remote sensing for archaeological studies and territory...

Remote sensing for archaeologicalstudies and territory management:case study of the roman city of LucusAsturum (Asturias, Spain)

Otilia Requejo PagésDepartment of Cultural Heritage. Principality of Asturias. SpainJavier F. CallejaDepartment of Physics. Universityy of Oviedo. Spain

Remote sensing for archaeological studies andterritory management: case study of the roman cityof Lucus Asturum (Asturias, Spain)

ObjectiveTo demonstrate the capability of spaceborn sensors to detect archaeological remains.

1.- Study area: Geography and History2.- Selection of the sensor3.- World View 2 features4.- Algorithms applied to the WV2 data5.- Conclusions and future work


Asturias

Study area: Geography

The area of study (Llanera, Principality of Asturias, Spain) is located in the central western part of the Cantabrian coast in the north of the Iberian peninsula.

Principality of Asturias

Iberian Peninsula


Asturias

Due to its geographic situation itforms part of the Asturiandepression bordering the sea, ahorizontal band which runs fromeast to west across the Asturianterritory and which topographicallyis a sunken flattened surface inbetween two lines of low summits.

The geomorphology mainly reflects the action offluvial processes being the course of the Norariver, which stands out, both in length andextension, as one of the fluvial meadows whichconstitute the most significant element in thelandscape of Llanera at its southern limit. To thenorth, the plain of Llanera is bordered by hillsand medium-sized slopes which do not surpassaltitudes of 500 meters and which separate thisvalley/coalfield from the coastal area.


Asturias

Lucus Asturum is one of three known and documented centers of population during the Romanperiod in Asturias. It is cited for the first time in the middle of the second century A.D by the Greekgeographer Claudio Ptolomeo, where it is mentioned alongside the names of cities and villages ofthe Astures.

Study area: History


Asturias

The center of this site is located on a surrounding area of flat ground of the parish church of Santa Mariade Lugo de Llanera, in the Eria de la Castañera, the central part of the territory of Llanera. Thearchaeological excavations undertaken in the area around the temple confirm material evidence of Romanoccupation beginning in the middle of the first century A.D. The archaeological remains identified formedpart of a vicus viari, a habitat or built-up area of secondary type composed of various separateconstructions or forming small groups with unoccupied medium-sized spaces. The area occupied by theentire complex would have been approximately 50 hectares.

Study area: History


Selection of the sensor

Figure taken from:R. Lasaponara, and N. Masini, Journal of Archaeological Science 34, 214-222, 2007

Optical range (VNIR). Working hypothesis: buried remains have an impact on vegetation

Radar ruled out: Lack of the required expertise, Vegetated humid ground


Selection of the sensor

Spatial resolution

Size of the searchedpatterns:

0.1 m – 10 m

Spectral resolution

Detection of vegetation changes:Visible , NIR

Temporal resolutionNot important.

Study area has notchanged much for years


World View 2

Spectral resolution Acquisition dateSpatial resolution

Pahchromatic0.5 m

8 Multispectral2 m

11 October, 2011


Algorithms

Real color Principal components Crop – Veg –Soil * NDVI

𝑁𝐷𝑉𝐼 =𝑁𝐼𝑅 − 𝑅

𝑁𝐼𝑅 + 𝑅

R = Band 5NIR = Band 7

𝐶𝑟𝑜𝑝 = −0.38ρ𝑏𝑙𝑢𝑒 − 0.71ρ𝑔𝑟𝑒𝑒𝑛 + 0.20ρ𝑟𝑒𝑑 − 0.56ρ𝑁𝐼𝑅

V𝑒𝑔 = −0.37ρ𝑏𝑙𝑢𝑒 − 0.39ρ𝑔𝑟𝑒𝑒𝑛 − 0.67ρ𝑟𝑒𝑑 + 0.52ρ𝑁𝐼𝑅

S𝑜𝑖𝑙 = 0.09ρ𝑏𝑙𝑢𝑒 + 0.27ρ𝑔𝑟𝑒𝑒𝑛 − 0.71ρ𝑟𝑒𝑑 − 0.65ρ𝑁𝐼𝑅

PC 1

PC 2

Blue = Band 1Green = Band 3Red= Band 5NIR= Band 6

* A. Agapiou, D. D. Alexakis, A. Sarris, D. Hadjimitis, Remote Sensing 5, 6560-6586, 2013.

R = Band 5G = Band 3B = Band 2


Real color RGB composite of the WV2 image

Study area


Real color RGB composite of the study area


Areas with special features/patterns will be marked in a red circle

When no special features/patterns are observed, the area will be marked in a yellow circle

In this work we focus our attention on three areas inside the study area


NDVIReal color RGB composite

YES Special features/patterns obseved

NONo features/patterns obseved


Real color RGB composite Principal Component 1 Principal Component 2


Real color RGB composite Crop Component Veg Component


NDVI Crop Component Veg Component


NDVI vs Crop Band NDVI vs Veg Band


Lidar Crop Component Veg Component


Kun adaptive filter on Crop Component


Conclusions and future work

1.- We have proven the capability of WV2 to detect regular patterns and features invegetated areas.2.- Different features are highlited using different algorithms:Principal components analysis seems to highlight any vegetation special feature.NDVI and Veg component are highly correlated, and they provide similar information.Crop component provides information not contained in the NDVI.Results are compared with Lidar results: Some features can be related to topography.3.- Features detected by different algorithms might have different origins. Some ofthem could be related to buried archaeological remains.4.- The results are consistent with the morphology of the site, archaeologicallycharacterized as a vicus viari: a secondary and dispersed agglomeration developedaround communications crossings5.- The results will contribute to an effective planning of the archaeological work onsite as well as facilitate the implementation of preventive and corrective measures inthe urban management of this territory.

Future work: DN conversion to TOA reflectance , atmospheric correction.Identification of the origin of the features.In-situ validation.


Acknowledgments

The authors would like to thank Fotoasturias S. L. for purchasing the World View 2data for this work

remote sensing for archaeological studies and territory...

Documents