The Art and Role of Climate Modeling Institute of Coastal Research, GKSS Research Centre Geesthacht, Hans von Storch Turku, 5-9 June 2001, CLIC Symposium 45 minutes incl.discussion "Climate Modelling„ Abstract Climate models are an indispensable tool for modern climate research; based on detailed descriptions of all first order climate processes, these models simulate climate variables in a quasi-realistic manner on time scales of a few hours to centuries of years and spatial scales of a few hundred kilometres. These models generate a virtual reality or laboratory which allow the reconstructing of past (historical and paleoclimatic) states and of plausible scenarios of future changes. In this talk, the art of quasi realistic climate modelling is reviewed. Its limitation - such as the failure to immediately constitute knowledge (insight into climate dynamics) or to provide regional detail, or the impossibility for positive verification - is discussed. The different modes of applications are sketched: simulation runs in the paradigm of initial/boundary value problems, and data analysis in the spirit of a state space formulation. The talk is concluded with a short discourse about the contemporary public role of climate models. based on: von Storch, H., S. Güss und M. Heimann, 1999: Das Klimasystem und seine Modellierung. Eine Einführung. Springer Verlag ISBN 3-540-65830-0, 255 pp von Storch, H., and G. Flöser (Eds.), 2001 Models in Environmental Research. Proceedings of the Second GKSS School on Environmental Research, Springer Verlag ISBN 3-540-67862 The Art and Role of Climate Modeling

Overview: Conceptual aspects of modelling Institut für Küstenforschung I f K Overview: Conceptual aspects of modelling Quasi-realistic climate models („surrogate reality“) Free and forced simulations for reconstruction of historical climate Laboratory to test conceptual models Climate change simulations

Institut für Küstenforschung Hesse’s concept of models I f K Hesse’s concept of models Reality and a model have attributes, some of which are consistent and others are contradicting. Other attributes are unknown whether reality and model share them. The consistent attributes are positive analogs. The contradicting attributes are negative analogs. The “unknown” attributes are neutral analogs. Validating the model means to determine the positive and negative analogs. Applying the model means to assume that specific neutral analogs are actually positive ones. The constructive part of a model is in its neutral analogs. Hesse, M.B., 1970: Models and analogies in science. University of Notre Dame Press, Notre Dame 184 pp.

Positive analog Neutral analog Application

Quasi-realistic Modelling Institut für Küstenforschung I f K Quasi-realistic Modelling

Models are Institut für Küstenforschung I f K Models are • • • smaller than reality (finite number of processes, reduced size of phase space) • • • simpler than reality   (description of processes is idealized) • • • closed, whereas reality is open   (infinite number of external, unpredictable forcing factors is reduced to a few specified factors)

Institut für Küstenforschung I f K

Institut für Küstenforschung I f K

Institut für Küstenforschung I f K

Models represent only part of reality; Institut für Küstenforschung I f K Models represent only part of reality; Subjective choice of the researcher; Certain processes are disregarded. Only part of contributing spatial and temporal scales are selected. Parameter range limited

Institut für Küstenforschung I f K Models can be shown to be consistent with observations, e.g. the known part of the phase space may reliably be reproduced.

Models can not be verified because reality is open. Institut für Küstenforschung I f K Models can not be verified because reality is open. Coincidence of modelled and observed state may happen because of model´s skill, or because of unknown external influences not accounted for by the model.

Models as surrogate reality Institut für Küstenforschung I f K Models as surrogate reality dynamical, process-based models • experimentation tool (test of hypotheses) • design of scenario • sensitivity analysis • dynamically consistent interpretation and extrapolation of observations in space and time (“data analysis”) • forecast of detailed development (e.g. weather forecast)

quasi-realistic climate models Institut für Küstenforschung I f K quasi-realistic climate models

Institut für Küstenforschung I f K Dynamical processes in the atmosphere

Institut für Küstenforschung I f K Dynamical processes in a global atmospheric general circulation model

