Global Land Use and Technological Evolution Simulations to Quantify Interactions Between Climate and Pre-industrial Cultures - Integrated Analysis of Interglacial Climate Dynamics (INTERDYNAMIC) - page 98

much did they disturb global and regional carbon and hydrological cycles? When
did the anthropocene begin (Kaplan et al. 2011 )? To understand the two-way
interaction between past societies and Holocene climate de

nes a challenge for
transdisciplinary research and for testing controversial hypotheses of a middle
Holocene in uence of humans on climate.

We addressed these questions by using an interactively coupled model system
composed of a cultural adaptation model [Global Land Use and technological
Evolution Simulator, GLUES (Lemmen and Wirtz 2010 , 2012 ; Lemmen et al.
2011 )] and an Earth System Model [Planet Simulator, PLASIM (Haberkorn 2013 ;
Haberkorn et al. 2012 )]. Cultural feedback on climate is implemented by land
surface changes. The realism of the interactive simulation of climate and culture is
improved by constraining the climate model with temperature and precipitation
proxies, and by constraining the cultural model with archeological data compila-
tions and vegetation proxies indicative of human land use. Abrupt climate changes
are included based on globally available time series of proxy-derived climate
variability (Wirtz et al. 2010 ). Models and data employed in our study cover the
period 11.5 thousand years (ka) before present (BP) to 3 ka BP and are global in
scope. Carbon emissions are evaluated regionally, and the sociotechnological
model is validated against regional archeological data for Northern Central Europe
and South Asia.
We propose to integrate the dynamic anthroposphere into today ' s state-of-the-art
Earth System Models (ESM) as a prerequisite to better understand current human-
climate interaction and adaptation to ongoing climate change. Current and antici-
pated users of our work are paleoclimate and paleovegetation modelers, paleoclimate
variability analysts, archeologists, and agricultural economists.

2 Materials and Methods

A large data set of 235 long-term (>4,000 years) and high-resolution (mostly
<100 years) time series of climate information have been collected from the liter-
ature and Interdynamic partners. Based on a change-point analysis, we partitioned
the Holocene into slightly overlapping periods, the early Holocene (11 5 ka BP)

-

and late Holocene (6 0 ka BP). For each interval, we evaluated each proxy time

series for statistically signi

-

cant periodic signals, using very strict and data-adaptive

thresholds for signi cance.

For simulations, we chose the PLASIM (Haberkorn 2013 ; Haberkorn et al.
2012 ) ESM which can be used to run climate simulations for multi-millennial time
scales in acceptable real time while relying on a fully dynamic core; it also offers
different vegetation couplers (Haberkorn 2013 and Fig. 2 ). We performed full
Holocene transient simulations at T21 and T42 resolutions with orbital, greenhouse
gas (GHG) forcing, and climatological sea-surface temperature (SST). A novel
scheme was devised to reconstruct past SST from the sensitivity of land temperature
to SST diagnosed from a comparison between present day climate and present day