EN中文

Journal of Geo-information Science >

2016 , Vol. 18 >Issue 4: 453 - 460

DOI: https://doi.org/10.3724/SP.J.1047.2016.00453

Orginal Article

Time Series of Raster-oriented Method for Marine Abnormal Events Extraction

  • LI Xiaohong , 1, 2 ,
  • YAN Jinfeng 1 ,
  • LI Yilong 2 ,
  • XUE Cunjin , 2, 3, *
Expand
  • 1. Shandong University of Science and Technology, Qingdao 266590, China
  • 2. Key Laboratory of Digital Earth Science, Institute of Remote Sensing and Digital Earth, Chinese Academy of Sciences, Beijing 100094, China
  • 3. Key Laboratory of Earth Observation, Sanya 572029, China
*Corresponding author: XUE Cunjin, E-mail:

Received date: 2015-10-13

  Request revised date: 2015-12-04

  Online published: 2016-04-19

Copyright

《地球信息科学学报》编辑部 所有
Fold

Abstract

Marine Abnormal Event (MAE) is an abnormal decrease or increase of marine environmental parameters, which covers a specified spatial domain and lasts for a specified temporal duration. Marine abnormal event can provide the temporal and spatial characteristics of regional sea-air interactions and global climate change, which has an important scientific significance. Based on the above information, we propose a novel algorithm to extract the MAE from the long-term raster datasets, named as MAESTEM (Marine Abnormal Event Spatio-Temporal Extraction Method). The MAESTEM has three key steps: the extraction of MAE at a temporal dimension, the extraction of MAE at a spatial dimension, and the tracking of MAE. At the temporal dimension, each grid pixel within an image is taken to be the one-dimensional time series, and its mean and standard deviation are taken as the criteria to define its abnormal snapshot status. If and only if the abnormal snapshot and its subsequent ones are not smaller than the specified threshold, i.e. T, the abnormal snapshots are defined as a temporal MAE, denoted by TMAE. We utilize the spatial neighborhood statistics method to count the number of spatial neighborhoods of a raster pixel which belongs to TMAE and to obtain the marine abnormal events at the spatial dimension, denoted as SAME, by using the spatial dimension abnormal extracting method. In the final step, we use the spatial topological relationship of SMAE to identify whether the SAMEs at the previous and post snapshots belongs to the same event. If they overlap, they are considered being in the same event. If the temporal duration of the event exceeds the temporal threshold, save the event, otherwise, delete it. Finally, the Pacific Ocean is taken as a research area, and its monthly averaged sea level anomaly (SLA), which is obtained from the remote sensing imagery during a period from January 1993 to December 2012, is used to test the feasibility and efficiency of MAESTEM.

Key words： Marine Abnormal Event (MAE); extraction methods; raster datasets; Pacific Ocean

Cite this article

LI Xiaohong , YAN Jinfeng , LI Yilong , XUE Cunjin . Time Series of Raster-oriented Method for Marine Abnormal Events Extraction[J]. Journal of Geo-information Science, 2016 , 18(4) : 453 -460 . DOI: 10.3724/SP.J.1047.2016.00453

