Soil freezing in northern aapa mires : freeze/thaw -detection using portable L-band radiometer
Viuho, Elmeri (2023-11-20)
Viuho, Elmeri
E. Viuho
20.11.2023
© 2023 Elmeri Viuho. Ellei toisin mainita, uudelleenkäyttö on sallittu Creative Commons Attribution 4.0 International (CC-BY 4.0) -lisenssillä (https://creativecommons.org/licenses/by/4.0/). Uudelleenkäyttö on sallittua edellyttäen, että lähde mainitaan asianmukaisesti ja mahdolliset muutokset merkitään. Sellaisten osien käyttö tai jäljentäminen, jotka eivät ole tekijän tai tekijöiden omaisuutta, saattaa edellyttää lupaa suoraan asianomaisilta oikeudenhaltijoilta.
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:oulu-202311203249
https://urn.fi/URN:NBN:fi:oulu-202311203249
Tiivistelmä
Seasonal soil freezing is one of the most significant sources of uncertainty in methane emissions from high latitude wetlands. Although soil freezing can be remotely sensed with current satellite-based instruments, the resolution is not high enough to detect small-scale variations within individual mires. In this study, a lightweight radiometer mounted on an unmanned aerial vehicle (UAV) was tested for detecting the freeze/thaw (F/T) state of the soil in aapa mires in Finnish Lapland. The three main research questions were the suitability of the radiometer for high resolution F/T detection, the existence of possible spatial patterns in the timing of soil freezing, and the effects of environmental factors on these spatial patterns.
As this was the first study to use a UAV-mounted radiometer for F/T detection, there was no established method for retrieving the F/T state of the soil from the measured brightness temperature values. In previous studies using satellite-based instruments, the F/T state of the soil is determined by a threshold method where the measured values are scaled pixel-wise between known reference values of thawed and frozen soils and classified based on a fixed threshold. This method was modified for use with UAV measurements. The performance of the radiometer was evaluated by comparing the measurement results with tower-based radiometer and in-situ measurements in the study area. Spatial patterns in the timing of soil freezing were investigated using analysis of variance and measures of spatial autocorrelation. The effects of environmental factors were investigated using generalized linear models (GLM), generalized additive models (GAM), and hierarchical partitioning with environmental variables derived from readily available remote sensing materials.
The F/T state of the soil was successfully determined from the UAV measurements, and the results were comparable to those of other measurements in the study area. Variation in the spatial distribution of the timing of soil freezing was detected at the local scale. The soil appeared to freeze as a result of two separate major freezing events and was therefore modeled as a binary response variable. Both GLM and GAM showed that the most significant factors contributing to the spatial patterns were the Enhanced Vegetation Index (EVI), the flark area and the standard deviation of the Topographic Wetness Index (TWI). Hierarchical partitioning highlighted the individual effects of EVI. All detected relationships were strongly correlated with the microtopographic structure of the mire, suggesting that seasonal freezing progresses differently on different surface types.
As this was the first study to use a UAV-mounted radiometer for F/T detection, there was no established method for retrieving the F/T state of the soil from the measured brightness temperature values. In previous studies using satellite-based instruments, the F/T state of the soil is determined by a threshold method where the measured values are scaled pixel-wise between known reference values of thawed and frozen soils and classified based on a fixed threshold. This method was modified for use with UAV measurements. The performance of the radiometer was evaluated by comparing the measurement results with tower-based radiometer and in-situ measurements in the study area. Spatial patterns in the timing of soil freezing were investigated using analysis of variance and measures of spatial autocorrelation. The effects of environmental factors were investigated using generalized linear models (GLM), generalized additive models (GAM), and hierarchical partitioning with environmental variables derived from readily available remote sensing materials.
The F/T state of the soil was successfully determined from the UAV measurements, and the results were comparable to those of other measurements in the study area. Variation in the spatial distribution of the timing of soil freezing was detected at the local scale. The soil appeared to freeze as a result of two separate major freezing events and was therefore modeled as a binary response variable. Both GLM and GAM showed that the most significant factors contributing to the spatial patterns were the Enhanced Vegetation Index (EVI), the flark area and the standard deviation of the Topographic Wetness Index (TWI). Hierarchical partitioning highlighted the individual effects of EVI. All detected relationships were strongly correlated with the microtopographic structure of the mire, suggesting that seasonal freezing progresses differently on different surface types.
Kokoelmat
- Avoin saatavuus [32049]