Machine Learning-Based Detection of Archeological Sites Using Satellite and Meteorological Data: A Case Study of Funnel Beaker Culture Tombs in Poland
Remote Sensing · 28.06.2025 · 17 (13) · 2225
DOI: 10.3390/rs17132225 ↗Kurz gefasst
Fachartikel in der Zeitschrift Remote Sensing, online erschienen am 28.06.2025. Verfasst von Krystian Kozioł und 5 weiteren Autor:innen. Die Arbeit wurde bisher 4-mal zitiert (Crossref) und nennt 78 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Ministry of Science and Higher Education (Poland) for AGH University of Krakow, “Excellence Initiative—Research University” program at the AGH University of Krakow.
Abstract (Originalsprache)
The detection of archeological sites in satellite imagery is often hindered by environmental constraints such as vegetation cover and variability in meteorological conditions, which affect the visibility of subsurface structures. This study aimed to develop predictive models for assessing archeological site visibility in satellite imagery by integrating vegetation indices and meteorological data using machine learning techniques. The research focused on megalithic tombs associated with the Funnel Beaker culture in Poland. The primary objective was to create models capable of detecting archeological features under varying environmental conditions, thereby enhancing the efficiency of field surveys and reducing associated costs. To this end, a combination of vegetation indices and meteorological parameters was employed. Key indices—including the Normalized Difference Vegetation Index (NDVI), Normalized Difference Water Index (NDWI), Normalized Difference Moisture Index (NDMI), and Normalized Archeological Index (NAI)—were analyzed alongside meteorological variables such as wind speed, temperature, humidity, and total precipitation. By integrating these datasets, the study evaluated how environmental conditions influence the visibility of archeological sites in satellite imagery. The machine learning models, including logistic regression and decision tree-based algorithms, demonstrated strong potential for predicting site visibility. The highest predictive accuracy was achieved during periods of high soil moisture variability and fluctuating weather conditions. These findings enabled the development of visibility prediction maps, guiding the optimal timing of aerial surveys and minimizing the risk of unsuccessful data acquisition. The results underscore the effectiveness of integrating meteorological data with satellite imagery in archeological research. The proposed approach not only improves site detection but also reduces operational costs by concentrating resources on optimal survey conditions. Furthermore, the methodology is applicable to diverse archeological contexts, enhancing the capacity to locate and document heritage sites across varying environmental settings.
Förderung
- Ministry of Science and Higher Education (Poland) for AGH University of Krakow
- “Excellence Initiative—Research University” program at the AGH University of Krakow
Zitieren
APA
Kozioł, K., Borowiec, N., Marmol, U., Rzeszutek, M., Santos, C. A. G., & Czerniec, J. (2025). Machine Learning-Based Detection of Archeological Sites Using Satellite and Meteorological Data: A Case Study of Funnel Beaker Culture Tombs in Poland. Remote Sensing, 17(13), 2225. https://doi.org/10.3390/rs17132225
Vancouver
Kozioł K, Borowiec N, Marmol U, Rzeszutek M, Santos CAG, Czerniec J. Machine Learning-Based Detection of Archeological Sites Using Satellite and Meteorological Data: A Case Study of Funnel Beaker Culture Tombs in Poland. Remote Sensing. 2025;17(13):2225. doi:10.3390/rs17132225
BibTeX
@article{Kozio2025,
author = {Kozioł, Krystian and Borowiec, Natalia and Marmol, Urszula and Rzeszutek, Mateusz and Santos, Celso Augusto Guimarães and Czerniec, Jerzy},
title = {Machine Learning-Based Detection of Archeological Sites Using Satellite and Meteorological Data: A Case Study of Funnel Beaker Culture Tombs in Poland},
journal = {Remote Sensing},
year = {2025},
volume = {17},
number = {13},
pages = {2225},
doi = {10.3390/rs17132225}
}
EUVIMED