Back to search

FRINATEK-Fri mat.,naturv.,tek

Satellite-based Permafrost Modeling across a Range of Scales

Alternative title: Permafrostmodellering med satellitt-data

Awarded: NOK 7.0 mill.

Permafrost is permanently frozen ground of which a shallow surface layer thaws every summer and refreezes again in fall. Permafrost is an important element of the global climate system, and it is included as an "Essential Climate Variable" (ECV) in the "Global Climate Observing System" (GCOS) operated by the "World Meteorological Organization" (WMO). The concept of ECVs is to implement continuous monitoring of environmental variables that are suited to characterize the ECV state. For permafrost, these state variables are the ground temperature and the thickness of the active layer (seasonal thaw layer), but so far only manual monitoring based on a very limited number of points has been implemented. Therefore, scientists, policymakers and stakeholders to date only have limited knowledge on the state and time evolution of the global permafrost extent, which is a significant shortcoming given its importance. The SatPerm project has to goal to create and demonstrate prototypes of novel algorithms that can monitor the permafrost ECV based on the ever growing record of satellite-based remote sensing data. SatPerm has delivered on this goal, developing two schemes that have subsequently been selected by the European Space Agency (ESA) for operational global permafrost ECV mapping and monitoring. As a first step, SatPerm presented an algorithm based on a computationally efficient equilibrium permafrost model (Westermann et al., 2015) which was timely nominated for the international ESA GlobPermafrost project ( to produce the first high-resolution map of the global permafrost extent based on remote sensing data (publicly available at As a main innovative step, this model can for the first time explicitly take the small-scale (<1km) spatial variability of ground temperatures into account, which allows reproducing the permafrost zonations in continuous, discontinuous and sporadic zones in modeling. However, this equilibrium scheme can only reproduce ground temperature and not the active layer thickness, so the permafrost ECV cannot be fully monitored. Therefore, as a second step, SatPerm demonstrated a fully transient model scheme from which not only trends of ground temperature, but also the thickness of the active layer can be obtained, so that the permafrost ECV can be fully monitored. After Westermann et al. (2017) published a first prototype for a small region in Siberia, the scheme was developed further in SatPerm, so that the performance could be significantly improved, especially with respect to the snow representation. In 2018, ESA selected this second SatPerm-developed scheme for its new Permafrost CCI project (, with the goal to implement it on high-performance computing clusters for operational and global modeling of the permafrost ECV. In addition, data assimilation methods developed and demonstrated in SatPerm (Aalstad et al., 2018) show high potential for adding data from newly launched satellite missions (e.g. the Sentinel satellites) to this scheme, which allows producing remote-sensing-based snow cover maps for input to permafrost modeling at so far unprecedented spatial resolution of as little as 10m. In conclusion, the development work in SatPerm has become the seed for operational permafrost ECV monitoring based on satellite data. The rapid adoption of SatPerm-derived algorithms by the European Space Agency ensures timely dissemination of the work both by the international science community and the interested public. For local dissemination, SatPerm PI Sebastian Westermann together with other researchers from the University of Oslo visited highschools in Kautokeino and Lakselv in Finnmark, which are situated close to SatPerm field sites. We presented how field measurements from these sites have been used to develop and validate new algorithms, in addition to a more general introduction to the global significance of permafrost. We believe that high-resolution permafrost products developed in SatPerm will make the rather abstract phenomenon "permafrost thawing" more apprehensible to people, thus leaving a long-lasting legacy also locally for the people living near permafrost. References Aalstad et al. (2018): doi:10.5194/tc-12-247-2018 Westermann et al. (2017): doi:10.5194/tc-11-1441-2017 Westermann et al. (2015): doi:10.5194/tc-9-1303-2015

-Novel field data sets on small-scale distribution of environmental parameters and permafrost state variables -Development of ensemble-based data assimilation schemes targeting permafrost and snow cover -Development of remote sensing - based processing chains for mapping permafrost distribution -Development of a remote sensing - based processing chain for change detection of permafrost temperatures and active layer thickness -Timely integration and application of developed algorithms in international research collaborations -SatPerm has made key steps towards operational mapping and monitoring of permafrost using satellite data and modelling

Permafrost is a key element of the cryosphere which is intimately meshed with the climate system through complex interactions. The current thermal state of the permafrost remains poorly quantified, and even its occurrence and distribution are not well known in many regions. Unlike for other elements of the cryosphere, such as glaciers and sea ice, Remote Sensing techniques have remained of limited use for permafrost monitoring since they are generally unable to observe physical properties of subsurface layers. However, many existing surface observations from satellites can be linked to the permafrost state by means of numerical modeling. SATPERM aims to create, implement and evaluate modeling schemes capable of employing satellite-derived land surface temperatures, snow water equivalent and snow extent for spatially distributed permafrost modeling. For the first time, ground thermal models forced by remote sensing products will deliver maps of the ground thermal regime for large spatial domains. In a second step, SATPERM will introduce state-of-the-art data assimilation techniques to permafrost modeling. By combining Ensemble Kalman Filter methods with ground thermal modeling, the cumulative information content of multi-source data sets can be exploited, and the best possible estimate of the system state and its associated uncertainties derived. A particular emphasis will be put on representing subgrid processes, such as spatially variable snow depths. SATPERM will focus on five field sites and regions, located in Northern Norway, Svalbard, Greenland, North-East Siberia and Mongolia, where research activities by the SATPERM project leader already exist. At these sites and surrounding permafrost areas, the performance of the conceived permafrost modeling schemes will be benchmarked in proof-of-concept studies, upon which future quantitative permafrost monitoring schemes based on satellite data can be built.

Publications from Cristin

No publications found

Funding scheme:

FRINATEK-Fri mat.,naturv.,tek