads.”
Quote in the paperSubstantial claim - Findingown data
The 99 percentile corridors are divided into several disconnected segments, mostly in the central part of the region.
“The 99 percentile corridors are divided into several disconnected segments, mostly in the central part of the region.”
Quote in the paperNot a substantial claim - Findingown data
In model A, a long continual section follows the piedmont of the Judaean and Samaritan Highlands from the northern Negev through the coastal plain, passing through Lower Galilee to the Golan Heights.
“In model A, a long continual section follows the piedmont of the Judaean and Samaritan Highlands from the northern Negev through the coastal plain, passing through Lower Galilee to the Golan Heights.”
Quote in the paperNot a substantial claim - Findingown data
Many Transjordanian ascents (Aqabat Fiq, Gadara, Mahatta, Kathrabba pass, Petra -Naqb el-Aqab) and Negev ascents (Maaleh Tamar, Maaleh Aqravim) are fully or partially captured by the model.
“many Transjordanian ascents (Aqabat Fiq, Gadara, Mahatta, Kathrabba pass, Petra -Naqb el-Aqab) and Negev ascents (Maaleh Tamar, Maaleh Aqravim) are fully or partially captured (Fig. 9A-C)”
Quote in the paperSubstantial claim - Findingown data
Critical slope and topographic variables are insufficient in modelling the roads in areas with low topographic variability.
“From the start, it became clear that critical slope and topographic variables are insufficient in modelling the roads in the areas with low topographic variability (Fig. 3A).”
Quote in the paperSubstantial claim - Findingown data
From the initial results, it seemed that the best model is the one less sensitive to slope categories and topography (using only two slope categories and higher conductivity values for topography and critical slope).
“From the initial results (Table 3), it seemed that the best model is the one less sensitive to slope categories and topography (using only two slope categories and higher conductivity values for topography and critical slope).”
Quote in the paperNot a substantial claim - Findingnone given
The model generally well captures potential movement corridors in hilly and mountainous zones.
“The model generally well captures potential movement corridors in hilly and mountainous zones, and so e.g., many Transjordanian ascents”
Quote in the paperSubstantial claim - Findingreasoning
In the case of the north-eastern corner, the contributing factor of the poor overlap is the flat region with little topographic variability.
“Although in the case of the north-eastern corner the contributing factor of the poor overlap is the flat region with little topographic variability.”
Quote in the paperNot a substantial claim - Findingnone given
A major corridor crosses bridge at Jisr el-Majami between Scythopolis and Gadara.
“A major corridor crosses bridge at Jisr el-Majami between Scythopolis and Gadara (Fig. 9A).”
Quote in the paperNot a substantial claim - Findingown data
After removing segments affected by the VRML edge effect (76 segments, 75.3 km), ca.95% of roads falls below 0.002332.
“After these segments were removed from consideration (76 segments, 75.3 km) ca.95% of roads falls below 0.002332.”
Quote in the paperSubstantial claim - Findingown data
Topographic variables had an influence on layouts of Roman roads.
“The fact that these variables had an influence on layouts of Roman roads is corroborated by comparison of isotropic models with the slope-based functions which showed that coincidence of Roman roads with isotropic models is higher, and combination of slope categories and topographic variables better explain locations of Roman roads than slope only.”
Quote in the paperSubstantial claim- The authors state this as an assumption: Roman roads are valid proxies reflecting past human movement and its limitations.
- Findingown data
The selected topographic variables and their limits clearly did not provide enough friction over large regions with little topographic variability, e.g., the coastal plain or the Transjordanian plateau.
“In the first place, the selected topographic variables and their limits clearly did not provide enough friction over large regions with little topographic variability, e.g., the coastal plain or the Transjordanian plateau.”
Quote in the paperSubstantial claim - Findingown data
After including the Bethoron road, the ranking of models changed.
“After including this road, the ranking of models changed.”
Quote in the paperSubstantial claim - Findingnone given
The model coincides well mainly with Roman roads going N-S along the coastal plain and along the length of the Rift Valley.
“The model coincides well mainly with Roman roads going N-S along the coastal plain and along the length of the Rift Valley (Fig. 9B-C).”
Quote in the paperNot a substantial claim - Findingown data
The 95 and 90 percentile corridors add connections in the Judaean and Samaritan Highlands, Negev, Transjordan, and upper Jordan Valley.
“The 95 and 90 percentile corridors add connections in the Judaean and Samaritan Highlands, Negev, Transjordan, and upper Jordan Valley.”
Quote in the paperNot a substantial claim - Findingown data
Most computed corridors run roughly north-south, the orientation of the research region, with few diagonals or east-west corridors.
“Most computed corridors run roughly north-south, the orientation of the research region, with few diagonals or east-west corridors.”
Quote in the paperNot a substantial claim - Findingreasoning
Some modelled corridors follow the general direction of a road and run parallel to it but do not overlap within the given neighbourhood.
“On the other hand, some modelled corridors follow the general direction of a road and run parallel to it but do not overlap within the given neighbourhood.”
Quote in the paperNot a substantial claim - Findingown data
There is some degree of optimization in Roman roads.