1 引言

海洋异常事件（Marine Abnormal Event,MAE）是指在特定的海洋区域内,海洋环境要素（如海洋表面温度、叶绿素-a浓度）相对于平均状态下的异常升高或异常降低。海洋异常事件通常持续一定的时间范围,覆盖特定的空间区域,其时空特性对海洋时空分析方法、区域海气相互作用、全球气候变化研究和人类社会都有重大的影响[1-2]。随着对地观测技术的发展,长时间序列的卫星遥感数据为获取宏观尺度的地理现象或特征提供了重要数据源[3],如何从长时间序列的卫星遥感数据中提取地理异常事件已成为研究热点。
不同领域的专家针对不同的应用目的,提出了不同的异常事件提取算法。例如,在异常探测领域,时空异常探测主要致力于寻找在时间上和空间上的异常属性值。Cheng等利用时空邻域的理论在栅格图像中发现时空异常值[4];Lu等利用图像分割技术分析气象图像序列中的时空异常值[5],并设计了相应的方法;Birant等利用基于密度聚类的方法寻找时间和空间上的异常值[6];Das等利用基于距离和基于邻域的异常探测查找全球气候数据的时空异常[7]; Benezeth等针对行为模拟和异常检测提出了一个基于位置的方法,用马尔科夫区域随机模型计算同时发生的可能性,检测视频中异常事件的发生,如异常交通事件中汽车的非法转弯等[8]。
在地理科学领域,传统的异常事件描述方法有区块极值模拟,利用广义极值分布（Generalized Extreme Value,GEV）模拟一组连续观测块的最大值[9];Yu等提出了一个统计模型-时空图形化模型,采用广义极值分布（GEV）模拟时空异常事件,模型可以用来估计异常事件的时间格局,根据当前数据的趋势,预测未来异常事件的分布,如极端降雨等,这为洪涝灾害预警和策略计划提供了重要的帮助[10];Huang等利用空间序列索引以及基于时域分片的算法查找连续的时空事件,讨论了时空事件数据集的序列模式[11];Tan等将地球科学原始数据转换为市场购物篮数据的形式来辨别时空事件,并挖掘事件序列之间的关联模式[12];Wu等提出了一个基于范围搜索最近邻域（Range-based Searching Nearest Neighbors,RSNN）的空间聚类算法,首先通过原始时间序列数据聚类减少数据大小和空间自相关,然后利用extractEvents算法从聚类的时间序列数据中提取异常事件序列,最后进行关联模式挖掘[1]。
尽管上述方法在异常事件提取方面取得了一系列成果,但Cheng、Kut等只是发现时空异常值,并没有考虑异常在时间和空间上的连续性;Lu等提取异常值并追踪区域异常值,Yu等模拟预测异常事件,Huang等查找连续的时空事件,Tan等辨别时空事件,Wu等提取异常事件序列,但都没有讨论异常事件在空间上的分布以及随着时间的变化性。大量的海洋现象具有连续渐变的过程特性[13],因此,在时空连续渐变的海洋异常事件提取方面仍然存在许多挑战。鉴此,本文从时空连续变化的特性出发,设计了海洋异常事件的时空提取算法。

2 海洋异常事件时空提取算法

海洋异常事件时空提取算法（Marine Abnormal Event Spatio-Temporal Extraction Method,MAESTEM）是利用长时间序列的海洋栅格数据集,根据海洋环境要素的时空变化特性,提取海洋异常事件。MAESTEM主要包括4个步骤：（1）数据预处理,用于消除季节波动引起的变异;（2）事件的时间维度提取,实现事件在时间上的连续性;（3）在此基础上的空间维度提取,实现事件在空间上的连续性;（4）MAE追踪算法,实现事件在时空上的连续性。具体流程如图1所示。
Fig. 1 An extraction flowchart of MAE

图1 海洋异常事件提取流程

2.1 数据预处理

相对于正常的季节性周期模式,地理学家更关注异常模式,如厄尔尼诺（El Niño）、拉尼娜（La Niña）等。地球科学数据具有很强的季节性成分,数据模式经常受季节性变化影响,只有去除气候时间序列的季节性成分后,异常气候事件（如干旱、洪水、热浪等）才更明显[14]。因此,在识别异常气候事件前,去除气候时间序列的季节性成分是关键。本文注重海洋异常变化,而长时间序列的海洋环境参数季节性变化主要受太阳辐射影响,因此在识别海洋异常事件前,先移除由季节性太阳辐射引起的变化。根据先前知识,标准化月均距平算法Z-score能够有效地去除季节性成分[14]。针对1-12月的任意月份,如1月,获取每年1月的值,构成时间序列,计算其平均值和标准差,并通过其均值和标准差,离散化任意年份1月的数值,如式（1）所示[15]。
X i , j ' = X i , j - X j ¯ δ j （ j = 1,2 , … , 12 ） （1）
式中：i为年; j为月; X j ¯ 和 δ j 分别为平均值和标准差; X i , j ' 和 X i , j 分别为长时间影像的原值和转换值。

