[S-TT35] Synthetic Aperture Radar and its application

[STT35-P08] Detecting ice conditions in northern Alaska tundra lakes using backscatter intensity of Sentinel-1

Shuto Miyazaki³, *Yoshihiro Iijima¹, Takahiro Abe² (1.Department of Geography, Tokyo Metropolitan University, 2.Graduate School of Bioresources, Mie University, 3.Faculty of Bioresources, Mie University)

Keywords:tundra lake, SAR, lake ice, reversal phenomenon

The North Slope region of northern Alaska consists of tundra, and about 40% of its land area is composed of shallow lakes. These lakes, called "tundra lakes," start to freeze at the end of September and continue to grow until May of the following year, reaching a thickness of about 2 meters or more. In the tundra lakes, ice covers the lakes for 8~9 months of the year. The growth and disappearance of ice in Lake Tundra is important for monitoring environmental changes in the polar regions and the release of greenhouse gases such as methane, so it is desirable to understand the ice conditions of the lake over a wide area. Previous studies monitoring the freezing and thawing of Lake Tundra using remote sensing have examined the backscattering intensity of synthetic aperture radar, and found that the scattering intensity increases with ice growth and that the backscattering coefficient increases or decreases depending on whether water exists under the ice or not. However, little is known about the ice thickness of tundra lakes, and it is necessary to clarify the relationship between ice formation and changes in backscattering intensity based on the seasonal change in ice thickness. Therefore, in this study, we used scatter intensity images from the Sentinel-1 synthetic aperture radar to see how the ice on Lake Tundra freezes and thaws over the course of a year, and at the same time, we attempted to examine the relationship between ice thickness and ice thickness. The target area was Teshekpuk Lake and its vicinity in northern Alaska. We used ENVI's SARSCape to analyze the monthly backscatter intensity images (dB values) for 2023, and examined how the ice grew and melted during the year. As a characteristic of the backscatter intensity seen throughout the season, it was determined that the ice starts to freeze around September, and the ice continues to freeze in spring until May. During the freezing and thawing changes, an inversion of the seasonal trend of the lake's scatter intensity was observed. The reversal phenomenon appeared in July, when thawing progressed, and in October, when freezing began; the reversal phenomenon in July (once rising from a decrease) was considered to be caused by the greater scattering on the surface of the mixture of ice and water (or melting snow). On the other hand, the backscatter intensity reversal in October (from an increase to a decrease) was considered to be caused by the ice beginning to grow and the backscatter return is less likely in thin ice conditions. Comparison of backscatter intensity images with local bathymetry data revealed that the spatial distribution of intensity corresponds to bathymetry, with higher intensity values in shallow areas of the lake (<3 m depth) and lower values in deeper areas (>5 m depth).