“Alternatively, the assumption that the Roman roads were globally optimised to minimise cost could be wrong, but as the results show there is some degree of optimization and the variables not considered in this research might potentially explain the rest.”
Quote in the paperSubstantial claim- The authors state this as an assumption: The Roman roads were globally optimised to minimise cost.
- Premisecitation
High spatial accuracy of the Roman road dataset was achieved through a combination of modern and historical topographic maps, high-resolution satellite imagery, and modern orthophotos.
“High spatial accuracy was achieved through a combination of modern and historical topographic maps, high-resolution satellite imagery, and modern orthophotos (De Soto et al., 2025).”
Quote in the paperSubstantial claim0 of 1 sources checked- cites De Soto et al., 2025 source
Checked before — its earlier verdict is held back until a new check runs.
- Interpretationreasoning
The results suggest that additional anthropic, cultural variables have a stronger influence on the layout of the Roman road network.
“It is argued that these results suggest that additional anthropic, cultural variables have a stronger influence on the layout of the Roman road network”
Quote in the paperSubstantial claim - Interpretationown data
The selected topographic variables and slope categories have limited explanatory power in understanding the layout of the Roman road network as a whole.
“suggesting that the selected topographic variables and slope categories have limited explanatory power in understanding the layout of the Roman road network as a whole”
Quote in the paperSubstantial claim - Premisecitation
The southern Levant is more represented in the Roman road dataset due to certain research bias.
“although the southern Levant is more represented due to certain research bias (Pažout, 2025).”
Quote in the paperSubstantial claim0 of 1 sources checked- cites Pažout, 2025 source
- Premisenone given
Only Van Lanen et al. (2015) in the case of the Roman Netherlands incorporates a wide range of environmental and anthropic variables to analyse potential for presence of roads in the landscape.
“Only Van Lanen et al. (2015) in the case of the Roman Netherlands incorporates wide range of environmental and anthropic variables to analyse potential for presence of roads in the landscape.”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites Van Lanen et al. (2015) source
- Premisecitation
The Roman road dataset for the Levant is derived from Itiner-e: The digital atlas of ancient roads.
“The Roman road dataset for the Levant is derived from the Itiner-e: The digital atlas of ancient roads (Brughmans et al., 2024).”
Quote in the paperSubstantial claim0 of 1 sources checked- cites Brughmans et al., 2024
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
- Interpretationown data
The calculated natural corridors of movement clearly illustrate several issues with the proposed simple isotropic model.
“The calculated natural corridors of movement clearly illustrate several issues with the proposed simple isotropic model.”
Quote in the paperSubstantial claim - Premisecitation
Part of this region lies in (semi-)arid zone so availability of water sources would be more influential variable than topography.
“Part of this region lies in (semi-) arid zone so availability of water sources would be more influential variable than topography (Manière et al., 2021).”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites Manière et al., 2021 source
A check is scheduled.
- Interpretationreasoning
Topographic variables explain an additional 7-10% of locations after slope.
“If slope algorithms explain ca.14.5-17% of the locations and Models A and B ca. 25%, we may surmise that topographic variables explain additional 7-10% of locations after slope.”
Quote in the paperSubstantial claim- The authors state this as an assumption: Roman roads are valid proxies reflecting past human movement and its limitations.
- Interpretationreasoning
Regions with gentle topography would necessitate creation of a different model based on additional variables.
“Such regions with gentle topography would necessitate creation of a different model based on additional variables.”
Quote in the paperSubstantial claim - Interpretationreasoning
The low overlap is likely caused by the scale of the analysis not permitting capturing enough potential natural corridors of movement.
“The low overlap is likely caused by a combination of two factors: (1) models with selected topographic variables do not perform well in flat regions, as was discussed above; and (2) the scale of the analysis does not permit capturing enough potential natural corridors of movement.”
Quote in the paperSubstantial claim - Interpretationreasoning
The main drivers behind the decision process leading to the creation of road layout must be anthropic, cultural variables, such as settlement patterns, exchange networks, persistency, state interventions, land use and land ownership.
“The main drivers behind the decision process leading to the creation of road layout therefore must be anthropic, cultural variables, such as settlement patterns, long-and short-distance exchange networks, persistency of the road network, state interventions (for strategic or military reasons), land use and land ownership, etc.”
Quote in the paperSubstantial claim- The authors state this as an assumption: Roman roads are valid proxies reflecting past human movement and its limitations.
- Interpretationreasoning
The size of the neighbourhood used to aggregate LCPs to corridors (1000 m) might partly cause some modelled corridors to run parallel but not overlap.
“This might be partly caused by the size of the neighbourhood used to aggregate the calculated LCPs to corridors (1000 m), although not all corridors are limited to 1000 m width, as multiple LCPs can converge close to each other in an area of several kms creating a wide corridor.”
Quote in the paperNot a substantial claim - Interpretationreasoning
The layout of the Roman roads and of the Roman road network must be explained by other factors not considered in this research.
“The layout of the Roman roads and of the Roman road network therefore must be explained by other factors not considered in this research.”
Quote in the paperSubstantial claim- The authors state this as an assumption: Roman roads are valid proxies reflecting past human movement and its limitations.
