Mean sea level variability in the North Sea: Processes and implications
Mean sea level variability in the North Sea: Processes and implications
复制标题
DOI：
10.1002/2014jc009901
复制
发表时间：
2014-10-01
影响因子：
3.6
通讯作者：
Jensen, Juergen
中科院分区：
文献类型：
--
作者：
Dangendorf, Soenke;Calafat, Francisco M.;Jensen, Juergen
在考虑不同强迫因子的情况下，研究了自19世纪后期以来北海在一系列时间尺度上的平均海平面（MSL）变化。我们使用多元线性回归模型来确定海洋对大气强迫波动的正压响应，这些模型已在20世纪下半叶通过潮汐+浪涌模式的输出进行了验证。我们发现，局地大气强迫主要引发了数年时间尺度上的MSL变率，倒压效应主导了英国和挪威海岸线的变率，风控制了南部从比利时到丹麦的MSL变率。在年代际尺度上，MSL变率主要反映空间变化，这种空间变化在很大程度上是远程强迫的。对测高观测和网格立体高度的空间相关性分析表明，有证据表明从挪威大陆架一直延伸到加那利群岛的相干信号。这与沿北大西洋东部边界的海岸风强迫理论相吻合，该理论导致沿海困波沿着大陆斜坡传播数千公里。这些发现的意义用统计蒙特卡洛实验进行了评估。结果表明，去除已知的可变性增加了信噪比，结果是：(1)可以更准确地估计线性趋势；（ii）可以更早地检测到可能的加速（如预期的，例如由于人为气候变化）。这些信息对于海岸的预先管理、工程和规划至关重要。
Mean sea level (MSL) variations across a range of time scales are examined for the North Sea under the consideration of different forcing factors since the late 19th century. We use multiple linear regression models, which are validated for the second half of the 20th century against the output of a tide+surge model, to determine the barotropic response of the ocean to fluctuations in atmospheric forcing. We find that local atmospheric forcing mainly initiates MSL variability on time scales up to a few years, with the inverted barometric effect dominating the variability along the UK and Norwegian coastlines and wind controlling the MSL variability in the south from Belgium up to Denmark. On decadal time scales, MSL variability mainly reflects steric changes, which are largely forced remotely. A spatial correlation analysis of altimetry observations and gridded steric heights suggests evidence for a coherent signal extending from the Norwegian shelf down to the Canary Islands. This fits with the theory of longshore wind forcing along the eastern boundary of the North Atlantic causing coastally trapped waves to propagate over thousands of kilometers along the continental slope. Implications of these findings are assessed with statistical Monte-Carlo experiments. It is demonstrated that the removal of known variability increases the signal to noise ratio with the result that: (i) linear trends can be estimated more accurately; (ii) possible accelerations (as expected, e.g., due to anthropogenic climate change) can be detected much earlier. Such information is of crucial importance for anticipatory coastal management, engineering, and planning.