Applications of CPRCMs

The applications of CPRCMs are limited by their computational cost. It is too costly to run the model for longer than 30 continuous years, which means they are not suitable for simulating many long term climate phenomena. As such, experiments trying to predict the effect of certain forcings on future climates will run the models for “time-slices” (typically 10 years). These slices will be started during the beginning and end of the studied period to observe how the climate has changed in the meantime. Between the runs, the conditions of the models are changed using data intended to reflect the conditions of the future climate. These changes often concern values such as GHG concentrations. Often, CPRCMs are couched within two boundary layers. The innermost boundary separates the high-resolution CPRCM from an RCM-esque model, which operates at a resolution between that of the CPRCM and the outermost GCM. CPRCMs can be extremely dependent on the optimizations done to improve the performance of this boundary layer. In the long term this can make their predictions unstable or relatively less accurate than similar predictions made with simpler RCMs. Thusly, CPRCMs are in many cases not viable tools for performing hindcasting research.
CPRCMs are useful often when researchers are attempting to model phenomena that require a high spatial resolution, not achievable on state-of-the-art GCMs. For example, CPRCMs have majorly increased and refined the ability of researchers to predict monsoon patterns in the Himalayan regions of Nepal. Karki et al.’s 2017 paper showed a direct (though not linear) relationship between resolution and accuracy of monsoon predictions. It is easy to understand how this technology could benefit people the world over. Arming humanity with the ability to accurately predict the weather during an entire season of the year would allow us to adapt to increasingly irregular weather patterns wrought on by climate change, helping farmers decide when they should plant their crops. Similarly, work done by Gentry and Lackmann in 2010 showed that CPRCMs are better at simulating cyclones themselves. The low pressure center that exist at the center of the cyclonic storms are too small to be depicted by GCMs and RCMs, so only CPRCMs are able to generate predictions that show a low pressure center in the middle of these storms.

Looking Ahead

It is always dangerous and difficult to predict what the future has in store for technology. However, history tells us that today’s CPRCMs will most likely be tomorrow’s RCMs, and tomorrow’s GCMs will operate at the same resolution that we’re experimenting with now. The fact that researchers are now able to generate better results in small areas tells us that the path forward is clear. Once computers get good enough, the boundaries can go away and the first convection permitting global circulation model will be created.