Benchmarking the war against global warming

We analyzed the HadCRUT3 reconstruction of the instrumental global temperature record for 1850 through 2008 to decompose thirty-year temperature trends into signal and noise components. The signal represents multidecadal trends and the noise represents annual variability about those trends. Historical estimates of temperature variability (e.g., noise) are used with seven temperature projections to simulate global warming time series. These trends include the 1979 through 2008 trend, four trends taken from the Intergovernmental Panel on Climate Change (IPCC) Fourth Assessment Report (AR4) simulations (A1B, A2, B2, and the constant composition commitment [CCC] scenarios) and trends representing approaches to 1.5°C and 2°C warming by 2100. Each series is simulated 1,000 times. The results are compared, statistically, to the current warming rate of 0.016°C year^-1. We calculate the time until those trends become statistically different from the trend observed over the most recent thirty-year period (1979-2008). The results indicate that it will probably be decades before distinct changes from the current warming rate become apparent. For the A1B scenario, only 25 percent of the simulations indicate difference by 2040. For the CCC, A2, 1.5°C, and 2°C scenarios, the 25 percent level is reached in about 2030, 2040, 2065, and 2075, respectively. Only about 10 percent of the B1 simulations indicate a difference before 2100. These results indicate that we should expect decades to pass before impacts of the war against global warming become apparent.