When it comes to climate change, Portlanders might wonder how they can help. Sure, you can personally cut your carbon emissions and lobby your lawmakers to take action, but what else?
May we suggest doing a little citizen science by way of donating idle computing time on your home computer to help run climate models that generate simulations of regional climates?
You can do so as part of an ongoing research project by the Oregon Climate Change Research Institute at Oregon State University in connection with a volunteer computing and climate-modeling project at climateprediction.net.
Climateprediction.net has a project running called Weather@Home that’s focused on regional weather. It and OCCRI aim to fill a void in current climate modeling in which fine details about regional climates and a high number of simulations have been lacking.
The fine details part relates to a concept called spatial resolution, in which you’re trying to tease out and distinguish small variables and phenomena from a whole lot of data and noise. In climate modeling, this can mean creating virtual weather stations that measure smaller, more localized areas. It also may mean you have more virtual stations running per area in your models. All of this can lead to more precise modeling results.
The number of simulations, meanwhile, is important because it’s akin to sampling, and if you’ve had any statistics, you know that the more you sample, the bigger your sample size will be. And The bigger your sample size, the fewer errors you have in the data you’re analyzing, so your results are more accurate or representative.
By improving spatial resolution and increasing the number of simulations, climate scientists would be able to boost precision and accuracy in the models they run of regional climates, contributing to better model data, better predictions for scientific testing and a better overall understanding of how climate works down to the local level. This can then be used to better assess the human-induced climate disruption we’re seeing now and hone in on the most effective possible interventions.
However, to do all that — have more virtual measuring stations in your models and run a lot of simulations — you need far more computing power than individual, or even consortium, research labs typically have access to on a daily basis. And this is where you come in by donating idle computing time.
To contribute some of yours, go to climateprediction.net’s “Getting started” page.
To read about the OCCRI-led research so far, check out a paper published in the February 2016 issue of the Bulletin of the American Meteorological Society.