Least Cost Path Modeling Between Inka and Amazon Civilizations

Least Cost Path Analysis (LCPA) is a GIS-based approach for calculating the most efficient route between a start and end point, often in terms of shortest time or least amount of energy. The approach is often applied in archaeology to estimate locations of sites, and routes between them. We applied LCPA to estimate how sites in the Andes in the eastern portion of the Inka empire may have connected to sites in the western Amazon Basin. Our approach further used the known Inka Road network to test performance of two types of LCP models (linear vs. areal calculation) and four types of cost functions. LCPs can be calculated with an areal approach, where each cell of the DEM is given one overall slope value, or linearly, where the direction of travel across a cell affects the slope value. Four different algorithms were tested: Tobler's Hiking Function (1993), Tobler's Hiking Function with a vertical exaggeration of 2.3 based on human perceptions of slope (Pingel 2010), Pingel's empirical estimation approach (2010), and Pandolf et al.'s energy expenditure equation (1977) using both an areal and linear approach for all the algorithms. An initial study was conducted in the Cusco region and results were compared to the Inka Road network using the linear accuracy assessment method of Goodchild and Hunter (1997) and Güimil-Fariña and Parcero-Oubiña (2015). The findings suggest that the empirical estimation and caloric cost methods were the most accurate and performed similarly, both were more accurate than travel-time based costs, and linear methods were better than areal based methods when using higher resolution DEM inputs.