Contents lists available at ScienceDirect Journal of Archaeological Science journal homepage: www.elsevier.com/locate/jas High-resolution remote sensing and advanced classification techniques for the prospection of archaeological sites’ markers: The case of dung deposits in the Shashi-Limpopo Confluence area (southern Africa) Olaotse Lokwalo Thabeng a,b,* , Stefania Merlo b , Elhadi Adam b a University of Botswana, 4775 Notwane Rd, Private Bag, UB 00703, Gaborone, Botswana b University of the Witwatersrand, 1 Jan Smuts Avenue, Braamfontein, 2000, Private Bag 3, Wits, 2050, Johannesburg, South Africa ARTICLE INFO Keywords: Remote sensing Farming communities Vitrified dung Non-vitrified dung Survey Prospection Advanced classification ABSTRACT Archaeological prospection through remote sensing is based on the contrast between areas of archaeological interest and their surroundings. It has been used as the cheapest and the fastest way of locating and documenting areas of archaeological interest since the 1920s, initially with the aid of film-based aerial photographs. In recent years, there has been a shift towards the use of multispectral satellite data in prospecting for archaeological sites because of their ability to give information on spectral characteristics of archaeological material beyond the visible spectrum. However, spectral signatures for identifying archaeological sites are not universal, and an assessment of the applicability of remote sensing techniques in different archaeological landscapes is needed. This study tests the feasibility of prospecting for archaeological sites previously occupied by farming commu- nities in the Shashi-Limpopo Confluence Area of southern Africa, using very high-resolution satellite WorldView- 2 images. It also assesses the performance of advanced classification algorithms (support vector machine and random forest) and the contribution of new WorldView-2 bands in detecting archaeological sites. Two in- dependent accuracy assessments were carried out, using a data set collected by Huffman (2011, 2009a) and a randomly generated holdout test dataset, respectively. The results demonstrate the potential of remote sensing methods in prospecting for archaeological sites previously occupied by farming communities using very high- resolution satellite images and advanced classification algorithms. Very high overall accuracies were achieved by: random forest, 95.29% using holdout sample and 97.71% using independent dataset; support vector ma- chine, 88.82% using holdout sample and 95.88% using independent dataset, respectively. The new WorldView-2 bands were of least importance (compared to traditional bands) in detecting sites in Shashi-Limpopo Confluence Area. Despite high classification accuracies achieved by both classifiers, there were some misclassifications between vitrified dung sites and river sand. Therefore, to address this problem, this study recommends the use of robust classifiers such as object-based algorithms because of their ability to segment an image into homogenous objects and classify them using a combination of spectral, textual, sub-pixels, spatial, relational and contextual methods. 1. Introduction Archaeological prospection through remote sensing offers a prac- tical and economical means to detect and characterise different types of archaeological sites, over traditional field walking survey methods (Beck, 2007). It also reduces the biases inherent in field walking survey methods, which can be influenced by the researcher's tendency to sample hotspots, leaving out other parts of the study area (Bintliff, 2000; Hawkins et al., 2003; Renfrew and Bahn, 2012). Most im- portantly, the use of aerial and satellite images gives archaeologists a synoptic view of archaeological landscapes and helps them understand patterns which may not be visible when on the ground (Crawford, 1923; Hritz, 2010). The increased spatial, temporal and spectral re- solutions of satellite remotely sensed data offer opportunities for dis- criminating different archaeological traces and detecting their change over time (Agapiou et al., 2015; Doneus et al., 2014; Hadjimitsis et al., 2009; Verhoeven and Vermeulen, 2016), including the monitoring of looting activities (Hesse, 2015; Parcak, 2015; Van Ess et al., 2006). Additionally, remotely sensed digital data can be integrated with var- ious other datasets for advanced analysis (Hesse, 2015; Megarry et al., https://doi.org/10.1016/j.jas.2018.12.003 Received 25 March 2018; Received in revised form 21 November 2018; Accepted 11 December 2018 * Corresponding author. Postal Address: Private Bag, UB 00703, Gaborone, Botswana E-mail addresses: thabengl@ub.ac.bw (O.L. Thabeng), stefania.merlo@wits.ac.za (S. Merlo), elhadi.adam@wits.ac.za (E. Adam). Journal of Archaeological Science 102 (2019) 48–60 0305-4403/ © 2018 Elsevier Ltd. All rights reserved. T