Climate trends data

Historical climate information in the graph has been estimated based on surrounded stations using interpolation techniques implemented by the Queensland Climate Change Centre of Excellence. These estimates will be most accurate after the mid-1950s. Prior to this time many records are unavailable in digital form and station density was also lower. There are relatively few records from the 1890s and the early 1900s, so estimates are least reliable for these times.

The graph also shows the historical trend in climate calculated over sequential 30-year periods. For example, the mean at 1995 is for the period 1980 to 2010. This trend line has been projected forward since 1995 based on information from 23 global climate models prepared by CSIRO in 2007. The middle dotted line represents the average projection based on all 23 global climate models. The outer dotted lines indicate the range in projections for 80 per cent of models. This means that projections from 10 percent of models lie above the upper dotted line and projections from 10 percent of models lie below the lower dotted line.

It is important to recognise that these models were not run specifically for each location but rather on large-scale global grids. The projections for individual locations are based on an interpolation of the coarse-scale grids over Australia prepared by CSIRO in 2007 at a scale of approximately 100 km2. The base information is the same as that used by the Queensland Climate Change Centre of Excellence to prepare 13 Regional Climate Change Fact Sheets for Queensland as part of the Queensland Government's ClimateQ initiative.

The climate projections are based on a high (A1FI) emissions scenario which global greenhouse gas emissions tend to be currently tracking. However it is important to note that the 23 global climate models on which this analysis is based were actually not run for the high (A1FI) emissions scenario. The projections for the high emissions scenario are extrapolated from model results based on lower emissions scenarios using a technique developed by CSIRO. The results are least reliable further into the future as the impact of emissions scenarios become more important and the extrapolations become greater.