Skip to main content
European Commission logo print header

Arctic sea ice, biogeochemistry and impacts on the atmosphere: Past, present, future

Final Report Summary - ASIBIA (Arctic sea ice, biogeochemistry and impacts on the atmosphere: Past, present, future)

Due to the sudden and unfortunate passing of the PI, Prof Roland von Glasow, much of its original scope was cut from the project. However, despite this major setback due to the reduced time available and the loss of Prof. von Glasow’s key expertise, two key objectives were achieved, i.e. the development and construction of a brand new, unique and world class sea-ice research facility (Work Package 1, WP1) and the soon-to-be public release of a community version of the coupled one-dimensional air-sea-ice-snow photochemistry model MISTRA (Work Package 2, WP2). Both milestones will provide a fitting legacy to Prof. von Glasow's work.

WP1: The ASIBIA sea-ice chamber, now re-named the “Roland von Glasow Air-Sea-Ice Chamber (RvG ASIC)”, has been completed and fully commissioned. The glass “ocean” tank (2.4 m x 1.4 m x 1.1 m) has been tested under a range of temperature conditions (–40 ⁰C to +30 ⁰C) with up to 25 cm of sea-ice growth. A Teflon FEP film atmospheric enclosure has been added to the system to carry out gas transfer experiments. A UV-VIS lighting rig has been installed and fully tested.

The development of a centralised measurement system for such a facility is a major achievement, allowing real-time continuous data collection of a wide range of parameters in each of the mesocoms: in the ocean: salinity, temperature, pressure, UV-Vis radiation. in the ice: salinity, solid fraction, temperature, pressure, UV-Vis radiation. in the atmosphere: mole fractions of CH4, CO2, NOx, NOy, O3, wind speed, humidity, pressure, temperature, spectral radiation. Discrete samples of atmosphere and water (under the ice) can be taken without disruption and a full suite of post experiment samplings such as coring and ice block sectioning have been successfully performed.
Three internationally collaborative campaigns have taken place at the RvG ASIC focussing on varying aspects of sea-ice physics or chemistry, demonstrating the use to the community, and the knowledge and expertise that has been sought to improve the facility during the testing phase.
1. Sea-Ice Physics and Sea-Ice Physical Modelling.
2. Gas Fluxes of N2O, CO2 and CH4 through sea-ice.
3. Sea-Ice optical transmission.

WP2: The researcher employed on this work package had started their work only 6 months before the death of the PI. This motivated the decision to try to release a community version of the MISTRA model, previous versions of which the PI had a leading role in developing. This required a shift of the work to document the existing model, ensure its robustness, and improve the code to make it easier to use and modify in future.
Achievement 1: Re-coding to allow compatibility of the existing snow module with the latest version of MISTRA
The snow module was not working with the latest version of the atmospheric part of MISTRA (v7.4.1) but only with an old version (v7.3.3). The snow module is now compatible with v7.4.1.
Achievement 2: Debugging of the whole code and removal of outdated code features
In order to publicly release the model, thorough debugging and removal of several obsolete code features were necessary. Many errors and compiler warnings as well as some informative messages have been corrected/fixed using Fortran (Forcheck) analysis.
Achievement 3: Improvement of the code for numerical efficiency
For this purpose, an analysis of the computing costs of various parts of MISTRA were carried out, and the slowest parts of the model were improved. This meant a reindexing of many arrays (in Fortran, the inner-most index in nested do-loops has to be the left-most array index). This work allowed reducing the duration of runs by nearly 40 %.
Achievement 4: Update of Kinetic Pre-Processor (KPP)
The chemical mechanism included in MISTRA is pre-processed by KPP, whose purpose is to translate chemical equations into a solvable set of equations. The oldest version of KPP was previously used (v1.0) and this version was no longer available online. Thus, the interface between MISTRA and KPP has been updated to fit the latest version of KPP (v2.2.3).
Achievement 5: Adapt the model to fit the requirements for and exchangeable code
Up to now, MISTRA has been a research model and was not freely distributed. Significant improvements to the quality of the code have been made, to make it suitable for community use, e.g. respecting common coding practises based on Fortran 90 features, adding headers to all subroutines and documenting the code in situ. In the past, the user had to modify (hard-code) input data in several places, which was error prone. Now, the model reads all input values from configuration and parameter files.
Achievement 6 (85 % complete): Legal and technical aspects for a public release
Public release of the model code, along with proper documentation and accompanying paper, requires appropriate licensing to make sure that this work will be used for similar purpose, and that future work making use of MISTRA will also be publicly released, so that further improvements to the model can in turn benefit to the scientific community.