Home Geospatial Applications Archaeology Archaeological site selection

Archaeological site selection

Asis Bhattacharya and K. Vinod Kumar
NRSA, Hyderabad

G. S. Narasimhan
Archaeological Survey ofIndia, Hyderabad

The launch of high resolution PAN data onboard IRS-1C has tremendously enhanced the application capability for many thematic mapping. The application of remote sensing technology in the field of archaeology is in its infancy in the country and archaeologists are realising the importance of this tool for their survey. There is very little published literature in the field of application using this technology. The present study may interest the researchers to apply this tool for finding new archaeological sites. The conventional survey involves a detailed fieldwork wherein the archaeologist has to visit village by village and find out any new archaeological artefacts from local sources. The use of remote sensing helps them to narrow down the target area for search through some geomorphic indicators. The PAN data has a spatial resolution of 5.8 metres with synoptic coverage of 72×72 km. hence, large areas can be observed with good spatial resolution. Further, expert knowledge from the field of archaeology can also be used to deduce the signature of known sites in these satellite images for searching in the unknown area. An integrated approach using collateral information along with remote sensing data in GIS environment is ideal for this type of study.

Study Area
The study area is located 60 km. west-northwest of Hyderabad city on Hyderabad-Mumbai highway near Kondapur village in Medak district.

The site was first explored by Henry Cousens in 1900. It is noted for its Middle Stone Age factory site near the stream. The tools found here include thick lunates, blades, flakes and tortoise cores. Interestingly, the late Stone Age (Microliths) blades, cores and geometric continuity of both the traditions. Some of the Neolithic instruments were found from the nearby hill slopes. This site was discovered by the archaeologists as a low lying mound approximately 5 m from the ground level between Kondapur and Terpol village. The excavation of this site has revealed a fortified township of the Satavahanas dynasty, founded by Simuka, who at the zenith of their power, ruled a vast kingdom extending between the seas across the Deccan Plateau.

Methodology and Analysis
This site being known as archaeologically important and covering a large area, was taken as a reference to deduce spectral and geomorphic signatures from the satellite data for this particular terrain. Later, these signatures were correlated with other information like drainage, topoheight and water bodies to create a knowledge-based database. This database can be used for other unknown and unexplored sites. Thus, a rule-based search using weighted index method can be employed.

The analysis was carried out in the ERDAS imagine image processing and Arc/Info software in Dec ALPHA work station. The various thematic layers like geomorphology, water bodies, drainage, road network and slope were generated. A favourability index model was used to derive the most favourable zone by using existing information about the known site.

A geomorphological map was prepared from the satellite data to understand the general physiography of the area. Genetically, the area can be grouped under arid landforms. The landforms mostly seen are inselbergs, pediments and valley fills. The inselbergs are seen mostly near Nagapur village, northeast to the site. The other important landform is the shallow buried pediplain, which has a bright tone in the image. The weathered zone is very thin in this landform with bedrock exposed at shallow depth. The shallow buried pediplain grade into deep brued pediplain, which has a thick soil cover and supports vegetation. The weathering of granitic inselberg generally occurs along three sets of joints. They are the weak planes in this hardrock. Along these weak planes, drainages develop, which follow the joints of this rock.

Because of weathering in course of time, valley-fill deposits develop along these zones. This is the most fertile plain in this zone. The intersection of multiple joints results in deep depression because of the weakness. This develops into small natural water tanks. Many man-made tanks are also there, developed after constructing bunds across the streams. Hence, the area as seen in the satellite image has numerous water tanks. These are the favourable areas for settlements. Most of the settlements, past or present, are always located near to the water body.

Results and Discussions
It is observed that tha existing Kondapur site of excavation is located on a low-lying mound (topohigh) near a large water body within a high valley-fill, which was an ancient drainage channel. These parameters were given weightage, followed by water body and valley-fills. Using the known weightages, unknown sites were selected for future explorations. The database also contains the existing roads and other infrastructure facilities around this site. This information is of immense help for an exploration programme. The satellite data also gives an update information about the urban sprawls and settlement around the sites. These are useful in conservation work of the archaeologically important sites. The temporal data can be used to monitor these sprawls and detect any encroachment into such important sites.

Conclusion
The database can be updated regularly as and when new sites are discovered. This is also cost-effective technology since this can be developed in a personal computer (Laptop) and can be carried out in the field. The total cost of this project covering 50×50 sq. km. area comes to approximately Rs. 15,000 (data and processing cost). This work can be carried out in seven days in the lab or in the field. The data collection when supported by GPS is more accurate and precise. Thus, it can be summarised that this technology is cost-effective, timely and accurate. This is an essential tool for the archaeologists for their exploration programme.

(Based on an article published by the authors in NNRMS Bulletin, May 1998.)