- Interpretationreasoning
The channelling of LCPs into few corridors is the main reason for low overlap with Roman roads.
“This 'channelling' of LCPs into few corridors is the main reason for low overlap with Roman roads.”
Quote in the paperSubstantial claim - Premisecitation
The three roads Jerusalem-Lydda, Jerusalem-Emmaus, and Emmaus-Lydda represent varied topography from flat coastal plain interspersed with shallow wide wadis, low rolling hills, to steep ridges and deeply incised valleys.
“These three roads represent varied topography from flat coastal plain interspersed with shallow wide wadis, low rolling hills, to steep ridges and deeply incised valleys.”
Quote in the paperSubstantial claim0 of 1 sources checked- cites Fischer et al., 1996 source
- Interpretationreasoning
The selection of 20° slope as an upper limit for road location (98.5 percentile for FABDEM, 98 percentile for EDTM) seems to accord well with data from these two improved global DEMs.
“The selection of 20° slope as un upper limit for road location (98.5 percentile for FABDEM, 98 percentile for EDTM) seems to accord well with data from these two improved global DEMs.”
Quote in the paperSubstantial claim - Interpretationnone given
Despite these limitations, it is possible to observe some interesting coincidences between modelled corridors, Roman roads, and location of important nodes (Roman sites).
“Despite these limitations, it is possible to observe some interesting coincidences between modelled corridors, Roman roads, and location of important nodes (Roman sites).”
Quote in the paperSubstantial claim - Interpretationreasoning
The low overlap is likely caused by models with selected topographic variables not performing well in flat regions.
“The low overlap is likely caused by a combination of two factors: (1) models with selected topographic variables do not perform well in flat regions, as was discussed above; and (2) the scale of the analysis does not permit capturing enough potential natural corridors of movement.”
Quote in the paperSubstantial claim - Premisecitation
Meggido-Legio area is historically attested as major crossroad of international routes.
“For example, there is no major corridor in Meggido-Legio area, which is historically attested as major crossroad of international routes (Isaac & Roll, 1982).”
Quote in the paperSubstantial claim0 of 1 sources checked- cites Isaac & Roll, 1982
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
- Premisenone given
FABDEM provides the best resolution covering the whole research area at 1-arc second ca.34 m at this latitude.
“It provides the best resolution covering the whole research area at 1-arc second ca.34 m at this latitude.”
Quote in the paperNot a substantial claim - Premisecitation
The 1000-m neighbourhood was selected because it was previously demonstrated that at this scale the effect of movement corridors on the location of Roman sites is most pronounced.
“A 1000-m neighbourhood was selected as it was previously demonstrated that at this scale the effect of movement corridors on the location of Roman sites is most pronounced (Lewis, 2024a).”
Quote in the paperSubstantial claim0 of 1 sources checked- cites Lewis, 2024a
- Interpretationreasoning
The influence of selected topographic variables on the layout of Roman roads, while it existed, is limited.
“and their influence on the layout of Roman roads, while it existed, is limited.”
Quote in the paperSubstantial claim- The authors state this as an assumption: Roman roads are valid proxies reflecting past human movement and its limitations.
- Interpretationreasoning
A model combining topographic and anthropic variables would be better suited to explain the layout of Roman roads.
“In light of this conclusion, it seems that a model combining topographic and anthropic variables would be better suited to explain the layout of Roman roads, and for this least-cost path networks might offer suitable methodology (Carroll & Carroll, 2023;Groenhuijzen & Verhagen, 2017;Herzog, 2023;Prignano et al., 2019).”
Quote in the paperNot a substantial claim0 of 3 sources checked- cites Carroll & Carroll, 2023 source
- cites Groenhuijzen & Verhagen, 2017 source
- cites Herzog, 2023 source
- The authors state this as an assumption: Roman roads are valid proxies reflecting past human movement and its limitations.
- Interpretationreasoning
The selected topographic variables (critical slope, range of TPI values, critical value of VRML) do not explain the location of the Roman roads well.
“the inevitable conclusion is that the selected topographic variables (critical slope, range of TPI values, critical value of VRML) do not explain the location of the Roman roads well”
Quote in the paperSubstantial claim- The authors state this as an assumption: Roman roads are valid proxies reflecting past human movement and its limitations.
- Probably not a substantial claim. 7
- It may state two separate claims as one. 27
Probably not a substantial claim. The paper’s conclusions may not depend on it.
Findingown dataIn total, 3877.8 km of roads in the dataset located in the Roman Levant belong to the 'certain' category.
“In total, 3877.8 km of roads in the dataset located in the Roman Levant (Fig. 2) belong to the 'certain' category”
Quote in the paperNot a substantial claimProbably not a substantial claim. The paper’s conclusions may not depend on it.
FindingreasoningWith 118.4 km of roads removed, the analysis is performed for a total of 3,759.4 km of roads.
“With 118.4 km such removed, the analysis is performed for a total of 3,759.4km of roads found throughout the research area in various environments”
Quote in the paperNot a substantial claimIt may state two separate claims as one.
FindingreasoningThe shape of the computational region, which roughly extends N-S, makes LCPs and corridors to have the same orientation with few E-W and diagonal connections, typically in places where the LCPs are forced through mountain passes or ridgeways between the coast and the hinterland.
