Can hydrological forecasting contribute to Science & Art projects?

Contributed by Maria-Helena Ramos and Louise Arnal.

In 1991, L. D. James starts his book chapter stating that “[i]n content, hydrology is a science; in practice it is an art”. The discussion that follows reflects on the links between science and practice at the time of writing, questioning how much “the art of hydrology [was becoming] increasingly outdated” and highlighting the “pressure for new methods” to keep on securing decision makers in water resources planning with “sound scientific information for objective decisions”.

In 2009, H. H. G. Savenije wrote an opinion article in HESS with the title “The Art of Hydrology” (you may also want to check the discussions here). The author argues that “[t]he process of modelling requires imagination, inspiration, creativity, ingenuity, experience and skill” and adds that “[t]hese are qualities that belong to the field of art.” The point raised by the author is that “[h]ydrological modelling requires art”, which lies in the “ability to reconstruct the architecture of a largely unknown system from a few observable signatures that characterise its behaviour.”

More recently, in April 2017, K. Beven wrote a blog post on “The Science and Art of Hydrology”, where he notes that the “dynamic features of the landscape (…) has long attracted artists”. He also discusses the fact that it requires good skills to “represent the dynamics of flowing water in two-dimensional images”, and presents beautiful photos of water taken through his eyes of scientist working on dynamic hydrology and uncertainty.

Is there also art in hydrological ensemble prediction?

Certainly, many of us are creative and skilful to produce novel science concepts and beautiful graphs from field and modelling experiences (and even nice Hepex logos!). Our question here refers to how we can open a dialogue between the science of forecasting and art.

Science and Art

Science and art is not new (look at Leonardo da Vinci, for example, artist and scientist already in the 1400s). In meteorology, scenic approaches to meteorological phenomena and science-based interpretations of paintings depicting different types of weather have been used for a few decades now.

In the early-1900s, L.C.W. Bonacina wrote a series of articles to advocate the need to combine the sensible impression of landscapes with the scientific study of the weather and climate (as well as other geographical sciences), which he termed the “scenic approach to meteorology”. His argument was that this close relationship between aesthetic and scientific problems is vital in reaching a unifying vision of our landscapes.

L.C.W. Bonacina’s articles were largely triggered by his growing impression that the description of meteorological events and phenomena could not entirely be done by numerical measurement and verbal categorisation, but needed to be completed by scenic or pictorial qualification. For example, the contemplation of beautiful scenes can raise scientific questions which might have not otherwise been thought of. L.C.W. Bonacina also argued that scientific knowledge can enhance aesthetic appreciation. He wrote:

“it is difficult, for example, to believe that a person could completely come to terms with the majestic beauty of a thunder cloud, who had not at least a partial understanding of the mode in which water in mid-air is able to pile itself up into mountainous formations with peaks, domes, cliffs and gulleys, and every uncanny simulation of rock-like solidity.” – L.C.W. Bonacina (1941)

“The scientific knowledge of the meteorologist should help rather than hamper artistic enjoyment.” L.C.W. Bonacina pleaded for the physical meteorologist and the artistic meteorologist to work alongside and constantly throw questions at each other, which neither of them would have thought about in isolation. This should help them realise that they are dealing in their respective fields with counter parts of one reality, a vital step towards the unification of outlook.

Would you try a scenic approach to hydrological forecasting?

Inspired by L.C.W. Bonacina’s (1941) depiction of landscape-weather scenes (you should check the paper!), we decided to try a “scenic approach to hydrological forecasting”. Here is what we came up with:

“Just after debugging, input and initial condition uncertainty combined to give a reliable confidence interval in light-blue shadows to an unperturbed forecast delivered by a single-value dark blue dotted line prediction in which the 2-year critical threshold and the 5-year close by, would at time intervals peer” – Maria-Helena Ramos

