论文部分内容阅读
海洋异常事件(Marine Abnormal Event,MAE)可为区域海气相互作用和全球气候变化研究提供重要的时空特征参考,具有重要的科学意义。鉴此,本文基于长时间序列的栅格数据集,提出了一种海洋异常事件时空提取算法(Marine Abnormal Event Spatio-Temporal Extraction Method,MAESTEM)。MAESTEM的核心步骤包括事件的时间维度提取、事件的空间维度提取和事件追踪。在时间维度提取方面,将每一个栅格像元作为一维时间序列,计算其平均值和标准差作为判断每个时刻是否异常的标准,并根据异常持续发生的时间长短来提取时间维度的海洋异常事件(Temporal MAE,TMAE)。在空间维度提取方面,利用空间邻域统计方法,统计栅格像元的空间邻域中属于TMAE的个数,并通过空间维度异常判断规则获取空间维度的海洋异常事件(Spatial MAE,SMAE)。利用时刻状态的SMAE的空间拓扑关系,根据事件前后时刻覆盖的空间区域是否重叠以及事件持续的时间长短,实现异常事件的追踪。最后,通过提取太平洋海域1993年1月至2012年12月的月均海面高度异常(Sea Level Anomaly,SLA)事件,验证了该算法的有效性和实用性。
Marine Abnormal Event (MAE) is of great scientific significance for providing important spatial and temporal characteristics of regional sea-air interaction and global climate change research. In view of this, this paper presents a Marine Abnormal Event Spatial-Temporal Extraction Method (MAESTEM) based on a long-time raster dataset. The core steps of MAESTEM include time dimension extraction of events, spatial dimension extraction of events and event tracking. In terms of time dimension extraction, each grid cell is regarded as a one-dimensional time series, and its average value and standard deviation are calculated as a criterion for judging whether each time is abnormal or not. The time dimension of the ocean is extracted according to the duration of the abnormality Abnormal event (Temporal MAE, TMAE). In terms of spatial dimension extraction, the number of TMAE belonging to the spatial neighborhood of the grid cells is calculated by using the spatial neighborhood statistical method, and Spatial MAE (SMAE) is obtained through spatial dimension abnormality determination rules. The spatial topological relationship of SMAE at the moment is utilized to track abnormal events according to whether the spatial areas covered by the events before and after overlap and the duration of the events. Finally, the validity and practicability of this algorithm are verified by extracting the Sea Level Anomaly (SLA) events from January 1993 to December 2012 in the Pacific Ocean.