CICERO - Center for International Climate Research
NO
Meny
TOPICS
Energy Food and forests Climate finance Local Solutions China ​International Climate Policy The Arctic
do you want to help run regional climate models to find out how climate change affects our weather? all you need is a computer and a few extra minutes.

do you want to help run regional climate models to find out how climate change affects our weather? all you need is a computer and a few extra minutes.

A little contribution to save the world

Climate news - News and opinions about climate science

Published 27.07.2017

Do you sometimes feel like you want to save the world? Actually, me, you, and everyone can. It is called citizen science.

A few years ago, I attended a climate science summer school for young researchers in Switzerland. The last minutes of the school had a lasting impact on me, maybe even more than all the interesting presentations and discussions of the whole week.

A few participants wanted to perform a very informal poll. They asked us this simple question: “Did you become a climate scientist because you are interested in the science or because you want to save the world?” As you can imagine, everyone started laughing, because really, this sounded very silly when spoken out loud.

How can one individual have the pretention to “save the world”?

I laughed too, but then, after a few seconds of reflection, I found myself raising my arm in the “save the world” category. Of course, I find climate science interesting, but from early on, I think what I secretly wished to do is to “save the world”, as ridiculous and impossible as it sounds. And I was not the only one, quite a few other participants raised their hand too, although the majority still voted for the “because it is interesting” category.

Even though I am completely aware that my personal contribution to “saving the world” will be marginal at best, it was nice to know that other scientists felt the same. I am sure that such grand ambitions are widespread among scientists of all fields, but also among non-scientists, and my experience seems to prove it.

Distributed computing

A few months after this summer school, I moved to the UK to work at the University of Oxford. The research group I had the chance to join, analyzes model simulations of the Earth’s climate performed on the personal computers of volunteers. This fascinating project, called climateprediction.net weather@home, relies on the fact that the resources of personal computers are underused most of the time, for example when you surf the internet.

Let me explain: We scientists prepare climate model simulations, and after volunteers (citizens) have installed a program called BOINC on a home computer, these idle resources are used to perform climate model simulations in the background, fully automatically. Together, these home computers make for a "citizen supercomputer", if you will. This technology, used for several applications, is called distributed computing.

Currently there are over 23’000 participants in the weather@home project, from 156 different countries, including Norway. The little program, BOINC, required to receive, calculate and send us back the climate simulations is very easy and safe to install. Once this is done, whenever your laptop is on, it works on a unique simulation, using only the capacity of the computer that you do not use. As a bonus, you have the possibility to see your current simulation as your screensaver.

Citizen science

While in Oxford, I attended a workshop about citizen science – people from the public that help scientists with tasks that we would else not be able to complete. Weather@home is only one of many citizen science projects.

I was fascinated by all the different projects, from classifying photos of animals in Tanzania or galaxies in the universe, to simply providing your laptop to run simulations or search for signs of extra-terrestrial intelligence. But what struck me the most were the testimonials given by the participants: they were all from different backgrounds and most of them, for one reason or another, did not become scientists. And yet they were still very interested in the science. Like me, these people had this inner drive to “save the world”.

The awesome power of citizen science, enthusiastically explained by the SciShow.

Hearing from these participants was very humbling, because unlike me, who is paid to do research, they provided these much-valued contributions to science on their free time. However, they all commented how enriching it had been to get involved, and that they felt fulfilled and useful as a result. All these projects are communal in that they have forums where participants can ask questions or report issues to the scientists, and more fun, simply discuss the interesting features they have discovered.

When I started working with weather@home, I found it really exciting to set up simulations, see them be distributed to the participants, and then returned after a few days. People would also ask us about more details on the forum, for example to comment on an article about climate change that they read in the press. As a scientist, you can feel a bit lonely with your research project on a day-to-day basis, but when I was working with weather@home, I felt part of a community and was eager to produce results to show the participants what they had contributed to.

Extreme weather events close to home 

So, why can’t the scientists perform these simulations themselves?

There are several types of climate model simulations that help us answer different questions. The most common type is probably the very long simulations of the Earth’s system from the past into the future. For these, scientists use supercomputers as they are the best tool to handle these calculations.

However, with the weather@home project, the goal is to study extreme weather events closer to home like heatwaves and floods. For this purpose, rather than a few century-long simulations, we need a very large number of short simulations because extreme events rarely occur.