“The plot was complex and hazy with elements of doubtful import, producing a fine quality of colours, lending in turn, as not uncommonly in research papers, stature to those stately groups of lines. One line with a peculiarly graceful rising limb, standing alone, darkly silhouetted against the plot’s background, was conveying in the particular signal prevailing, an emphatic message of beauty which the recipient will never forget. There was nothing very unusual or spectacular about the signal that time of year, but just for that very reason the suggestion was all the stronger that the expression of the lines on the plot, never constant, depended on a delicate balance of conditioning factors in the atmosphere.” – Louise Arnal

Although we are very proud of our creativity, we are not really sure that we can exploit this potential in our publications, so let’s come back to our business here…

Towards two-way collaboration between artists and scientists

L.C.W. Bonacina’s writings plead the case for collaboration between artists and scientists. This is not the very one directional idea that is sometimes put forward, according to which artists should communicate scientists’ work. It goes beyond that because, as mentioned by L.C.W Bonacina, artists and scientists need each other to have a unified vision of things, by learning from each other’s tools, understandings and perspectives.

Nowadays, many museums and galleries, exhibitions and artists explore the beauties of science (data, products, old sketches, paintings or diagrams) to engage the public and increase awareness about the importance of our societies’ growing knowledge. Lise Autogena and Joshua Portway are two artists based in London, who have already created several science and art projects, exhibited internationally. “Most Blue Skies (I+II)“ is one of their science and art projects, where a computer generated installation runs continuously to find “the bluest skies” in the world – by measuring the passage of light through particulate matter in the atmosphere and calculating the exact colours of the sky at billions of places on Earth. By combining “atmospheric research, environmental monitoring and sensing technologies with the romantic history of the blue sky and its fragile optimism”, this project “addresses our changing relationship to the sky space as the subject for scientific and symbolic representation.”

An ongoing exhibition in London at the V&A museum, called “Chance and Control: Art in the Age of Computers”, celebrates 50 years of computer-generated art. The artworks displayed question the intersection between the expected and the unexpected (it may ring a bell to those involved in real-time operational probabilistic forecasting!)

Aiming at bringing together scientists and artists, the EGU Science and Art session provides a platform where artists and scientists can explore, discuss, interact and finally exhibit their co-creations (check here for an overview of last year’s EGU Science and Art session submissions). Also, since 2015, the Joint Research Centre of the European Commission has been organising festivals and summer schools to experiment with elements of science, policy making and the arts “to explore different ways of discussing scientific, ethical and policy relevant issues”. Their 2018 SciArt Summer School (3^rd edition) was on “Big Data”. Working with a lot of good quality data is probably daily routine for many people of the Hepex community. This is at least the case of my [Louise Arnal’s] PhD, which was also a motivation to participate to the summer school (see this blog post for an account of Louise’s experience at this event).

It is in fact impressive to see how much (art and science!) can be done with so much data accumulated. The ongoing exhibition “1 2 3 data” at EDF Foundation in Paris, for instance, presents the works of about forty data designers, for whom the millions and millions of data generated worldwide is the raw material for their artwork (see news in English here and much more about it here). The exhibition, very interactive, does not only show the power of combining data and creativity, but also illustrates how we are constantly acting as data providers ourselves. All in all, it is interesting to think that, no matter how much of it we have, data is only an abstraction of reality – it doesn’t tell us the full story and is just one out of many realities.

Art and Science in Hydrological Forecasting

Checking if reality matches our forecasts is a typical exercise of forecasting verification. Forecasting can be summarised as an attempt to deduce (or guess) the future from the present. We know that our best guesses are based on the data we have and science: understanding the drivers, the natural processes and the human influences; estimating the initial conditions and the model uncertainties; understanding how to communicate risk and inform decision-makers on future scenarios, etc.

Finally, with this blog post, what we would like to know is: is there “art in forecasting”? If yes, where is it?

Some might argue that it is not in the techniques, but in the applications. Some might say that it is in the creative and innovative forecasting solutions, but not in the verification of forecast accuracy. What do you think?

• Do you see art in forecasting the future?
• Can art be more explored as means of communicating forecasts, concepts behind ensemble prediction, uncertainty, etc. in the science and practice of hydrometeorological ensemble prediction?