“Another issue is that the shape of the computational region, which roughly extends N-S, makes LCPs and corridors to have the same orientation with few E-W and diagonal connections, typically in places where the LCPs are forced through mountain passes or ridgeways between the coast and the hinterland.”
Quote in the paperNot a substantial claimIt may state two separate claims as one.
Findingown dataThe 95 and 90 percentile corridors are more connected, but there are also many parallel routes or wide corridors in plains and valleys.
“They are more connected, but we see also many parallel routes or wide corridors in plains and valleys.”
Quote in the paperNot a substantial claimIt may state two separate claims as one.
Findingown dataThe results obtained from all four slope-based algorithms are very similar, with a major continuous corridor running along the length of the Jordan Valley.
“The results obtained from all four slope-based algorithms are very similar, with a major continuous corridor running along the length of the Jordan Valley.”
Quote in the paperSubstantial claimIt may state two separate claims as one.
Findingown dataSmall changes between model A and B can be observed in flat areas with wider corridors and in hilly areas, where corridors are not channelled to one route, but some parallel corridors exist.
“Small changes can be observed in flat areas with wider corridors and in the hilly areas, where corridors are not channelled to one route, but some parallel corridors exist.”
Quote in the paperNot a substantial claimIt may state two separate claims as one.
FindingreasoningSlope algorithms explain ca. 14.5-17% of the locations, and Models A and B ca. 25%.
“If slope algorithms explain ca.14.5-17% of the locations and Models A and B ca. 25%, we may surmise that topographic variables explain additional 7-10% of locations after slope.”
Quote in the paperSubstantial claim- The authors state this as an assumption: Roman roads are valid proxies reflecting past human movement and its limitations.
It may state two separate claims as one.
Findingnone givenThis effect can be observed also in areas of gentle topography such as piedmonts and low hills where LCPs converge either in discrete places (crossroads) or along short segments which causes breaking of the corridors into non-continuous segments.
“This effect can be observed also in areas of gentle topography such as piedmonts and low hills where LCPs converge either in discrete places (crossroads) or along short segments which causes breaking of the corridors into non-continuous segments.”
Quote in the paperNot a substantial claimIt may state two separate claims as one.
Findingown dataAnother corridor follows the central part of the Jordan Valley, meeting the previous corridor in the northern Negev.
“Another follows the central part of the Jordan Valley, meeting the previous corridor in the northern Negev.”
Quote in the paperNot a substantial claimIt may state two separate claims as one.
Findingown dataTwo combinations of conductivity values performed best in modelling the roads in the test case: one more sensitive to critical slope and topography (values 10 and 2 respectively), and one less sensitive (values 20 and 10).
“Two combinations of values performed best in modelling the roads in the test case (Fig. 3D): one more sensitive to critical slope and topography (values 10 and 2 respectively), and one less sensitive (values 20 and 10).”
Quote in the paperSubstantial claimIt may state two separate claims as one.
Findingown dataAnother corridor branches in the central Jordan Valley north of the Dead Sea, ascends the Transjordanian plateau and runs south, bypassing the deep wadis.
“Another corridor branches in the central Jordan Valley north of the Dead Sea.It ascends the Transjordanian plateau and runs south, bypassing the deep wadis.”
Quote in the paperNot a substantial claimIt may state two separate claims as one.
Findingown dataSlope and VRML values are both positively skewed.
“Slope and VRML values are both positively skewed.”
Quote in the paperSubstantial claimIt may state two separate claims as one.
Findingown dataVia Nova Traiana is well captured between Khirbet es-Samra to Philadelphi and then south to Wadi Mujib where crossing of this wadi is captured very accurately.
“Via Nova Traiana is well captured between Khirbet es-Samra to Philadelphi and then south to Wadi Mujib where crossing of this wadi is captured very accurately (Fig. 9C).”
Quote in the paperNot a substantial claimIt may state two separate claims as one.
Findingown dataThe overlap with Roman roads is weaker for all four slope-based algorithms, with maximum overlap in 90 percentile frequency category between ~14.7 and 17%, so 7-10% less than in the isotropic models.
“The overlap with Roman roads is weaker for all four algorithms, with maximum overlap in 90 percentile frequency category between ~ 14.7 and 17%, so 7-10% less than in the isotropic models.”
Quote in the paperSubstantial claimIt may state two separate claims as one.
Findingnone givenPart of the overlap between modelled corridors and Roman roads is caused by perpendicular crossings, which are accidental and a poor match.
“Moreover, part of the overlap is caused by perpendicular crossings, which are accidental and a poor match.”
Quote in the paperSubstantial claimIt may state two separate claims as one.
Findingnone givenThe movement corridors modelled using slope-based algorithms tend to follow valleys and flat areas more systematically, resulting in fewer corridors in hilly and mountainous areas.
“The movement corridors modelled using slope-based algorithms tend to follow valleys and flat areas more systematically, which results in fewer corridors in the hilly and mountainous areas.”
Quote in the paperNot a substantial claimIt may state two separate claims as one.
Findingown dataBoth TPI minimum and maximum values are close to the normal distribution, with median values around 0.
