Modelled Ground Fields in Fennoscandia

Horizontal ground magnetic field

dB/dt field

Electric field

Modelled GIC


The purpose of this product is to demonstrate how GIC activity in power grids can be estimated. GIC is calculated from the following input: recent geomagnetic field recordings, ground conductivity models and power grid models.

The plots show the modelled sum of GIC divided by the number of nodes in the Finnish and Norwegian demo grids. This value is considered to be large when it exceeds a threshold corresponding to the top 0.1% of all values in long-term statistics in 1996-2008. This is indicated in the plots by red colour. For the Finnish grid, the threshold is 5.5 A per node and for the Norwegian grid, it is 3.7 A. Cyan dots on the time axis refer to time steps at which data from less than half of magnetometer stations are available.

The following quantities are shown on interactive maps for the previous hour: the interpolated horizontal ground magnetic field and its rate of change, the modelled geoelectric field, and the modelled GIC. Additionally, snap shot maps illustrate the time step at which the GIC sum reached it largest value within the latest 24 hours. For the magnetic field, the snap shots show also measured values. Modelled values are also given as text files; the GIC text file contains only the Finnish grid.

There are specific points to be understood when interpreting results shown in this product:

– The models of the power grids do not describe the present configurations.
– The Finnish model is the configuration of 1978 (for parameter values, see Viljanen et al., 2012).
– The Norwegian model is very similar to the EURISGIC model (Viljanen et al., 2012) with an extension to the north (Myllys et al., 2014).
– Some substations and transmission lines are omitted, but the model still provides a realistic geographic coverage.
– Only a few Swedish substations close to the Norwegian border are included.
– Substations with autotransformers are approximated as single nodes.
– The node resistances in the Norwegian model are approximated to 0.60 ohm (cf. Myllys et al., 2014, Table 3).
– Transmission lines between nodes are assumed to be straight.
– The ground conductivity models for calculating the geoelectric field are approximate.

Despite several simplifications, the use of fixed models of the power grids and of the ground conductivity ensures that GIC activity is characterised in a meaningful quantitative way. In other words, if this product indicates a high GIC level in the demo grids then large GIC likely occur in the true grids too.

This product could be tailored for true power grids, also outside of the region considered here, if precise network parameters are available. Additionally, local real-time geomagnetic recordings and ground conductivity models are necessary. GIC recordings would help in validating power grid and ground conductivity models.


All publications and presentations using data obtained from this site should acknowledge the Finnish Meteorological Institute, Tromsø Geophysical Observatory and the ESA Space Situational Awareness Programme.

For further information about space weather in the ESA Space Situational Awareness Programme see:

Access the SSA-SWE portal here:

These results are very much based on projects that received funding from the ESA Space Situational Awareness Programme’s network of space weather service development activities under ESA contract number 4000113185/15/D/MRP, the European Community’s Seventh Framework Programme (FP7/2007-2013) under grant agreements no 260330 (EURISGIC) and 263325 (ECLAT). Fingrid Oyj and Gasum Oy are acknowledged for their long-term contribution to FMI’s GIC research, and Statnett is acknowledged for collaboration in Norwegian GIC studies.

Credits: Finnish Meteorological Institute (FMI)