We asked these questions to an artist and a scientist. The first question was answered by Tiziana Lanza, science writer and novelist at the National Institute of Geophysics and Volcanology (Istituto Nazionale di Geofisica e Vulcanologia) in Rome. The second question was answered by Jutta Thielen-Del Pozo, Head of the Scientific Development Unit at the JRC (see a previous Hepex interview with her here).

Do you see art in forecasting the future? by Tiziana Lanza:

“In the contemporary world, Nature cannot be seen anymore only through the eyes of Galileo Galilei who was describing it as a book open to be investigated just with the language of Mathematics. Today human beings are interacting with the natural world and transforming it day by day through technology, a by-product of science, that, in turn, creates a new reality where scientists are already creators.

Then, I do believe that in their very first approach in exploring the natural world, each with their own method, scientists and artists are in the same mood. There is no action without inspiration neither for the scientist nor for the artist. Nevertheless, the main difference is that for a scientist, inspiration is mainly the first step, since immediately after a dialogue starts with the natural world. Instead for an artist inspiration should remain and eventually become more intense during all the creative process.

So, I am convinced that Art can play an important role for the future of hydrogeological forecasting, since artists can take inspiration from the work of scientists, that, on their turn, take inspiration from the work of artists. But, most important, artists in their emotional approach to reality can contribute to re-awaken in scientists what was so important to Albert Einstein, the intuitive mind, especially today and especially when we deal with forecasting the future. This is an activity that has already widely shown how very often reality is so complex and flighty to resemble to an artist’s masterpiece, forcing scientists to continually tune their models and methods of investigation.”

Can art be more explored as a means of communicating forecasts, concepts behind ensemble prediction, uncertainty, etc? by Jutta Thielen-Del Pozo:

“When I became first involved with the JRC Art and Science project a few years ago I was sceptical that this new approach could be successfully weaved into the day to day scientific research in an impactful way. However, since then I have started to realise how powerful collaborations of artists and scientists working together can be. In particular, art has the power of providing context to a subject in an intuitive way that pure facts and figures have difficulty to convey. Facts and figures rationalise our understanding, whilst art has the power to touch on our emotions. Together art and science allows us to explore and communicate underlying complexities  – at a glance, with a touch, through sounds. Just like music puts the actions of a movie into an emotional context, art can frame scientific results in a meaningful way.

Now to answer the question: yes, even if the collaboration between artists and scientists was reduced to aspects of visualisation only it would already be an advantage for ensemble prediction and related research because it could help to extract the meaningful information for different viewers in a better way. In particular, visualising uncertainty in a meaningful way is a challenge and any improvement in communicating it better is an added value. However, in my opinion, this would be a serious shortcoming in exploiting the collaboration potential between artists and scientists. We experience that users are reluctant to act appropriately on uncertain information and observe irrational decision making in the face of extreme events.

I believe that artists could help scientists to understand where this reluctance comes from and then nudge users towards acting according to the facts and figures even if it appears to be counter intuitive. Also, artists may be in able to help hydrologists in understanding how the citizens perceive risks and their potential impacts which could lead to better risk management information and policies. Also, how we deal with the probabilities of changing spatio-temporal conditions whether it is upcoming extreme events such as flash floods or a slowly developing climatic shifts is an issue that can be explored together.

Finally, it is important to understand that working with artists does not mean that scientists stop doing science – “So, now should we all start doing art instead of science?” is a provocative question I am hearing often and which only illustrates the limitation of the one who is asking. On the contrary, scientists need to continue doing their research vigorously and perhaps even more so than in the past. By involving artists in the research from the start and throughout a project may actually allow viewing issues from different angles, to understand the concerns and fears of society in the context of the research better and thus to formulate the research questions differently from the beginning. And, of course, it may also help to better exploit the results and reach out to different audience including the public. The end result might astonish some of the initially reluctant scientists, and help bring previously dry topics to the attention of the public and get scientists out of their ivory tower.”

_________

What are your views on the topic of this post? Are there any examples of Science and Art exhibitions/projects/initiatives that would you like to share with us? Add them using the comment box below!