Climate Change 2001:
Working Group I: The Scientific Basis
Other reports in this collection

10.4 GCMs with Variable and Increased Horizontal Resolution

This section deals with the relatively new idea of deriving regional climate information from AGCMs with variable and increased horizontal resolution. Although the basic methodology is suggested in the work of Bengtsson et al. (1995), where a high resolution GCM was used to simulate changes in tropical cyclones in a warmer climate, it is only in the last few years that such models have been used more widely to predict regional aspects of climate change. Even so, only a limited number of experiments have been conducted to date (see Table 10.1) and hence what follows is not a definitive evaluation of the technique but an initial exploration of its potential.

Table 10.1: Enhanced and variable resolution GCM control and anomaly simulations. Resolution is given as either the spectral truncation or grid-point spacing depending on the model's formulation (and with a range for variable resolution models). The equivalent grid-point resolution of spectral truncation T42 is 2.8°32.8° (scaling linearly).
Institution Model Horizontal Resolution Control Forcing Anomaly Forcing Region of interest
MPI ECHAM3/4 T106 Obs ECHAM/OPYC Euro/Global
UKMO HadAM2b 0.83°x1.25° Obs   Global
UKMO HadAM3a 0.83°x1.25° Obs HadCM3 Euro/Global
MRI JMA T106 Obs MRI/GFDL/+2ºC Tropics
CNRM ARPEGE T213 T21, T106 Obs/HadCM2 HadCM2 Euro/Global
LGGE LMDZ 100 to 700km Obs CLIMAP Polar regions

10.4.1 Simulations of Current Climate

Analysis of current climate simulations has considered both deviations from the observed climate and effects of changes in resolution on the model's climatology. Most studies have considered just the mean climate and some measures of variability, either globally or for a particular region of interest. The only extreme behaviour studied in any detail was the simulation of tropical cyclones. Even for mean climate, no comprehensive assessment of the surface climatology of variable or high resolution models has been attempted. Europe has been the most common area of study to date, although southern Asia and the polar regions have also received attention Mean climate

Figure 10.8: Mean sea level pressure for DJF in: (a) HadAM3a at high resolution (100 km), (b) ECMWF reanalysis (ERA), (c) HadAM3a high resolution minus ERA, (d) HadAM3a at standard resolution (300 km) minus ERA. Adapted from Stratton (1999b).

The mean circulation is generally well simulated by AGCMs, though relatively large regional-scale biases can still be present. Many features of the large-scale climate of AGCMs are retained at higher resolution (Dèquè and Piedelievre, 1995; Stendel and Roeckner, 1998; May, 1999; Stratton, 1999a). A common change is a poleward shift of the extra-tropical storm track regions. It has been suggested that this is linked to a general deepening of cyclones, noted as a common feature in high-resolution atmospheric models (Machenhauer et al., 1996; Stratton 1999a). More intense activity is also seen at higher resolution in the tropics. For example, a stronger Hadley circulation was observed in ECHAM4 and HadAM3a that worsened agreement with observations (Stendel and Roeckner, 1998; Stratton, 1999b).

The repositioning of the storm tracks generally improves the simulations in the Northern Hemisphere, as it reduces a positive polar surface pressure bias which is present in the models at standard resolution. In the case of HadAM3a, this leads to substantial improvements in Northern Hemisphere low level flow in winter (Figure 10.8). In the Southern Hemisphere, the impact on the circumpolar flow is not consistently positive across models (Figure 10.8; Krinner et al., 1997). In ECHAM4 and HadAM3a, increased resolution has little impact on the negative surface pressure bias over the tropics but improves the low-level South Asian monsoon flow (Lal et al., 1997; Stratton, 1999b).

The existence of these common responses to increased resolution suggests that they result from improved representation of the resolved variables. In contrast, an increase in the intensity of sub-tropical anticyclones observed in ECHAM4 results from a tropospheric warming promoted by excessive cirrus clouds attributed to a scale-dependent response in the relevant parametrization (Stendel and Roeckner, 1998).

The aim of increasing resolution in AGCMs is generally to improve the simulation of surface climatology compared to coarser resolution models (Cubash et al., 1995). Early experience shows a much more mixed response. ECHAM3 at T42 improved the seasonal cycle of surface temperature in seven regions, compared to the driving AOGCM, but overall surface temperature was too high (by 2 to 5ºC). Increasing the resolution to T106 did not improve winter temperatures and, in summer, the spatial patterns were better but the regional biases worse (Cubasch et al., 1996). For precipitation, spatial patterns were improved in summer but degraded in winter. The summer warming was due to excessive insolation from reduced cloud cover and overly transparent clear skies (Wild et al., 1995). Improved physics in ECHAM4 reduced some of the radiation errors but the precipitation and temperature biases remained (Wild et al, 1996; Stendel and Roeckner, 1998). In simulations of European climate with ARPEGE (Dèquè and Piedelievre, 1995) and HadAM2b/3a (Jones, 1999; Stratton, 1999a), improved flow at higher resolution generally led to better surface temperatures and precipitation. However, over south-eastern Europe, precipitation biases increased in both models, as did the warm temperature bias in HadAM3a.

The increased summer temperatures in Europe in HadAM3a were caused by reduced cloud cover at higher resolution (Jones 1999) and warming and drying, in summer, was seen over all extra-tropical continents (Stratton, 1999b). This clearly demonstrates a potential drawback of increasing the resolution of a model without comprehensively retuning the physics. Krinner et al. (1997) showed that, to obtain a reasonable simulation of the surface climatology of the Antarctic with the LMD variable resolution AGCM, many modifications to the model physics were required. The model was then able to simulate surface temperatures to within 2 to 4ºC of observations and to provide a good simulation of the ice mass balance (snow accumulation), with both aspects being better than at standard resolution.

Other reports in this collection