How old weather becomes new science with the help of citizens. Source: NOAAPMEL.

For example, if a flood on the coast of Norway has a 1% chance of happening every year, you would see this event every 100 simulation on average. But then if you want to know whether the chance of experiencing that type of flood changes in a warmer climate, you need thousands of the simulations in the present and the future climate in order to be able to draw conclusions that are statistically robust.

Why we need simulations in the thousands

Quantifying these things is not trivial but everyone can understand this concept with the example of a loaded dice. Imagine you are throwing dices with a friend. She throws it 10 times and gets five times a 6, while you get none at all with your dice. Does it mean that her dice is loaded? Not necessarily, because if you throw it 10 more times, she might get just no 6, but you 5. Instinctively, you would want to throw the dices many more times before drawing a conclusion.

We are facing the same problem with understanding how the current weather (and in particular extreme events) have changed compared to that in the past, and how it will change in the future due to increasing CO2 concentrations in the atmosphere. Researchers only feel confident in communicating statements about such changes if we have performed a very large number (often in the order of 10’000) of simulations.

This is why performing many short simulations is very important to study extreme events. Unfortunately, the supercomputers available are not the best tool to handle this task. However, a very large number of laptops is, and this is where the idea of asking volunteers to help us running the model simulations came from.

A success story from Oxford

The first winter I lived in Oxford, it rained a lot and parts of the city and surrounding areas got flooded. It was my first experience of flooding, and since we lived very close to the Thames river, it was a very stressful situation.

Every morning we would look outside the window and see how high the water level had risen. Luckily, our street did not flood, but many of our neighbors had to be evacuated. By the end of February 2014, we knew that it had never rained as much in Oxford since scientists started to record rainfall in 1767.

At the same time, people were asking whether this extreme event was caused by climate change. While this is not a question that can be answered by a yes or no, what we can say, thanks to the weather@home project, is by how much human influence has changed the chance of having that much rain during winter. This is similar to doctors being unable to say that smoking will cause lung cancer for everyone, but being able to quantify that smoking will increase the risk of getting cancer by a certain percentage.

Still, we wanted to answer the question the public was asking us climate scientists instead of letting speculations become the answer people would remember. Therefore, at the beginning of March 2014, we prepared hundreds of thousands of simulations and asked our volunteers to calculate them. Within a few weeks we had around 100’000 simulations and we were able to say that indeed, in this specific case, human influence had increased the risk of having so much rain during the winter months by around 40%.

Nathalie Schaller explaining the UK flooding in 2014, and citizen science. Source:  Environmental Change Institute. 

Obtaining this result so quickly would not have been possible without the volunteers. If we had had to use a supercomputer, it would have probably taken years before being able to make such a statement, and by that time, people – except the ones directly affected – would have forgotten about this extreme event.

We believe that it is important to show the public that climate change is happening here and now, and it is much easier to do this with their attention already turned towards the current weather, rather than years after.

Become part of the "citizen supercomputer"!

If you too have this drive to “save the world” or you are just “interested in climate science”, I cannot encourage you warmly enough to sign up for a climate-related citizen science project like weather@home. We won’t save the world with citizen science, but I believe it enables us to make great progress in better understanding the world and climate we live in.

To participate, you don’t need to have previous knowledge about the climate system. All you need to make a great contribution to science is your computer and a few minutes of your free time to install BOINC. And, if you want to, you will have the possibility to learn more about the climate and engage with other participants and scientists in the forums.

The environmental issues we are currently facing can appear so much larger than us, and it is easy to think that personal actions are too small to make an impact. This is however wrong, and when you join a citizen science project, you realize that many people around the world share similar interests and that together, we do produce results that will hopefully lead humanity to actions that will save the only planet we have.

 

Apart from weather@home, there are several other climate/weather citizen science projects you can join, depending on your particular interests:

Community Collaborative Rain, Hail and Snow Network: Community-based network of volunteers working to measure and map precipitation (rain, hail and snow).

USA National Phenology Network: Focused on collecting standardised ground observations of phenology by researchers, students and volunteers.

Old weather: Inputs historical maritime weather logs that can be used as observed weather data for climate models.

Hurricane satellite data: The HURSAT project provides Tropical Cyclone-"centric" satellite data.