本日選文（共 4 篇）。

[1843] 以GIT為基礎之場址響應去卷積建立歐地中海區域參考硬岩條件下的地震動模型

Ground‐Motion Model for Reference Hard‐Rock Conditions in the Euro‐Mediterranean Context by GIT‐Based Site‐Response Deconvolution

  • 期刊：Geophysics (GeoScienceWorld) — Advance Access
  • Published：Fri, 15 May 2026 00:00:00 GMT
  • DOI：10.1785/0120250024/730834
  • 原文連結：連結

Summary (EN) The paper addresses the lack of hard‑rock reference motions in strong‑motion databases and develops a method to retrieve and model reference ground motion for shallow crustal earthquakes in the Euro‑Mediterranean region. The authors merge multiple strong‑motion datasets, estimate station site responses using parametric and nonparametric generalized inversion techniques (GITs), select reference hard‑rock stations, and deconvolve site terms to virtually project recordings to very high VS (bedrock) conditions. Resulting ground‑motion models for reference hard‑rock show limited site‑to‑site variability (0.1–0.2 in natural log), indicating effective removal of site effects, and an average VS30‑based amplification model is proposed for cases without site‑specific soil estimates.

重點摘要（繁中） 本文針對強震資料庫中代表性硬岩參考動態不足的問題，提出在歐地中海區域擷取與建立硬岩參考地震動的方法。作者整合多個強震資料庫，利用參數與非參數廣義反演（GIT）估計站點響應，挑選硬岩參考站並以站點項目反卷積紀錄以虛擬方式還原到高VS（出露基岩）條件。所得的參考硬岩地動模型顯示站間變異性低（自然對數0.1–0.2），表示站點效應已被有效移除，並提出一個基於VS30的平均放大模型以供無場址特定土壤響應時使用。

[1848] 地震環境噪訊的時間變化揭示格陵蘭康吉亞 Sermeq Kujalleq 基底條件的季節性變化

Temporal Variations in Seismic Ambient Noise Identify Seasonal Changes in Basal Conditions at Sermeq Kujalleq in Kangia, Greenland

  • 期刊：JGR: Solid Earth (AGU/Wiley) — eTOC
  • Published：Fri, 15 May 2026 23:51:01 -0700
  • DOI：10.1029/2026GL122613
  • 原文連結：連結

Summary (EN) Continuous seismic (HVSR) and GNSS observations at Sermeq Kujalleq were used to probe summer basal conditions. By late June a low-frequency resonance peak appeared in HVSR and coincided with surface acceleration and uplift, consistent with elevated basal water pressure and a transition from a strongly coupled ice-bed state to partial decoupling driven by sustained meltwater. Rayleigh-wave modeling shows these observations can be reproduced by reduced rigidity of basal sediments, highlighting that slower-flowing glaciated areas around ice streams can still undergo rapid seasonal changes.

重點摘要（繁中） 研究使用連續地震（HVSR）與GNSS觀測於Sermeq Kujalleq檢視夏季基底條件。到六月下旬，HVSR出現低頻共振峰，並與地表速度加快及垂直抬升同時發生，符合基底水壓上升，代表由有限融水維持的強耦合狀態轉為由持續融水驅動的部分解耦。雷利波建模顯示觀測到的變化可由基底沉積物剛性降低來重現，指出圍繞冰流的較慢冰蓋區域仍可能出現快速的季節性動態。

[1855] 以無監督機器學習揭示瑞士布拉滕滑坡的地震前兆

Seismic Precursors to the Blatten, Switzerland Landslide Revealed by Unsupervised Machine Learning

  • 期刊：JGR: Solid Earth (AGU/Wiley) — eTOC
  • Published：Fri, 15 May 2026 22:58:29 -0700
  • DOI：10.1029/2025GL121175
  • 原文連結：連結

Summary (EN) The paper analyzes seismic data recorded in the 20 days before the glacier collapse at Blatten, Switzerland (28 May 2025) and reports signals consistent with a nucleation process prior to failure. Using an unsupervised machine learning approach, the authors identified three main groups of events and separated rockfalls from sliding-related seismic signatures within the glacier. The seismicity accelerated in the final two days and the observations are interpreted as consistent with slip-weakening behavior, illustrating the potential of unsupervised learning to detect precursors to collapse.

重點摘要（繁中） 本文分析了2025年5月28日布拉滕（Blatten）冰川塌方前20天的連續地震資料，發現與失穩前的成核過程相符的訊號。研究者以無監督機器學習方法將事件分為三大類，並區分出岩崩與與冰川滑動相關的地震特徵。地震活動在最後兩天加速，觀察結果與滑移弱化行為一致，顯示無監督學習有助於辨識崩塌前的地震前兆。

[1857] GRENet：GNSS 增強的雷達外推網路用於降水即時預報

GRENet: GNSS‐Enhanced Radar Extrapolation Network for Precipitation Nowcasting

  • 期刊：JGR: Solid Earth (AGU/Wiley) — eTOC
  • Published：Fri, 15 May 2026 05:37:12 -0700
  • DOI：10.1029/2025GL120787
  • 原文連結：連結

Summary (EN) Accurate precipitation nowcasting is challenging because radar reflectivity alone can miss key initial conditions of complex weather systems. This study introduces GRENet, a generative network that fuses high-temporal-resolution GNSS water vapor information with radar observations to improve forecasts. In a heavy rainfall case study, GRENet more accurately predicted the range and location of the precipitation center than a radar-only baseline, yielding better critical success index and fractions skill score and showing that GNSS-derived water vapor improves nowcasting skill.

重點摘要（繁中） 降水短時預報很具挑戰性，僅依賴雷達回波常無法充分捕捉複雜天氣系統的初始條件。本文提出 GRENet，一種將高時間解析度的 GNSS 水汽資料與雷達觀測融合的生成式網路，以提升預報表現。針對一場強降雨案例，GRENet 在降水中心的範圍與位置預測上優於僅用雷達的基準模型，並提升了關鍵成功指標與分數技能指標，顯示 GNSS 水汽資訊能顯著增強短時預報能力。