“Both TPI minimum and maximum values are close to the normal distribution, with median values around 0.”
Quote in the paperNot a substantial claimIt may state two separate claims as one.
Findingnone givenPart of the overlap in isotropic models is caused by accidental perpendicular overlap, though it seems less prevalent than with slope-based functions.
“It also needs to be admitted here that part of this overlap is caused by the accidental perpendicular overlap, although it seems to be less prevalent than with the slope-based functions.”
Quote in the paperNot a substantial claimIt may state two separate claims as one.
Findingown dataCopernicus GLO-90 is smoothed and tends to have lower slope values overall, while SRTM and especially ASTER are skewed to higher values.
“We may see that Copernicus is indeed smoothed and tends to have lower values overall, while SRTM and especially ASTER are skewed to higher values.”
Quote in the paperNot a substantial claimIt may state two separate claims as one.
PremisecitationLCP algorithms predominantly used nowadays are slope-based, assuming that movement through terrain is either a time-saving or energy-saving process.
“LCP algorithms predominantly used nowadays are slopebased assuming that movement through terrain is either time-saving or energysaving process and methods were developed to assess which might have played more role in decision making when establishing a path (Field et al., 2023).”
Quote in the paperSubstantial claim0 of 1 sources checked- cites Field et al., 2023 source
Probably not a substantial claim. It may state two separate claims as one.
PremisecitationThe DEM used for the analysis is FABDEM, which is based on the Copernicus GLO-30 DEM.
“The Digital Elevation Model (DEM) used for the analysis is FABDEM (Forest and Building removed DEM, Neal & Hawker, 2023), which is based on the Copernicus GLO-30 DEM (European Space Agency and Airbus 2022).”
Quote in the paperSubstantial claim0 of 1 sources checked- cites Neal & Hawker, 2023
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
It may state two separate claims as one.
InterpretationcitationThe identified topographic limits of Roman road locations offer a wide niche in the landscape where Roman roads can be located, and the road layout is a result of additional processes related to settlement structures, economy, persistency, and local circumstances.
“where Roman roads can be located, and the road layout (in the sense of an actual realised route in the landscape, Watteaux, 2017) is a result of additional processes related to the development of settlement structures, economy, persistency (building new roads is more expensive than maintaining existing roads), and local circumstances (Earle, 2011;Llobera, 2020;Pažout et al., 2024).”
Quote in the paperSubstantial claim0 of 3 sources checked- cites Watteaux, 2017 source
Checked before — its earlier verdict is held back until a new check runs.
- cites Earle, 2011
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
- cites Llobera, 2020 source
- The authors state this as an assumption: Roman roads are valid proxies reflecting past human movement and its limitations.
It may state two separate claims as one.
PremisecitationBy removing forest and building cover, FABDEM achieves higher vertical accuracy compared to similar global DEMs and better represents the underlying topography.
“Moreover, by removing the forest and building cover, it achieves higher vertical accuracy compared to similar global DEMs (Hawker et al., 2022) and better represents the underlying topography.”
Quote in the paperSubstantial claim0 of 1 sources checked- cites Hawker et al., 2022 source
Checked before — its earlier verdict is held back until a new check runs.
Probably not a substantial claim. The paper’s conclusions may not depend on it. The paper may give no reason why it holds. It may state two separate claims as one.
Study design and conductreasoningThe spatial accuracy of certain roads in the dataset might be too low in certain places, though given the width of the corridors, this should not be an issue.
“Additionally, the spatial accuracy of certain roads in the dataset might be too low in certain places, although given the width of the corridors, this should not be an issue.”
Quote in the paperNot a substantial claimIt may state two separate claims as one.
PremisecitationGerasa is disconnected from all corridors in Transjordan despite being a crossroad of several main roads in the Roman period.
“Likewise, Gerasa is disconnected from all corridors in Transjordan despite being a crossroad of several main roads in the Roman period (Mittmann, 1970).”
Quote in the paperSubstantial claim0 of 1 sources checked- cites Mittmann, 1970
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
Probably not a substantial claim. The paper’s conclusions may not depend on it. The paper may give no reason why it holds. It may state two separate claims as one.
Premisenone givenBoth FABDEM and EDTM are supposed to have improved vertical accuracy and mean root square error compared to other global DEMs which often include unfiltered buildings and vegetation cover.
“Both FABDEM and EDTM are supposed to have improved vertical accuracy and mean root square error compared to other global DEMs which often include unfiltered buildings and vegetation cover.”
Quote in the paperNot a substantial claimProbably not a substantial claim. It may be too vague to check against evidence.
InterpretationreasoningLeast-cost path networks might offer suitable methodology for such a combined model.
“In light of this conclusion, it seems that a model combining topographic and anthropic variables would be better suited to explain the layout of Roman roads, and for this least-cost path networks might offer suitable methodology (Carroll & Carroll, 2023;Groenhuijzen & Verhagen, 2017;Herzog, 2023;Prignano et al., 2019).”
Quote in the paperNot a substantial claim0 of 3 sources checkedIt may state two separate claims as one.
Interpretationnone givenPossible water availability might provide a better explanation in the (semi-)arid regions of the Roman Near East, but it will probably not be a decisive variable in the Mediterranean climate zone with better availability of water.
