Saving Arctic climate data

More and more climate data is being collected. While the large amounts of data certainly provide a better basis for precisely measuring changes in the global climate, it is often challenging to collect, process and store such volumes so the data can be of benefit to other researchers. The Nordic Centre of Excellence eSTICC is looking to solve this problem in a project that will enable climate scientists and computer experts to work together closely.

"This project is unique in that we have received funding for something that is essential for our research and yet never gets funding otherwise," explains centre director Andreas Stohl, a senior researcher and meteorologist at the Norwegian Institute for Air Research (NILU).

“Normally we get funding for a particular research activity, but very often we simply lack the tools we need to carry out the research. The whole purpose of the eSTICC centre is to develop tools and make them more effective. This is incredibly valuable.”

The acronym eSTICC stands for “eScience Tools for Investigating Climate Change at High Northern Latitudes”. The centre specialises in different types of eScience tools, e.g. computational tools or components of larger models, for use in Arctic climate research. Specific activities at the eSTICC centre encompass data processing, inverse modelling, climate modelling, high-performance computing and training and education.

“The Arctic is warming faster than the rest of the earth,” says senior researcher and mathematician Ignacio Pisso, project coordinator of the eSTICC centre. “At the same time it is an under-researched region. We still understand too little about the processes occurring in the Arctic as a result of climate change.”

There is therefore a need to take more measurements as well as develop better measurement tools and models for analysing these. Having IT specialists review some of the scientific software used by climate researchers has made it possible to optimize the Norwegian Earth System Model (NorESM), quadrupling the computational rate of some parts of the program. This has a major impact on the number of calculations that can be performed.

Improved models

A core area at the eSTICC centre is improvement of models used to verify greenhouse gas emissions. One of the centre’s six projects has been focusing on optimising a model for atmospheric methane concentrations. Greenhouse gas concentrations in the atmosphere are typically calculated using greenhouse gas inventories, which are based on emissions figures collected for various greenhouse gases. But there is high uncertainty around these inventories.

“It always gets more complicated when the biosphere is involved,” explains Dr Stohl. “For example, how much greenhouse gas does the forest take up? This question has yet to be properly answered. We also don’t know how much methane the wetlands emit.”

In cases such as these, researchers use “inverse modelling”, first measuring atmospheric methane concentrations and then calculating the emissions that would explain the measurement data. These calculations are then compared to the greenhouse gas inventories to uncover any discrepancies in the inventories. An inverse model for measuring methane gas concentrations has been reviewed by the eSTICC centre’s IT specialists, who were able to optimise the computational rate here by a factor of four as well.

Reusing Arctic data

There is a great need to store the measurements and results already recorded from Arctic climate research in a way that makes them available for other researchers. Another project at the eSTICC centre is dealing with systematising and publishing Arctic climate data from three former Nordic Centres of Excellence (NCoEs) awarded funding under the Top-level Research Initiative: Cryosphere-Atmosphere Interactions in a Changing Arctic Climate (CRAICC), Stability and Variations of Arctic Land Ice (SVALI) and Impact of a changing cryosphere - Depicting ecosystem-climate feedbacks from permafrost, snow and ice (DEFROST).

The eSTICC centre has engaged one or more people from each of those former NCoEs for the purpose of identifying and systematising the climate data collected.

“The centres produced a wealth of data,” says Dr Pisso in explaining the eSTICC centre’s work with the data compiled. “Some of these data were published directly and are thus available via articles, databases or websites. But there is also data that, without our centre’s efforts, could have wound up solely on someone’s hard drive and only known to a very few people.”

Publicly available data

This was the case with, among other things, sets of aerosol data from the former CRAICC centre. These data will eventually be stored in NILU’s EBAS database, a world data centre for aerosol. The eSTICC centre plans to use the CRAICC data in upcoming projects. “Even more importantly,” adds Dr Stohl,“ these data should in principle be publicly available, and when they are, anyone interested can use them in the years to come.”

At the same time, it is important for the eSTICC personnel that their efforts actually make it easier for researchers to navigate their way through the various data and databases. The eSTICC website recently opened a data portal providing an overview of relevant climate data.

“This is a kind of metadata that identifies datasets in the major official databases,” continues Dr Pisso. “There were people who worked to make data more accessible before the eSTICC centre, of course, so it is vital that we don’t duplicate their work. We provide links in areas where relevant and maintained data is available.”

Face-to-face with technicians

Carrying out the eSTICC centre’s activities requires many meetings between the climate researchers and IT specialists. “Researchers typically do not meet personally with the IT people,” says Dr Stohl. “It is all done by e-mail. For this reason the IT experts never really learn about climate research and we don’t learn why the infrastructure is the way it is. But at our centre, this is something we have all really come to understand.”

One benefit resulting from this cooperation is an advanced model for how icebergs calve. The model was derived from work done at the SVALI centre, but before the eSTICC centre’s time this consisted of several separate models using, among other things, different units of measurement. Making the model function required many cumbersome manual calculations and data transfer. But after meetings arranged by the eSTICC centre, the model was streamlined. “This is a good example of something that did not automatically happen simply because the researchers do not have the right expertise,” says Dr Pisso.

“We researchers are usually not interested in formalizing things too much because documenting and formalising require extra effort,” concludes Andreas Stohl. “But sometimes it’s actually worth the time.”

Read more about The Top-level Research Initiative

Read more about eSTICC


Text: Anne Munk Christiansen

Photo: Terje Heiestad

This article has previously been published in NordForsk Magazine 2017