Snow, artificial intelligence and future visions: a researcher’s journey from the Alps to Oslo

Far from her mountains and her sea, between Norway's icy landscapes and the lines of code that simulate snow dynamics, Giulia Blandini faced a challenge that was both personal and scientific.

A PhD candidate in snow hydrology at CIMA Research Foundation and the University of Genoa, Giulia focuses her research on the use of artificial intelligence to improve the understanding and modelling of snow and hydrological processes. During a research period at the Department of Geosciences of the University of Oslo (UiO), she continued her investigation, working with distributed models, satellite data, and new machine learning algorithms.

In this interview, she shares what it means to work with artificial intelligence, to be young in science, and to imagine new ways of interpreting climate change. Hers is not a voice looking for shortcuts, but one that moves consciously through the complexity of the scientific method, sharing doubts, discoveries, and the courage to step into the unknown.

Giulia, your research period in Oslo is coming to an end. What have you been working on in recent months?

Before coming to Norway, I had developed a data assimilation technique for one-dimensional snow models applied to point-scale stations. In Oslo, I began working on a two-dimensional version, applied to entire catchments, but still designed for operational use, meaning it requires fewer computational resources. The aim is to build a distributed framework to estimate snow in a realistic way, combining station-based and satellite observations.

Even during my stay in Oslo, I continued working with data from the Italian Alps, with the goal of developing a modelling approach that is theoretically sound but applicable in concrete, complex contexts such as those found in the Mediterranean region. This allowed me to maintain scientific consistency while enriching my perspective through international experience.

I delved deeper into statistical and machine learning methods, using models like Gaussian Processes, which lie somewhere between statistics and AI-often described as infinite neural networks. These models allow spatial interpolation of variables I previously modelled only at specific points. It's a way to build something more scalable, while preserving scientific rigor and applicability.

So, not just neural networks?

Exactly. My goal is not to replace physical models, but to complement them. Traditional models are based on physical equations: we assume known relationships among variables and build on them. But when those relationships are unknown or too complex to formulate, artificial intelligence can step in. In particular, I use AI to emulate the statistical component of data assimilation-creating an emulator capable of reproducing corrections from the physical model, but with lower computational cost and faster execution.

So it's not just a matter of performance. It's about bridging physics and statistics, preserving theoretical control while enabling application in operational environments with limited resources.

You've also worked with satellite data?

Yes, and I've learned a lot from it. In Oslo I deepened my knowledge of satellite data, both for data assimilation and for machine learning. The goal is to move from point-based station data to a distributed and temporally continuous observation system. That helped me better understand the links between modelling, observation, and forecasting.

My ambition is to create a framework that can be applied across the full range of Italian and Mediterranean mountain chains. But to do that, you need a solid foundation and a vision that integrates theory, data, and application. These six months were dedicated to finding that balance.

From a personal point of view, what has this experience given you?

Much more than I expected. I chose Oslo as a personal challenge-one that went well beyond the scientific. It meant immersing myself in a place far from where I grew up, used to the brightness and openness of the Mediterranean, between Liguria and Sicily. I wanted to see how I would adapt in a landscape that, at least on the surface, didn't seem to reflect me. It was a conscious, almost symbolic choice: to push myself out of my comfort zone, into a space where even my way of doing research would change. And that's exactly what happened.

A concrete example? I had never practiced winter sports, and suddenly I found myself skiing and ice skating on a frozen lake. For me, it was symbolic. I thought: if I can stay upright on this ice, maybe I can build a model too. And I did. I learned to take risks, to trust my own abilities. To take my first steps alone, even in research.

What struck you about the way research is done in Norway?

Two things. First, the relationship with snow is different. In Italy, snow is a scarce and increasingly critical resource. We study it to manage water availability, to assess risk-we're already in an emergency context. In Norway, snow is also studied for recreational reasons, tourism and sports. That changes the approach, and made me realise how much context shapes the meaning of research.

Second, the balance between work and personal time. In Italy, we often stay glued to the screen well into the evening. In Norway, it's normal to take a break-to enjoy the sunlight, for example. It doesn't mean working less; it means slowing down in order to think better. I've learned that even empty time can be part of the creative process.

How are you feeling now that you're about to return to Italy?

I'm very satisfied with these six months. I used to think, "life is better outside Italy." But now I know I want to come back. I want to live in a place that reflects who I am-my rhythm, my character, my way of relating to others. I'm sad to leave behind the friendships and academic routine in Oslo, but I feel this is the right moment to return and bring home what I've learned-and the person I've become.

What would you say to someone considering a PhD?

That if there's a field worth investing in, it's environmental science applied to natural hazards. Because the impacts are already visible-we experience them every day. These are no longer future scenarios: this is our present.

And I'd say that young people have a voice. My experience abroad taught me that a few years in the field doesn't mean you lack experience. It just means you have a different perspective. The more you progress, the broader your view becomes. But even a smaller perspective can be just as valid, precise, and valuable.

Finally: what role do you think young people play in science today?

We are a generation of transition. We remember how things used to be, and we are witnessing how they're changing. I've been lucky enough to look at a snow-covered mountain and recognise in that view something I've known all my life. But I also know that view might not last-at least not like that. We may have fewer resources, but we have more energy, more courage, more will to imagine solutions. And that's exactly why science-as I see it-must be based on this feeling of possibility.