Documenting more than 5000 ha in Mongolia in five days using low-cost drones

Keywords: Mongolia, UAV, low-cost, 3d reconstruction

Abstract:
Over the last decades, progress in developing Unmanned Aerial Vehicles (UAV) and 3d reconstruction software is noticeable, which extends into documentation methods in archaeology as a standard component. In a little while, UAVs record data automatically and 3d reconstruction software produces georeferenced 3d models to map i.e. excavation steps accurately. Different tools are available to record data automatically. The archaeocopter team (HTW Dresden, Germany) ,and the German Archaeological Institute (DAI) , the Mongolian Academy of Science (MAS) and the Mongolian State University work together in a documentation campaign in Mongolia in October 2018. The objective of the campaign is the documentation of remaining surface structures of two major archaeological sites in Mongolia: Karabalgasun, the capital of the Uyghur Empire from 745 to 840 CE and Karakorum, the capital of the Mongol Empire from 1220 to 1260 CE. Both are located in the UNESCO world heritage site of the Orkhon Valley and are subject to Mongolian-German joint research efforts since 1999. Beside relatively small documentations in Karakorum of about 500 ha, the team records large-scale areas of about 8000 ha close to Karabalgasun. It was necessary to find a way to document such large area with low-cost drones in about 10 days. As previously reported, with accurate planning beforehand the archaeocopter team was able to document 13 desert fortresses in Bukhara/Uzbekistan in 2015 in only 5 days. During the preliminary stages, all flights and a potential parallelization of operations which are executed by several teams need to be elaborated. The main challenges are the coordination of the teams and flights in an area without internet and visual markers, which would facilitate the orientation in the field. Also, a sufficient but not too large overlapping of different flight sectors is necessary. The planning part beforehand took several months and this paper presents findings and solutions as guidance for further documentations of such huge areas which do not have a well-developed infrastructure. This paper starts with the calculations and considerations of test flights in the preparatory phase. After presenting the results of this campaign the validity, robustness and scalability of these results for large-scale documentations will be discussed.