Institut für Küstenforschung I f K Dynamical processes in the ocean

Institut für Küstenforschung I f K Dynamical processes in a global ocean model

Institut für Küstenforschung I f K

Institut für Küstenforschung I f K 1880–2049 ECHAM3/LSG 1973–1993 ERA ECMWF

Institut für Küstenforschung I f K

Institut für Küstenforschung I f K How well are these processes represented in climate models? atmosphere ocean Bray and von Storch, 1999 Results of a survey among climate modellers

Laboratory to test conceptual models Institut für Küstenforschung I f K Laboratory to test conceptual models

Example: Stommel model of the North Atlantic overturning Institut für Küstenforschung I f K Example: Stommel model of the North Atlantic overturning Ft, Ht freshwater and heat flux Fp, Hp Subtropical Atlantic Tt,St Subpolar Atlantic Tp, Sp Trans-port

Institut für Küstenforschung Rahmstorf‘s model Stommel‘s theory I f K Rahmstorf‘s model Stommel‘s theory Rahmstorf, 1995

Testing the of multimodality of large scale atmospheric dynamics Berner and Branstator, pers. comm

Free and forced simulations for reconstruction of historical climate Institut für Küstenforschung I f K Free and forced simulations for reconstruction of historical climate

Free Simulation: 1000 years no solar variability, no changes in greenhouse gas concentrations, no orbital forcing Institut für Küstenforschung I f K Temperature (at 2m) deviations from 1000 year average [K] Zorita, 2001

Forced Simulation 1550-2000 simulation Changing solar forcing and time variable volcanic aerosol load; greenhouse gases

Climate model used Institut für Küstenforschung Atmosphere: ECHAM4 I f K Climate model used Atmosphere: ECHAM4 horizontal resolution T30 ~ 300 km at mid latitudes Ocean: HOPE-G horizontal resolution T42 ~ 200 km at mid latitudes increased resolution in the tropics Model provided as community climate by Model & Data Group at MPI for Meteorology and run at German Climate Computing Centre (DKRZ) and computing facilities at FZ Jülich

Institut für Küstenforschung I f K

Institut für Küstenforschung I f K Temperature conditions in Switzerland according to Pfister‘s classification. From Luterbacher, 2001

Late Maunder Minimum 1675-1710 vs. 1550-1800 validation Reconstruction from historical evidence, from Luterbacher et al. Late Maunder Minimum Model-based reconstuction 1675-1710 vs. 1550-1800

Global 1675-1710 temperature anomaly Institut für Küstenforschung I f K Model as a constructive tool Global 1675-1710 temperature anomaly

Institut für Küstenforschung Model as a constructive tool I f K Model as a constructive tool Simulated differences of ice coverage, in percent, during the LMM event 1675-1710 and the long term mean 1550-1800.

Institut für Küstenforschung I f K control forced

Climate change simulations Institut für Küstenforschung I f K Climate change simulations

Institut für Küstenforschung I f K

Institut für Küstenforschung I f K Scenario A2 Annual temperature changes [°C] (2071–2100) –(1961–1990) Scenario B2 Danmarks Meteorologiske Institut

Institut für Küstenforschung I f K Consistency of regional change in precipitation in different climate model scenarios. (consistency: 7 out of 9 scenarios agree)

Quasi-realistic climate models Institut für Küstenforschung I f K Quasi-realistic climate models are useful tools to describe our knowledge about the climate processes and their interactions in a comprehensive manner. are a reduction of the real complexity of the climate system. They may be used to construct knowledge, but do not by themselves represent more knowledge than was inserted into. are the only tool to conduct experiments with the environmental system. are a valuable tool to reconstruct past climate variations and to envisage possible future climate developments. cannot be “verified”, i.e., shown to respond “correctly” to anomalous forcing if this has not been observed before.

Quasi-realistic climate models are useful tools to describe our knowledge about the climate processes and their interactions in a comprehensive manner. are a reduction of the real complexity of the climate system. They may be used to construct knowledge, but do not by themselves represent more knowledge than was inserted into. are the only tool to conduct experiments with the environmental system. are a valuable tool to reconstruct past climate variations and to envisage possible future climate developments. cannot be “verified”, i.e., shown to respond “correctly” to anomalous forcing if this has not been observed before.