Hydrological Modeling and Prehistoric Settlement on Santa Rosa Island, California, USA

Abstract

Fresh water availability was an important variable that influenced prehistoric human settlement on California's northern Channel Islands. Previous attempts to understand settlement on the islands use watershed size as a proxy for water at canyon mouths. In semi‐arid regions, this approach has limitations because streams may lose much or all of their flow to groundwater. We developed a distributed hydrological model for Santa Rosa Island that incorporates geospatial and temporal data for climate (precipitation, solar radiation, wind speed, relative humidity, temperature), soils, vegetation, and topography to simulate the complex land‐surface‐groundwater behavior of island hydrology for hypothetical wet, dry, and median centuries. Our simulations show that water flow is greatest in drainages on the northwest and east coasts of the island. This correlates with some of the earliest and most persistent settlement on the island. During the most extreme droughts of the last 2000 years during the Medieval Climatic Anomaly (1150–600 cal BP), island populations contracted to a small number of large coastal villages. We argue that this was related in part to the greater availability of surface water at these locations. This study expands the theoretical and methodological scope of past studies that have applied hydrological simulation to archaeological investigations.

A Test of Ideal Free Distribution Predictions Using Targeted Survey and Excavation on California’s Northern Channel Islands

“…These drainages do not have the same fresh water flow nor are they as resilient to drought as the highranking drainages along the northwest coast (Jazwa et al 2015b). The middle Holocene occupation at Cow/Lobos Canyons is more extensive than that at La Jolla Vieja Canyon, perhaps because of their relative locations.…”

“…Watersheds were chosen as a unit of analysis for reasons given in Kennett (2005) and Kennett et al (2009). Fresh water availability is watershed dependent (Jazwa et al 2015b;Winterhalder et al 2010), and large primary NCI settlements are almost all along the coast at the mouths of major drainages (Kennett et al 2009;Winterhalder et al 2010;). These suitability rankings were then compared with archaeological evidence for age of first permanent settlement and the persistence of use.…”

“…Occupation of these sites indicates persistent settlement in a given location. We use watersheds as our unit of analysis, but refine the estimates of fresh water flow and drainage resilience using hydrological simulations (Jazwa et al 2015b) to assess the assumption by Winterhalder et al (2010) that watershed size is a good proxy for fresh water availability. We use weights of environmental contributions to suitability similar to those of Kennett et al (2009).…”

“…Climate records suggest that the middle Holocene, particularly the period from about 6500 to 3800 cal BP, was a dry interval dominated by drought conditions (e.g., Antevs 1948Antevs , 1952Antevs , 1955LaMarche 1973LaMarche , 1974Lindström 1990;Morgan et al 1991;Thompson 1992;Cole and Liu 2004;Quade et al 1998;Hughes and Graumlich 2000;Benson et al 2002;Erlandson et al 2005;Kennett 2005;Kennett et al 2007). To examine the local effects of this climate change, Jazwa et al (2015b) modeled fresh water flow and drainage resilience at the mouth of the 19 major drainages on Santa Rosa Island (Table 1). We use these results to generate several different sets of drainage rankings by altering the weights given to the different environmental variables (Table 11).…”

Open science in practice: Learning integrated modeling of coupled surface‐subsurface flow processes from scratch

“…The idea behind the development of PIHM was to support and realize the concept of “community models” for environmental predictions. After decades of development, the utility of PIHM spreads out to the areas of archeology [ French and Duffy , ; Jazwa et al ., ], critical zone processes [ Duffy et al ., ; Shi et al ., ], flood assessment [ Chen et al ., ], forest management [ Yu et al ., ], hydrologic data assimilation [ Shi et al ., , ], landscape evolution [ Zhang et al , ], land surface energy balance [ Shi et al ., ], scalable computing [ Kumar and Duffy , ], and snowmelt runoff simulation [ Kumar et al ., ; Wang et al ., ; Jepsen et al ., ]. A typical application workflow is shown in Figure .…”