A Basic Introduction to Climate Modeling and Its Uncertainties

Abstract

Climate models are nowadays the most important tools for estimating future climate changes. The data produced by climate models are widely used not only by climate researchers, but also by a growing number of researchers from other disciplines, such as ecologists, coastal engineers, social scientist, etc. The use of climate models in the assessment of climate change impacts requires, however, certain care and the knowledge of some caveats. This chapter attempts to introduce some basic concepts relevant for climate modelling and for the interpretation of climate simulations. It is aimed at those with only a basic knowledge of climate physics and is intended to provide a rough guide to this complex field, mainly in the form of examples from current results with climate models. Although the Baltic Sea does not play a central role in this chapter, the chosen examples bear some relevance to potential problems which climate simulations encounter in this geographical area.

Appendix

4.1.1 A Brief Description of the Climate Models Mentioned in this Chapter

ECHO-G is a global coupled atmosphere-ocean model. It is composed of the atmospheric model ECHAM4 and the ocean model HOPE. ECHAM4 is a spectral model with a horizontal resolution of about 3.75° × 3.75° and 19 levels in the vertical direction. HOPE is a finite-difference model with a horizontal resolution of about 2.8° × 2.8° and 20 levels in the vertical direction. Due to the relatively coarse horizontal resolution, the fluxes of energy and momentum between both models contain errors that would cause a small but persistent long-term drift of the coupled model. This is quite detrimental for long simulations and to avoid this drift, a flux adjustment is applied to couple both models.

ECHAM4-OPYC is also a global coupled climate model as ECHO-G but with another ocean sub-model. In the place of HOPE, another ocean model OPYC is coupled to ECHAM4. A special characteristic of OPYC is that its levels in the vertical are defined at surfaces of constant density, and not of constant depth.

ECHAM5-OM is a global climate model that consist of the atmospheric model ECHAM5 (a newer version of ECHAM4) and the ocean model OM. The horizontal resolution of the model ECHAM5 is typically 1.9° × 1.9°. It does not require flux adjustment as ECHO-G does. In the set-up used for the IPCC simulations, this model includes a vegetation model and a carbon cycle model.

CCSM3 is a global atmosphere-ocean climate model developed at the National Centre for Atmospheric Research in Boulder (USA). The atmospheric sub-model is a spectral model with a resolution, for the IPCC simulations, of about 2.8° × 2.8°, and 20 levels in the vertical direction. It does not require a flux adjustment.

MM5 is a regional atmospheric climate model. It needs to be nested in another global model, i.e. it requires values of the temperature, humidity and wind every 6 h at the boundaries of the simulation domain by a global climate model, or by meteorological re-analysis. Its resolution may be chosen by the user within a certain permissible range. Typically, resolutions of about 50 × 50 km are used, but simulations with finer resolutions, for instance half these values, are also performed. An interesting property of this model is that it allows for the so called double nesting set-up. In this set-up, a smaller model domain with a higher resolution is nested within a larger domain with a somewhat coarser resolution, which in turn is nested in a global climate model. In the example presented in this chapter for simulations of the climate of the Iberian Peninsula, the high-resolution inner domain covered the Iberian Peninsula with a resolution of 45 × 45 km and this domain was embedded in the outer model domain, covering the North Atlantic-West European sector with a resolution of 120 × 120 km. This model allows to choose among different parametrizations of sub-grid processes involved in cloud formation, convection, etc., so that different results may be obtained depending on the values of these parameters. In general, these parameters are selected according to the climatic region that is simulated.

WRF is essentially a newer version of the regional model MM5.

RCAO is a regional coupled atmosphere-ocean model developed at the Swedish Rossby Centre. It comprises a regional atmospheric model RCA and an ocean regional model RCO that includes a sea-ice model. It has been mainly applied for the Baltic Sea basin. The atmosphere resolution is similar to those in other regional atmospheric models such as MM5. The ocean resolution may typically be of the order of 6 nautical miles, with 40 levels in the vertical direction.

BALTIMOS is, like RCAO, a regional coupled atmosphere-ocean model, based on the regional climate model REMO. It has been developed at the Max-Planck-Institute for Meteorology and one of its main applications has been to simulate North Sea water intrusions into the Baltic Sea.