|Paper title||Satellite and Tower-based SAR Observations of Freezing Canopy and Soil – Improving Freeze/Thaw Detection in Boreal Forests|
|Form of presentation||Poster|
Satellite and tower-based SAR observations of boreal forests were investigated for studying the influence of temperature changes on the SAR backscatter of ground surface and forest canopy during the winter. Soil freezing increases the penetration of microwave radiation into the soil, thus reducing the observed backscatter in a wide range of microwave frequencies. Recent studies show that decreasing winter air temperatures causing gradual freezing of the tree canopy increases the canopy transmissivity (optical depth) for microwaves in L-W frequency bands (Li et al., 2019; Schwank et al., 2021). Similarly, radar backscatter from vegetation has been observed to decrease due to freezing for P- to L-bands (Monteith et al., 2018). However, the observed backscatter over forest canopies with the Sentinel-1 C-band SAR increased in very cold winter conditions following canopy freezing (Cohen et al, 2019). The structure and the freezing process affecting the microwave signature of boreal forest canopies are complex. The influence of decreasing air temperature and the consequent canopy freezing on the SAR backscatter has not yet been deeply investigated. Understanding the effect of freezing canopy on the backscatter in below-zero temperatures is important for instance in satellite SAR based retrieval of the freeze/thaw (F/T) state of the soil, as well as in the detection of other surface parameters.
In this study, we analyzed more than 50 ALOS-2 L-band and a similar number of Sentinel-1 C-band SAR satellite acquisitions acquired during winters 2019-2020 and 2020-2021 from Northern Finland. We also performed continuous tower-based SAR measurements in L-, S-, C- and X-bands during the same time periods over a test plot of boreal forest located in the Sodankylä Arctic Space Centre, Northern Finland. A simple water cloud model (Attema and Ulaby, 1978) was applied to simulate the SAR observations of the different frequencies, for retrieving the components affecting the total observed backscatter, such as the ground and canopy backscatter and the canopy transmissivity, in various winter conditions. Special attention was given to the influence of below-zero air temperature changes on the backscatter of the forest canopy, and the implications on satellite SAR based detection of the soil F/T state in the boreal forest environment.
Our preliminary results show that for all analyzed frequencies canopy freezing increases the transmissivity of the forest canopy, when comparing reflecting targets set beneath the forest canopy to reference targets in open areas. On the other hand, for the same forests, changes in the observed backscatter over the forest canopy caused by very cold winter air temperatures were the opposite for high (C, X) and low (L, S) frequencies. As observed previously for Sentinel-1, freezing of the canopy increased the backscatter observed over the forest canopy for C-band SAR. For the higher frequency X-band, the increase in canopy backscatter following canopy freezing was even more prominent. However, for the lower frequency S- and L-bands, canopy freezing led to reduced overall backscatter over forest canopies. Concerning satellite-based soil F/T detection with L-band SAR, these results are encouraging, as the freezing of both soil and canopy lead to lower observed backscatter over boreal forests. Instead, for C-band, the freezing of soil decreases the backscatter from the ground, but the canopy freezing increases the observed backscatter over the canopy, adding complexity to the satellite-based soil F/T detection. Additional research regarding the relation between the canopy transmissivity and canopy backscatter following air temperature changes is required, in order to gain better understanding on the overall behavior of the forest canopy in SAR remote sensing.
Attema E. P. W. and Ulaby F. T., (1978). Vegetation modelled as a water cloud. Radio Science, vol. 13, no. 2, pp. 357-364.
Cohen J., Rautiainen K., Ikonen J., Lemmetyinen J., Smolander T., Vehviläinen J., and Pulliainen J., (2019). A Modeling-Based Approach for Soil Frost Detection in the Northern Boreal Forest Region With C-Band SAR, IEEE Transactions on Geoscience and Remote Sensing, vol. 57, no. 2, pp. 1069-1083.
Monteith A., and Ulander L., (2018). Temporal Survey of P- and L-Band Polarimetric Backscatter in Boreal Forests. IEEE JSTARS, vol 11, no. 10, pp. 3564-3577.
Schwank M., Kontu A., Mialon A., Naderpour R., Houtz D., Lemmetyinen J., Rautiainen K., Li Q., Richaume P., Kerr Y, and Mätzler C., (2021). Temperature effects on L-band vegetation optical depth of a boreal forest, Remote Sensing of Environment, vol. 263.