May 2th, 2016

Climate models are one means of studying climate dynamics at global and regional scales. A variety of these models exist and each uses different approaches to solve the mathematical equations that describe the physical phenomena and processes driving our planet’s climate. They allow scientists to ask important questions about how the climate system works, and how it is expected to change in the future under different greenhouse gas emissions scenarios. In the most recent assessment report published by the Intergovernmental Panel on Climate Change (IPCC) in 2013-2014, a set of 30-40 climate models were used to evaluate and understand the impacts of climate change in the past, present and future. These experiments are part of the foundation of information used to guide the Conference of the Parties (COP21) international climate agreement of December 2015 in Paris.

Climate models are being continuously enhanced to better represent a wider range of physical phenomena. Thus, to make sure these climate models are reliable, each model’s performance must be evaluated against its ability to reproduce present and past climatic conditions. For instance, model simulations should correctly reproduce the Earth’s energy budget. Increasing atmospheric concentrations of greenhouse gases, like carbon dioxide, tilts the Earth’s energy balance by trapping additional outgoing radiation emitted from the surface of the planet. The increased energy available in the Earth system is distributed among the oceans, continents, ice sheets and atmosphere and is ultimately the cause of current climate change.

To examine the ability of state-of-the-art climate models to properly partition energy among the different parts of the Earth system during climate change, M.Sc. students Francisco José Cuesta-Valero and Almudena García-García (both in StFX’s NSERC CREATE program in Climate Sciences), Dr. Hugo Beltrami (St. Francis Xavier University, Canada Research Chair in Climate Dynamics) and Dr. Jason Smerdon (Lamont-Doherty Earth Observatory of Columbia University) used 32 20th-century climate simulations from the latest IPCC ensemble to calculate the amount of energy partitioned underground in each simulation between 1950 to 2000. These simulated results were compared with estimates of global continental heat gain derived from hundreds of terrestrial borehole temperature profiles measured throughout the world. The team found that the IPCC simulations consistently underestimate the energy storage in the subsurface of continental landmasses for the latter half of the 20th century with respect to estimates from geothermal data. The results of the study were published this week in Geophysical Research Letters [1].

The inability of the climate model simulations to reproduce the continental energy storage is due to the shallow depth of the soil models used in the simulations. The authors suggest that changes in surface air temperature propagate into the ground with time, but in shallow soil models, air temperature signals cannot reach the depth necessary to permanently influence the ground’s thermal state. This implies that models do not store temperature changes from previous years, resulting in an energy storage deficit in the continental subsurface.

These findings suggest that climate simulations should be performed using deeper soil models, better suited to represent important soil processes that depend on the heat content near the surface of the continents, such as those determining the decay and stability of permafrost and soil carbon. These important processes may become positive feedbacks to the climate system, as the instability of permafrost and decay of old soil carbon pools could make soils stronger sources of greenhouse gas emissions, impacting the evolution of 21st-century climate. The group’s research was supported by the Natural Sciences and Engineering Research Council of Canada (NSERC) CREATE program, as well as from the Canada Research Chairs program.