“Besides topographic variables, possible water availability might provide better explanation in the (semi-)arid regions of the Roman Near East, but it will probably not be decisive variable in the Mediterranean climate zone with better availability of water.”
Quote in the paperSubstantial claimProbably not a substantial claim. It may be too vague to check against evidence. It may state two separate claims as one.
InterpretationreasoningMany of these anthropic variables are also related to topographic and environmental variables, but the way how people manipulated their environment might be highly specific.
“Arguably, many of these variables are also related to topographic and environmental variables, but the way how people manipulated their environment might be highly specific.”
Quote in the paperNot a substantial claimIt may state two separate claims as one.
PremisecitationThe resolution and quality of the DEM can influence both the calculation of the topographic variables and the length and shape of the calculated LCPs.
“The resolution and quality of the DEM can influence both the calculation of the topographic variables and the length and shape of the calculated LCPs (Herzog & Posluschny, 2011).”
Quote in the paperSubstantial claim0 of 1 sources checked- cites Herzog & Posluschny, 2011
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
The paper relies on this without support.
There is often limited use of additional environmental, topographic, or anthropic variables in LCP studies.
“Preoccupation with slope means that there is often limited use of additional environmental, topographic, or anthropic variables (typically land cover or hydrology).”
The paper relies on this without support.
Selection of additional variables is very often a deductive, top-down process that incorporates existing archaeological data only in its last (verification) stage.
“And if so, selection of those variables is very often deductive, top-down process that incorporates existing archaeological data only in its last (verification) stage.”
The paper relies on this without support.
A bottom-up inductive process of analysis of available data for past movement that is essentially road-oriented is missing in both predicting and post-dicting approaches.
“What seems to be missing in both predicting and post-dicting approaches adopted in the aforementioned studies is the lack of bottom-up inductive process of analysis of available data for past movement that would be essentially road-oriented.”
The paper relies on this without support.
Roman roads are one of the best documented remains of pre-modern transport infrastructure reflecting past human movement.
“Arguably, this was likely caused by lack of appropriate dataset of roads, although Roman roads are one the best documented remains of pre-modern transport infrastructure reflecting past human movement.”
The paper relies on this without support.
The southern Levant Roman road network is more representative to allow comparison with modelled natural corridors of movement.
“For the same reason, the southern Levant was chosen as the case study since the Roman road network in the region is more representative to allow comparison with modelled 'natural corridors of movement'.”
The authors state this as an assumption.
The final conductance surface is isotropic, assuming equal passability in all directions including the reclassified slope values.
“The final conductance surface is isotropic, i.e., direction-independent, assuming equal passability in all directions including the reclassified slope values.”
The paper relies on this without support.
Roman roads can serve as a proxy for past movement patterns.
“By starting the enquiry with Roman roads as a proxy for the past movement patterns, it is hoped to avoid some pitfalls of the 'traditional' LCP algorithms described above and to gain empirically informed evaluation of the influence of topography on actual”
The paper relies on this without support.
Caesarea was chosen as the main provincial harbour and capital rather for its maritime connectivity than for terrestrial mobility.
“Caesarea was chosen as the main provincial harbour and capital rather for its maritime connectivity than for terrestrial mobility.”
The authors state this as an assumption.
Roman roads are valid proxies reflecting past human movement and its limitations.
“assuming that Roman roads are valid proxies reflecting past human movement and its limitations.”
The authors state this as an assumption.
The Roman roads were globally optimised to minimise cost.
“the assumption that the Roman roads were globally optimised to minimise cost could be wrong”
Citations
Cites
Terrain analysis : principles and applications
Wilson, J. P., & Gallant, J. (2000) · Wiley eBooks · Cited by 1,145
A 30 m global map of elevation with forests and buildings removed
