Google Earth Engine as a Base for Developing Microregional Pre‐Hispanic Occupation Predictive Models: The High Valleys of Catamarca as a Case Study

ABSTRACT The use of satellite remote sensing data has become one of the most employed methodologies for the initial stages of prospections and detection of archaeological features, especially considering the increasing availability and low costs of these approaches. Nowadays, it is even possible to work with the capabilities of freely available planetary‐scale cloud computing platforms such as Google Earth Engine (GEE) and make use of the data accumulated over the last five decades in novel ways. Here, we introduce a methodology that combines ethnographic information and the calculation of a widely used vegetation index from several datasets in order to predict the location of archaeological occupations in areas which cannot be easily accessed. The selection of satellite images to be used takes three factors into consideration: temporal amplitude, spatial resolution and product quality. We are also interested in studying the differences that may exist in the application of the same methodology on different satellite image datasets. Due to geomorphological conditions and archaeological survey strategies, we have to deal with vast unprospected areas above 3300 m above sea level and the subsequent lack of information. Thus, we decided to work in the identification of vegetation (and water) patches that could have attracted the concentration of human occupations, paying special attention to recurrence and predictability, factors that could have favoured the return of people at different times, leaving more intense signs on the surface. This strategy has allowed us not only to advance in the identification of new spaces with a recurrent availability of water and vegetation, known in the literature and among locals as ‘vegas de altura’ (translated as highland marshes or peatlands), register features that could correspond to archaeological occupations and plan pedestrian prospections there. In addition, the applied methodology can also be used to monitor changes in the conditions of growth and decline of these fragile environments.