SatPerm - Satellite-based Permafrost Modeling across a Range of Scales

Permafrost is found in about a quarter of the land area in the northern hemisphere. Unlike snow and ice cannot permafrost be "observed" with remote sensing techniques. However satellites collect data regarding permafrost from sensors in field, and permafrost can be modeled numerically using above-ground data sets of temperature and snow depth. In SatPerm we aim to see if such data sets can be used for modelling of permafrost.

About the project

Permafrost is permanently frozen ground of which a shallow surface layer thaws every summer and refreezes again in fall. Permafrost is found in about a quarter of the land area in the northern hemisphere which makes it an important element of the Earth's Cryosphere. Unlike for other elements of the Cryosphere, such as glaciers and sea ice, Remote Sensing techniques have remained of limited use in permafrost science since satellite sensors are not able to "see" the state of the ground below the surface.

The physical variable which scientists use to characterize the thermal state of the permafrost is the temperature of the ground. It can be measured directly in boreholes drilled in the frozen ground, but also modeled numerically using above-ground data sets of temperature and snow depth.

These variables, however, are operationally measured by satellite sensors even on global scale, and SatPerm will investigate possibilities of using these data sets as input for permafrost modeling. We will explore different methods and models to achieve this goal.


Simple and fast methods can be applied to map ground temperatures and permafrost for large regions, while computationally demanding approaches will facilitate a higher accuracy for single points. For the latter, Ensemble Kalman Filter methods which have been successfully used in many science fields will be introduced to permafrost science.


For the Lena River Delta in North-East Siberia, a cold permafrost area with ground temperatures around -10°C, we implemented a modeling scheme based on satellite data and the CryoGrid permafrost model developed at UiO. The scheme could very well reproduce measured ground temperatures and thaw depths, and the study is currently in review in the journal “The Cryosphere”. (link to  

In addition, we have conducted field work e.g. on Svalbard and in Mongolia to better understand the variability of ground temperatures, and relate it to external factors such as the distribution of winter snow cover. For the area around Ny-Ålesund on Svalbard, we are currently testing a novel modeling scheme using Ensemble Kalman Filter methods which uses satellite data sets to estimate first the snow depth distribution and subsequently the ground temperature distribution.

The novel techniques developed by the University of Oslo have been recognized internationally: the European Space Agency has included us in its new “GlobPermafrost” project (link to in which satellite-based modeling of permafrost extent will be applied on the pan-arctic scale.


SatPerm will focus on five field sites and regions. The special focus areas are located in Norway, Svalbard, Greenland, North-East Siberia and Mongolia which are hotspots of permafrost research.

At these sites, we will test and benchmark the SatPerm results in close collaboration with partners from Denmark, Germany, Poland, Japan and Mongolia.


This project is funded through the Norwegian Research Council FRINATEK program, project number NFR no 239918.

The project is given in the category "Unge forskertalent".


This project is carried out in cooperation with several researchers from different institutions, see links in right column for participating researchers:

Tags: Permafrost, Remote sensing, Europe, Japan
Published Nov. 17, 2015 3:09 PM - Last modified Oct. 5, 2016 11:25 AM