Hawker, L., Uhe, P., Paulo, L., Sosa, J., Savage, J., Sampson, C., & Neal, J. (2022) · Environmental Research Letters · doi:10.1088/1748-9326/ac4d4f · Cited by 668
European Space Agency, & Airbus (2022) · doi:10.5270/esa-c5d3d65 · Cited by 418
ASTER Global Digital Elevation Model V003
NASA/METI/AIST/Japan Spacesystems and U.S./Japan ASTER Science Team (2019) · Earth Observing System Data and Information System · doi:10.5067/aster/astgtm.003 · Cited by 318
Shuttle Radar Topography Mission (SRTM) 1 Arc-Second Global
Earth Resources Observation and Science (EROS) Center (2000) · USGS DOI Tool Production Environment · doi:10.5066/f7pr7tft · Cited by 282
White, D. A., & Barber, S. B. (2012) · Journal of Archaeological Science · doi:10.1016/j.jas.2012.04.017 · Cited by 146
Computational approaches to archaeological spaces
(2014) · Choice Reviews Online · doi:10.5860/choice.51-3920 · Cited by 144
Modelling of pathways and movement networks in archaeology: an overview of current approaches
Verhagen, P., Nuninger, L., & Groenhuijzen, M. R. (2019) · Computational Social Sciences · doi:10.1007/978-3-030-04576-0_11 · Cited by 117
Zigzagging: Theoretical insights on climbing strategies
Llobera, M., & Sluckin, T. J. (2007) · Journal of Theoretical Biology · doi:10.1016/j.jtbi.2007.07.020 · Cited by 103
Using Global Positioning System Data to Understand Variations in Path Choice
Jan, O., Horowitz, A. J., & Peng, Z. (2000) · Transportation Research Record Journal of the Transportation Research Board · doi:10.3141/1725-06 · Cited by 103
Güimil-Fariña, A., & Parcero-Oubiña, C. (2015) · Journal of Archaeological Science · doi:10.1016/j.jas.2014.11.030 · Cited by 95
Roman Roads in Judaea II: The Jaffa-Jerusalem roads
Fischer, M., Isaac, B., & Roll, I. (1996) · doi:10.30861/9780860548096 · Cited by 63
The potential and limits of optimal path analysis
(2016) · Computational Approaches to Archaeological Spaces · doi:10.4324/9781315431932-14 · Cited by 57
Lewis, J. (2021) · Journal of Archaeological Method and Theory · doi:10.1007/s10816-021-09522-w · Cited by 53
Neal, J., & Hawker, L. (2023) · Bristol Research (University of Bristol) · doi:10.5523/bris.s5hqmjcdj8yo2ibzi9b4ew3sn · Cited by 51
Spatial analysis based on cost functions
Herzog, I. (2020) · Archaeological Spatial Analysis · doi:10.4324/9781351243858-18 · Cited by 45
Fonte, J., Parcero‐Oubiña, C., & García, J. M. C. (2017) · Mediterranean Archaeology & Archaeometry. International Journal/Mediterranean archaeology and archaeometry. International scientific journal · doi:10.5281/zenodo.1005561 · Cited by 43
Improved topographic ruggedness indices more accurately model fine-scale ecological patterns
Dilts, T. E., Blum, M. E., Shoemaker, K. T., Weisberg, P. J., & Stewart, K. M. (2023) · Landscape Ecology · doi:10.1007/s10980-023-01646-6 · Cited by 36
GEDTM30: global ensemble digital terrain model at 30 m and derived multiscale terrain variables
Ho, Y.-F., Grohmann, C. H., Lindsay, J. B., Reuter, H. I., Parente, L., Witjes, M., & Tomislav Hengl, A. (2025) · PeerJ · doi:10.7717/peerj.19673 · Cited by 33
van Lanen, R. J., Kosian, M. C., Groenewoudt, B. J., & Jansma, E. (2015) · Geoarchaeology · doi:10.1002/gea.21510 · Cited by 29
Tilt – Slope-Dependent Least Cost Path Calculations Revisited
Unknown, (. (2011) · Universitätsbibliothek Tübingen · doi:10.15496/publikation-2095 · Cited by 29
Manière, L., Crépy, M., & Redon, B. (2021) · Journal of Computer Applications in Archaeology · doi:10.5334/jcaa.67 · Cited by 28
Prignano, L., Morer, I., Fulminante, F., & Lozano, S. (2019) · Journal of Archaeological Science · doi:10.1016/j.jas.2019.02.007 · Cited by 27
Fábrega‐Álvarez, P. (2006) · DIGITAL.CSIC (Spanish National Research Council (CSIC)) · Cited by 23
CAA2015. Keep The Revolution Going
(2016) · Archaeopress Publishing Ltd eBooks · doi:10.2307/jj.15135955 · Cited by 21
Parcero-Oubiña, C., Güimil-Fariña, A., Fonte, J., & Costa-García, J. M. (2019) · Computational Social Sciences · doi:10.1007/978-3-030-04576-0_14 · Cited by 20
Communication Networks, Interactions, and Social Negotiation in Prepalatial South-Central Crete
Déderix, S. (2016) · American Journal of Archaeology · doi:10.3764/aja.121.1.0005 · Cited by 19
leastcostpath: Modelling Pathways and Movement Potential Within a Landscape
Lewis, J. (2020) · doi:10.32614/cran.package.leastcostpath · Cited by 17
Explaining known past routes, underdetermination, and the use of multiple cost functions
Lewis, J. (2024) · Journal of Archaeological Method and Theory · doi:10.1007/s10816-023-09621-w · Cited by 15
12. Routes through the Landscape: A Comparative Approach
Earle, T. (2010) · University of Pennsylvania Museum of Archaeology and Anthropology eBooks · doi:10.2307/j.ctt3fhjb3.18 · Cited by 15
Groenhuijzen, M. R., & Verhagen, P. (2017) · Journal of Archaeological Science: Reports · doi:10.1016/j.jasrep.2017.07.024 · Cited by 14
