Uncertainties
Methodology uncertainty
Many unknowns and uncertainties in the climate system are not reflected in the IMAGE scenarios. Some of the major uncertainties in the causal chain are the climate sensitivity and regional climate-change patterns. The direct effects of a changed climate are changes in carbon uptake by the biosphere and oceans and in the distribution and productivity of crops, as well as shifts in ecosystems. Indirectly, many other processes are influenced, which can lead to the concentrations of greenhouse gases in the atmosphere being built up differently and to different land-use patterns. IMAGE simulates the consequences of these changes in an integrated fashion, accounting for interactions and feedbacks. The outcome is thus not necessarily a linear function of climate sensitivity.
These climate uncertainties were addressed by providing additional simulations to illustrate the uncertainty in the climate sensitivity and in the regional climate-change patterns.
Climate sensitivity
Climate sensitivity refers to long-term (equilibrium) change in global mean surface temperature following a doubling of the atmospheric concentration in CO2 equivalents. According to IPCC, this climate sensitivity is between 1.5oC and 4.5oC. In earlier versions of IMAGE, the climate sensitivity generated by the climate model was 2.4oC. Due to the rigid structure of these earlier versions, we were unable to change this and assess the consequences of such a change.
In IMAGE 2.2 a simpler climate model MAGICC (see Upwelling-Diffusion Climate Model) is incorporated, allowing to define the climate sensitivity. The default value for IMAGE runs is 2.5, which is the median value of the IPCC range (median differs from mean because the range is logarithmic).
To test the uncertainty related to the climate sensitivity, runs with respectively a low (1.5oC) and high (4.5oC) climate sensitivity were created. A pattern-scaling procedure is used to obtain regional and seasonal climate-change patterns using the calculated increase in global mean temperature.
Runs with changed climate sensitivity are provided for the A1F (A1F low, A1F high) and B1 (B1 low, B1 high) scenarios on the main disc (IMAGE team 2001a). These scenarios span the full range of the SRES emission scenarios and therefore adequately illustrate the uncertainty of different climate sensitivities.
Regional climate-change patterns
Climate-change patterns are not simulated explicitly in IMAGE. The global mean temperature increase, as calculated by IMAGE, is linked with the climate patterns generated by a general circulation model (GCM) for the atmosphere and oceans. This linking takes place using the standardized IPCC pattern-scaling approach (Carter et al., 1994) and additional pattern-scaling for the climate response to sulphate aerosols forcing (Schlesinger et al., 2000; see Geographical Pattern Scaling, GPS). GCMs are currently the best tools available for simulating the physical processes that determine global climate dynamics and regional climate patterns.
GCMs simulate climate over a continuous global grid with a spatial resolution of a few hundred kilometres and a temporal resolution of less than an hour.
Most GCMs agree on the global patterns of climate change:
- temperature increases above land are faster than above the oceans
- high latitudes warm up more sharply than low latitudes
- winter warms up more sharply than summers
- total precipitation increases with increasing temperature
- maritime regions generally get wetter
- continental regions could get dryer.
Regionally, however, there are large differences between the different GCMs, especially in precipitation-change patterns.
IMAGE 2.2 runs with five different climate-change patterns are provided on the supplementary disc (IMAGE team 2001b, RIVM CD-ROM publication 481508019) for the A1F, B1 and A2 scenarios. The aim of this material is to illustrate the uncertainties in SRES climate-change scenarios resulting from these differences in GCMs. The first two scenarios span the full range of the SRES emission scenarios, the latter being based on a highly different narrative with different demographic and socio-economic assumptions. The three scenarios therefore adequately illustrate the uncertainty of different climate patterns. Differences in the runs for each scenario indicate some of the uncertainty caused by regional variation in climate-change patterns (not the global mean).
The scenarios for five different GCM runs from the IPCC data centre were implemented, which comprised:
- ECHAM4 of the Deutsches Klimarechenzentrum DKRZ in Germany
- CGCM1 of the Canadian Centre for Climate Modelling and Analysis in Canada
- GFDL-LR15-a of the Geophysical Fluid Dynamics Laboratory in the USA
- HADCM2 of the Hadley Centre for Climate Prediction and Research in the UK
- CSIRO-MK2 of Commonwealth Scientific and Industrial Research Organisation in Australia
Data sets uncertainty
The input data to the UDCM model is atmospheric concentrations of greenhouse gases and emissions of SO2, which by itself is calculated on the basis of the other IMAGE 2.2. Models and bare all uncertainties related to those models (see more in methodology uncertainly).
Rationale uncertainty
The observed increase in average air temperature, particularly during the recent decades, is one of the clearest signals of global climate change
The indicator shows trends in temperature data over time. Temperature is directly linked to the question of climate change and is a state variable that changes in response to the pressures of global warming.
There is growing evidence that anthropogenic emissions of greenhouse gases are (mostly) responsible for the recently observed fast increases in average temperature. Natural factors like volcanoes and sun activity could explain to a large extent the temperature variability up to mid of the 20th century, but they can explain only a small part of the recent warming.
Document Actions
Share with others