Two Decades of Arctic Coastal Sea Ice Phenology: A GNSS-IR Record from the Beaufort Sea (2003-2023)
收藏资源简介:
Arctic coastal sea ice phenology is a sensitive indicator of climate change, yet long-term, sub-kilometer observations remain scarce. This study presents a continuous 20-year (2003–2023) climate data record of coastal sea ice phenology derived from ground-based Global Navigation Satellite System Interferometric Reflectometry (GNSS-IR) at Tuktoyaktuk in the Canadian Beaufort Sea. By employing the Amplitude Integration Factor (AIF) method, we successfully constructed a robust time series that maintains continuity across decadal hardware changes. The 20-year complete climatology establishes a mean ice season duration of 254.1 ± 8.6 days. Trend analysis (excluding the anomalous 2018 ice year) reveals a statistically significant shortening of the ice season at a rate of -6.41 ± 6.31 days per decade (p=0.08). Critically, this declining trend remains robust even when analyzing the "corrected ice duration" (representing complete freeze-thaw cycles), which shows a steady shortening of -4.63~-4.19 days per decade regardless of whether the anomaly is included (p= 0.04~0.06). This decline is asymmetric, driven primarily by a delayed freeze-up (+3.40 days/decade) rather than earlier breakup (-1.42 days/decade). Comparison with 4-km gridded satellite products (IMS) demonstrates that GNSS-IR detects freeze-up onset 5.5 ± 3.7 days earlier, highlighting the sensor's unique capability to resolve initial nearshore frazil ice formation that is typically invisible to coarse-resolution satellite sensors. Thermodynamically, phenological transitions are strongly controlled by atmospheric forcing, with freeze-up timing exhibiting a high correlation (R2=0.74) with accumulated Freezing Degree-Days (FDD). This work validates GNSS-IR as a scalable, cost-effective tool for monitoring the complex land-ocean interface, providing essential ground-truth data for validating regional climate models. The dataset includes annual phenology metrics and hourly AIF time series derived from the TUKT GNSS station.



