Modeling of permafrost temperatures in the Lena River Delta, Siberia, based on remote sensing product

Julia.Boike [ at ]


The northeast Siberian lowlands are a climatically sensitive region dominated by permafrost, but monitoring the thermal ground conditions and predicting its future is challenging for such vast areas. A modeling scheme based on gridded remote sensing data, which was recently published for a single grid cell, was extended to the entire Lena River Delta using the transient permafrost model CryoGrid 2. The model is based on the heat transfer equation, calculating the evolution of the soil temperature for every grid cell. The horizontal grid cell size is determined by the remotely sensed forcing data of MODIS Land Surface Temperature (1x1km) and snow depth (1x1km) that was compiled from the GlobSnow Snow Water Equivalent and MODIS Snow Extent products. To assign subsurface properties for each grid cell, a spatially resolved stratigraphic classification was constructed. Based on field observations, such as studies of vegetation, geomorphology and geology, the Lena River Delta was divided into three stratigraphic classes which differ in their layers and layer characteristics, i.e. the volumetric contents of water/ice, mineral, organic and air. From this soil stratigraphy, the soil thermal properties, such as soil thermal conductivity and volumetric soil heat capacity required for the modeling can be inferred for each depth and grid cell. A validation of the MODIS LST forcing time series at one point in the delta revealed a cold bias of up to 3 °C when compared to in-situ measured land surface temperatures. When the gaps in the MODIS data series that occurred due to cloud covered scenes were filled with 2 m - air temperature of the ERA-interim reanalysis, the bias was reduced to -0.8 °C in the average. Therefore, the modeling was conducted with this modified temperature forcing. The model results, in particular ground temperatures and thaw depths, were validated at seven in-situ measurement sites distributed over the delta. For annual average ground temperatures, an agreement within 1°C was found for most validation sites, while modeled and measured thaw depths agreed within 10 cm or less. A sensitivity analysis revealed the influence of the soil stratigraphic classes on ground temperatures and thaw depths, showing differences between classes of more than 2 °C in annual average ground temperature and 50 cm in thaw depths for the same forcing data. The warmest modeled ground temperatures are calculated for grid cells close to the main river channels in the southern parts of the delta, while the coldest are modeled for the northeastern part, an area with low surface temperatures and snow depths. The lowest thaw depths are modeled for the so-called ‘Ice Complex’, an area with extremely high ground ice and soil organic contents. The deepest thaw depths are modeled for grid cells which feature low organic and ice contents and no organic upper layer. The remote sensing driven model scheme demonstrated to be a useful tool for monitoring the thermal state of permafrost and its time evolution in the Lena River Delta. Thus, the approach could be a first step towards operational permafrost monitoring using satellite sensors.

Item Type
Thesis (Master)
Primary Division
Primary Topic
Publication Status
Eprint ID
Cite as
Peter, M. (2015): Modeling of permafrost temperatures in the Lena River Delta, Siberia, based on remote sensing product , Master thesis, University of Leipzig.

[thumbnail of MasterThesisMariaPeter2015.pdf]

Download (5MB) | Preview
Cite this document as:

Add to AnyAdd to TwitterAdd to FacebookAdd to LinkedinAdd to PinterestAdd to Email

Geographical region

Research Platforms


Edit Item Edit Item