The importance of energetics in archaeological least cost analysis
Field, S., Glowacki, D. M., & Gettler, L. T. (2023) · Journal of Archaeological Method and Theory · doi:10.1007/s10816-022-09564-8 · Cited by 13
The Roman Road System in the Golan: Highways, Paths and Tracks in Quotidian Life
Pažout, A. (2017) · Journal of Landscape Ecology · doi:10.1515/jlecol-2017-0022 · Cited by 12
Self-citationItiner-e: A high-resolution dataset of roads of the Roman Empire
de Soto, P., Pažout, A., Brughmans, T., Vahlstrup, P. B., Auir, Á., Bongers, T., Christoffersen, J. E. B., Crépy, M., Johansen, M. H., Lewis, J., et al. (2025) · Scientific Data · doi:10.1038/s41597-025-06140-z · Cited by 11
Self-citationRemote Sensing and GIS Modelling of Roman Roads in South West Britain
Parcero‐Oubiña, C., Smart, C., & Fonte, J. (2023) · Journal of Computer Applications in Archaeology · doi:10.5334/jcaa.109 · Cited by 9
Lewis, J. (2024) · Journal of Archaeological Science · doi:10.1016/j.jas.2024.106055 · Cited by 7
Álvarez, P. A. A., & Fernández, P. T. (2025) · Archaeological and Anthropological Sciences · doi:10.1007/s12520-025-02175-w · Cited by 7
The Road Network in the Longue Durée: A Reading Key of the History of Territories
Watteaux, M. (2017) · Open Archaeology · doi:10.1515/opar-2017-0009 · Cited by 6
Memory at your feet: Modeling the agency of past trails
Llobera, M. (2020) · Journal of Archaeological Science: Reports · doi:10.1016/j.jasrep.2019.102177 · Cited by 6
Ho, Y.-F., Grohmann, C. H., Lindsay, J., Reuter, H. I., Parente, L., Witjes, M., & Hengl, T. (2025) · doi:10.21203/rs.3.rs-6280607/v1 · Cited by 6
Transportation Networks and Least-Cost Paths
Herzog, I. (2023) · The Oxford Handbook of Archaeological Network Research · doi:10.1093/oxfordhb/9780198854265.013.11 · Cited by 5
Pažout, A., Brughmans, T., & de Soto, P. (2022) · Elsevier eBooks · doi:10.1016/b978-0-323-90799-6.00044-6 · Cited by 3
Self-citationRoad and Transport Networks in Hellenistic and Roman Syria
Pažout, A. (2025) · The Oxford Handbook of the Hellenistic and Roman Near East · doi:10.1093/oxfordhb/9780190858155.013.85 · Cited by 2
Self-citationAnalysing Roman itineraries using the GIS tool: the case of the road XIX
Argüelles-Alvarez, P. A., & Trapero-Fernández, P. (2024) · doi:10.21203/rs.3.rs-4132506/v1 · Cited by 2
Carroll, E., & Carroll, F. (2023) · Journal of Archaeological Science: Reports · doi:10.1016/j.jasrep.2023.104140 · Cited by 1
(2013) · Dialnet (Universidad de la Rioja)
Fonte, J., Parcero-Oubiña, C., & Costa-García, J. M. (2017) · Zenodo · doi:10.5281/zenodo.1005562
Flagged first, then by citations
Research works printed but not linked: 5
- Isaac, B., & Roll, I. (1982). Roman Roads in Judaea I: The Legio-Scythopolis Road. British Archaeological Reports.
- Mittmann, S. (1970). Beiträge zur Siedlungs- und Territorialgeschichte des nördlichen Ostjordanlandes. Otto Harrassowitz.
- Naismith, W. W. (1892). Excursions: Cruach Ardran, Stobinian, and Ben More. Scottish Mountaineering Club Journal, 2(3), 135–136.
- Riley, S. J., DeGloria, S. D., & Elliot, R. (1999). A terrain ruggedness index that quantifies topographic heterogeneity. Intermountain Journal of Science, 5(1–2), 23–27.
- Tobler, W. (1993). Three presentations on geographical analysis and modeling: Non- isotropic geographic modeling; Speculations on the geometry of geography. NCGIA Technical Report 93(1), 1-26.
Cited by
Still being checked.
Metadata
- Authors
- Pažout, A.
- Affiliations
- Universitat Autònoma de Barcelona (ES)
- Corresponding author
- Pažout, Adam
- ORCID
- Pažout, Adam
- Venue
- Journal of Archaeological Method and Theory
- Published
- 2026-02-10
- Volume
- 33
- Issue
- 3
- ISSN
- 1072-5369 (ISSN-L)
- 1573-7764
- OpenAlex
- W7128510178
- Semantic Scholar
- 25d3716e1c0771c59135dc193825cb04b92ccaca
- Cited by
- 0
- References
- 44
- FWCI
- 0.00
- Type
- journal article
- Language
- en
- Open access
- hybrid
- doi.org
- Locations
- Journal of Archaeological Method and Theory (publishedVersion, cc-by)
- Research Square (submittedVersion, cc-by)
- Dipòsit Digital de Documents de la UAB (Universitat Autònoma de Barcelona) (submittedVersion, cc-by)
- Funders
- European Commission (101151931)
- Universitat Autònoma de Barcelona
- Topics
- Social Sciences › Archeology › Archaeological Research and Protection
- Social Sciences › Urban Studies › Urban Design and Spatial Analysis
- Social Sciences › Arts and Humanities › Classics › Classical Antiquity Studies
- Keywords
- Roman roads, highway location, topography, topographic position index, isotropic model, slope, road network layout, Near East, cost function
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
A language model drew the claims from the paper on 2026-11-10, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-11-10 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
