Towards NeIC2019: Interview with Jørn Kristiansen

NeIC2019 takes place at the Tivoli Hotel & Congress Center in Copenhagen, 14-16th May. One of the Keynote speakers at the conference is Jørn Kristiansen, the director of Development Centre for Weather Forecasting at the Norwegian Meteorological Institute. We had the chance to discuss about open data as well as its use in weather forecasting with him.

Why are you attending NeIC2019?

I am invited as a Keynote speaker. I was invited to NeIC2017 too, so this will be my second time. I think it's interesting to follow what's going on in the field of Nordic e-Infrastructure, and it's quite relevant to my work. At the conference, you get an update on the development in i.e. cloud computing, cloud storage - things that are currently going on. It is good to know that there is a Nordic community working on these things.

You are the Director of Development Centre for Weather Forecasting at the Norwegian Meteorological Institute. What does your unit do?

We do research, development, innovation, implementation and operational runs for numerical weather prediction; from observations to end-user products like those that you see on Yr (www.yr.no/en or app). The way the department is organized is quite unique, and it is a modernization and automation of the weather forecasting services driven by science for services. We are organized in and largely funded by (external) projects, have several collaborations from research to operations and our focus area are Norway, Arctic and adjacent areas. Everything we do is connected with getting better forecasts for users. Both in terms of quality and in the way it is presented. A forecast is not complete until it is understood and used.

How do open science and data sharing help you in your work?

We have had an open data policies since 2007. However, open data itself is not user-oriented, it is how you make the data available that is user-oriented. Open data and products are the backbone of Yr and several other internal and external services. There is no exclusivity in the data we use, and everyone can get the same back-end data to present and add value through new services for instance in combination with other data sources or co-production with end-users. Open data and local forecasts creates a strong connection with the users giving us good feedback. The same can be said for open software but there we are just starting to explore the potential. Open data and science are public digital goods and good investments coming from several different sources. For us it creates a strong commitment to translate the science into improved weather forecasts and the steady improvements in forecast quality is both transparent and sustainable across a large project portfolio – emphasizing quality, impact and relevance in our science.

You’re giving a presentation on something called Improved observation usage in numerical weather prediction. What does this mean?

The project – iOBS for short - contributes to improved weather forecast quality by using existing and emerging observation types by combining world-leading numerical weather prediction with future generation e-infrastructure. Within the framework of Nordic collaboration on operational numerical weather prediction, iOBS will first streamline and improve the joint collection and use of weather observations to initialize the numerical weather predictions. That is, get a better ‘diagnosis’ of the weather at a given moment as a start point for integrating its evolution in time (over the next few days). Then iOBS will explore the potential in using citizen observations from a dense network of stations for the same purpose. Since this is both a new system of observations and a very high geographical density network, there are several challenges that must be tackled to harness the opportunities. To this end, the e-infrastructure of the Nordic cloud and Glenna2 are important parts of iOBS.

How does iOBS impact the everyday life of the people?

If successful, we expect better forecasts like the predictions of rain showers for the next several hours, which is important to many user groups. Indirectly iOBS can also be a motivating factor for more citizen observations and science, and it will provide the (open) science for efficiently ingesting more and other types of observations in the weather forecast value chain.

What is your view on the EOSC-Nordic project?

In the Nordic countries, we already have several examples of the benefit of collaboration across disciplines and boarders. The cost for computation and power are increasing, and new technologies are emerging. The idea of an open science cloud is good opportunity for pushing the developments forward. An open science cloud is perhaps not as useful for operational purposes as for research. But if it is easy to transfer results from research to an operational environment and there is a sustainable business model dedicated to operational use, it is worth exploring jointly by in the Nordics. The amount of data is increasing, so it does not make sense for individual institutes, like national meteorological institutes, to archive the same data. An open science cloud can allow experts access the data and to do their computations close to the software and data.

Please find more information and register for NeIC2019 here. You can register until the 1st of May. We hope to see you in Copenhagen in May!