CLIMATE CHANGE MODELS: YOU WILL NEVER GUESS WHICH ONE IS RIGHT!

In 2021, the IPCC stated in its climate change report that crossing the global warming tipping point of 1.5℃ will lead to some irreversible effects. Anthropogenic release of greenhouse gases are estimated to have caused, on average, 1.2℃ of global warming since pre-industrial times (1850-1900), and are projected to rise based on global models. Global climate models encompass temperature readings from land and ocean which cover 29% and 71% of the Earth’s surface, respectively.

The question has to be asked, can global models and tipping points be used to predict climate at the regional/local level?

Regional climate ensembles are employed to predict climate at the smaller scale and can be made up of over twenty models. Great variation of over 5℃ between models becomes apparent when compared to the historical measurements.

In this study, we propose to use historical temperature trends to select the most accurate and precise model for climate predictions in the Niagara Region. First, historical data ranging from 1900 to 2021 was compared with trends of 19 regional climate models. These were extracted from the North American Coordinated Regional Climate Downscaling Experiment program data portal. These models are best suited for analysis of the Niagara Region because each include a lake component, simulating the effects of both Lake Ontario and Lake Erie on the local climate. The models were compared to the Region’s observed temperature trend over a timespan of 72 years, from 1950 to 2021. Each model was statistically assessed using a skill score, a matching score and the M3 score, where a M3 score of 1 means a perfect match to the historical trend. The model which bests fits Niagara’s observations is RCP8.5 model 8, which corresponds to business as usual with a high-emissions scenario. It has the highest M3 score of 0.96 and predicts an increase in mean annual temperature of 1.93℃ from 1900 to 2030. We recommend that global models or regional climate ensembles should not be used at the local level for future climate predictions. Only the regional climate model that best-matched past historical trends should be used to predict future climate. In this case, only RCP8.5 model 8 should be used to accurately and precisely predict climate change in the Niagara Region, and thus enable better preparedness of infrastructure and agricultural planning.