University of Oulu

Analysing of frozen ground in Finland : affecting environmental factors, trends in northern Finland and applicability of satellite data

Saved in:
Author: Rimali, Aleksi1
Organizations: 1University of Oulu, Faculty of Science, Geography
Format: ebook
Version: published version
Access: open
Online Access: PDF Full Text (PDF, 3.1 MB)
Pages: 118
Persistent link:
Language: English
Published: Oulu : A. Rimali, 2019
Publish Date: 2019-03-20
Thesis type: Master's thesis
Tutor: Hjort, Jan
Ruha, Leena
Rautiainen, Kimmo
Reviewer: Hjort, Jan
Ruha, Leena


The warming climate will lead to major changes in cold regions in the future. These changes will be more rapid and more severe in high latitudes. This would eventually also affect the seasonal soil frost depth, which has a significant impact on water and energy cycle between atmosphere and soil surface in cold regions. The frozen ground is understood as the soil layer which freezes and thaws annually.

In this study, we investigate the seasonal soil frost depth in Finland 1981–2010 in open, forest and bog environments with three main aims: 1) To study which variables affect the overall thickness of soil frost layer and its changes most 2) To explore are there trends or major changes in the past 30 years on soil frost time series in northern Finland 3) To evaluate the applicability of satellite data against in-situ data in Finland. The main data in this study are frost tube in-situ measurements conducted by the Finnish Environmental Institute in 1981–2010. As a satellite data, we are using National Aeronautics and Space Administration’s Earth System Data Record for Land Surface Freeze/Thaw State (FT-ESDR) and European Centre for Medium-Range Weather Forecast’s (ECMWF) reanalysis model of soil temperature ERA-Interim datasets. For the first objective, we apply GAM (generalized additive model) and LME (linear mixed-effects model) statistical models in multivariate analysis. In the second objective, we are using the Mann-Kendall trend test and the Sen’s slope estimate to conduct trend analysis. In the third objective, we evaluate the satellite-based measurement against in-situ observations with contingency tables.

Based on the multivariate analysis, the most statistically significant factors were air temperature, snow depth, precipitation and north coordinate. The interaction plots revealed that the effect of air temperature and snow depth to the maximum depth of soil frost is not a linear and varied in open, forest and bog environments. The yearly average of maximum depth of soil frost had decreased 2.12 cm/year on open, 2.75 cm/year on forest and 0.5 cm/year on bog sites in 1981–2010. The most distinct decreases were experienced in May in all three site types. The FT-ESDR and ERA Interim had the highest error rate percentages (avg. 68% and 56%) during shallow snow cover and soil frost depth. The accuracy increased steadily with the increasing soil frost and snow layer.

The study revealed that the seasonal soil frost depth has been decreasing between 1981 and 2010 in Finland. This study aimed to give more insight about the multidimensional process of frozen ground. Results can be applied in future research planning. The way to improve the current setting would require information about factors like soil moisture, groundwater, and extensive data from a longer period of time.

see all

Copyright information: © Aleksi Rimali, 2019. This publication is copyrighted. You may download, display and print it for your own personal use. Commercial use is prohibited.