Digging up the past can change the ground beneath it

Istanbul, Türkiye – Archaeological excavation uncovers the past but also changes the landscape that has preserved sites for centuries or millennia. These changes influence water flow, soil erosion, and the stability of exposed remains. Researchers from Istanbul Technical University and the U.S. have shown how computer models can assist archaeologists in understanding and planning for these effects, especially regarding erosion risks. Using an Early Bronze Age settlement in Türkiye as a case study, they reconstructed the landscape before and after excavation. Their findings indicate that excavation significantly impacts runoff and erosion, with the effects heavily influenced by the location of soil removal, placement of excavated materials, and terrain shape and steepness. The study was published in Geoderma Regional.

When protecting history changes the landscape

The researchers investigated Resuloğlu Mound in north-central Anatolia, a settlement approximately dating from 2300 to 2100 BCE. Archaeological excavations occurred there from 2003 to 2019. The site presents a unique conservation challenge, as erosion had been shaping the landscape long before modern excavation efforts. Interestingly, erosion played a role in exposing graves, which contributed to the decision to excavate. However, excavation introduced new disturbances, such as exposing archaeological structures and altering slope shapes and properties. Additionally, excavated material was often deposited downslope, which could redirect water flow and lead to new erosion-prone areas. This situation highlights a key dilemma in archaeology: exposing cultural heritage can inadvertently alter environmental conditions and threaten the site's future stability.

“Archaeological excavation is essential for understanding our past, but it also changes the landscape. Our study shows that we can examine those changes scientifically and use that information to think about how archaeological sites are excavated and protected,” said Kemal Koçaklı, of the Eurasia Institute of Earth Sciences, Department of Evolution and Ecology, at Istanbul Technical University.
 

Reconstructing the landscape before and after excavation

To determine what had changed, the team integrated various data sources including historical aerial photos, drone surveys, satellite images, soil and weather records, archaeological archives, and detailed terrain analysis. Aerial photos from 1990 helped recreate the landscape before excavation, while drone images from 2019 offered a detailed view after excavation. The researchers then utilized QGeoWEPP, a QGIS interface for the Water Erosion Prediction Project (WEPP), to model water and soil movement across eight hillslopes, six disturbed by excavation and two undisturbed reference slopes. The model was run with 24 years of weather data from 2000 to 2024, enabling comparisons of average runoff, soil loss, and sediment yield under reconstructed pre- and post-excavation conditions. Unlike simpler erosion models, this method captures how water and sediment travel along specific slopes and how changes like bare soil, compaction, and soil deposition influence that movement. This allows for identifying erosion hotspots and assessing practical options, such as optimal locations for excavated soil placement. 

“Among today’s soil erosion models, QGeoWEPP is, in our view, the best suited for protecting areas at risk of erosion, because it simulates the physical processes slope by slope instead of producing rough averages. That strength comes from decades of high-quality work on WEPP at the USDA’s National Soil Erosion Research Laboratory (NSERL) at Purdue University in West Lafayette, Indiana, and this study would not have been possible without the expertise of our partners there,” said Kemal Koçaklı.

Excavation changed erosion in different ways

The results indicated that excavation effects varied and were not uniform. When excavation removed soil cover and exposed harder, pavement-like archaeological surfaces, rainfall infiltrated less freely, causing more water to flow across the surface. For example, on one cemetery slope, the modeled average annual runoff increased from 12.6 to 20.1 millimeters per year after excavation. Meanwhile, average annual soil loss went up from 1.1 to 2.1 kilograms per square meter, and sediment yield increased from about 11 to 21 tonnes per hectare per year. In other cases, depositing excavated soil created different issues; thicker soil sometimes enhanced infiltration and lowered runoff, but loose, fine-grained material remained prone to erosion, especially on steep slopes without vegetation or stabilization. One settlement slope exhibited a modeled increase in average annual soil loss from 1.1 to 4.9 kilograms per square meter after excavation. Overall, the changes’ magnitude and direction varied significantly across slopes, highlighting the importance of assessing erosion risks based on specific terrain and disturbance rather than assuming all excavations behave similarly.

From measuring damage to planning excavations

The researchers see broader potential in examining individual slopes. Soil erosion at archaeological sites has been previously studied, with GIS methods used to identify heritage locations at risk. However, this approach uniquely models excavation as part of the erosion process itself, comparing landscapes before and after excavation and analyzing how decisions made during excavation can affect future erosion. This could, in time, allow erosion modeling to be integrated into excavation planning. Archaeologists and conservation efforts could evaluate vulnerable slopes, determine optimal locations for soil placement, and identify areas needing stabilization or protection before digging begins. For example, at Resuloğlu, the team found that placing excavated material on steep slopes increased erosion risk. They suggest that positioning and stabilizing this material on gentler slopes, or using it to refill excavated structures, might help reduce erosion.

“The goal is not to tell archaeologists not to excavate. It is to give them another tool for deciding how to excavate with less impact. If we can identify where erosion is likely to occur before work begins conservation can become part of excavation planning rather than something considered only after damage appears. In a sense, the model gives archaeologists the opportunity to excavate digitally before they excavate in the field: we can try different scenarios on screen and choose the one that leaves the landscape most stable,” explained Kemal Koçaklı.

Protecting more than the excavation trench

The study highlights a broader lesson: protecting an archaeological site involves considering more than just the visible ruins. At Resuloğlu, factors such as surrounding gullies, livestock movements, water sources, and the placement of excavated soil all influence erosion processes. The authors suggest that conservation efforts should take into account the entire landscape, not just individual trenches or structures.

They also warn that their findings are primarily based on modeling rather than direct erosion measurements at the site. While simulated rainfall and crop yields were compared with regional data to validate the hydrological model, future work should include field measurements of runoff and sediment to improve accuracy.

This approach promotes a new perspective for archaeological sites dealing with increasing environmental issues. Excavation not only uncovers historical landscapes but also forms a new environment that needs to last. By anticipating water and soil behavior before, during, and after excavation, archaeologists can learn about the past while also preparing the terrain for future conditions.