2.2 时间维度的异常提取算法

经过数据预处理,长时间序列的遥感影像变为月均距平栅格数据,根据MAE的定义,MAE应该持续一定的时间段,通过时间邻域窗口大小来表达这个时间段。假设时间邻域窗口大小为WT,时间标准阈值为TCT,且 WT ≥ TCT ， 则时间维度的MAE提取步骤如下：
（1）将每一个栅格像元作为一个时间序列,获取每个时间序列的栅格像元值,假设 V ( row , col ) = ( v t 1 , v t 2 , v t 3 , … , v tn ) ,其中row和col分别为栅格数据中的行号和列号,t1、t2、…、tn为时刻状态,n为数据的时间序列长度;
（2）计算V(row,col)的平均值M(row,col)和标准差SD(row,col) ;
（3）对于每一个时间间隔t,获取栅格像元的值Vt,根据式（2）判断是否异常;
S ( V t ) = 1 V t ≥ M + kSD 0 其他 - 1 V t ≤ M - kSD （2）
式中：1代表异常升高;-1代表异常降低;0代表正常;k是比例因子。
（4）重复步骤（3）,直到所有时间序列的栅格像元都处理完毕;
（5）根据式（3）提取时间维度MAE,如果式（3）为真,则此栅格像元在持续的时间段内被定义为一个海洋异常事件;
∑ t = - WT 2 t = WT 2 T s ( v t ) ≥ TCT （3）
式中： T s ( v t ) 代表异常持续的时间长度。
（6）重复步骤（1）-（5）,直到所有的栅格像元都处理完毕。

2.3 空间维度的异常提取算法

在时间维度异常提取的结果基础上,进行空间维度的异常提取。时间维度上的海洋异常事件保证了每一个栅格像元事件在时间序列上的连续性,然而,面向栅格影像,任意时刻栅格像元之间的空间关系被割裂。空间维度的海洋异常事件旨在实现空间区域的连续性,即覆盖一定的空间区域（该区域中栅格像元的属性值是一致的）,属性值为-1表示该区域发生了海洋异常降低事件,属性值为1表示该区域发生了海洋异常升高事件。因此,空间维度的海洋异常事件应该去除空间域的噪声信息,把相邻有意义的空间区域连接起来。在地理科学中,空间邻域是空间异常探测的重要组成部分[16],一般采用空间邻域低通滤波去除空间噪声。由于海洋环境要素受外界影响较小,方形的空间邻域更具应用性,因此本文利用方形邻域去除空间噪声。假设空间邻域窗口尺寸为w,空间维度的异常事件提取算法具体步骤如下：
（1）T时刻,获取一个栅格像元的值,假设为 V ( row , col ) ;
（2）获取该栅格像元w×w邻域的值,统计空间邻域内属性值为-1、0、1的个数,分别用 Num ( - 1 ) 、 Num(0)、Num(1)表示,则MV=max(Num(-1)、Num(0)、 Num ( 1 ) ) ;
（3）根据式（4）,做去噪处理;
V ( row , col ) = - 1 MV = Num ( - 1 0 MV = Num 0 1 MV = Num 1 （4）
（4）重复步骤（1）-（3）,直到处理完所有栅格 像元。

2.4 海洋异常事件追踪算法

时间维度的海洋异常事件提取算法保证了事件在每一个栅格像元上时间序列的连续性,空间维度的海洋异常事件提取算法保证了事件在时刻状态上空间区域的连续性。海洋异常事件是一个连续渐变的时空过程[17],覆盖特定的空间区域,覆盖的空间区域会随着时间而改变,但是不同时刻的空间区域会重叠,因此可以利用拓扑关系（如拓扑相交、相并）来追踪海洋异常事件。
在一定的时间间隔内,若空间覆盖区域有叠加,且事件类型相同（升高或降低）,则认为是同一个事件。在图2中,假设灰色范围代表异常降低事件的覆盖范围,橙色范围代表异常升高事件的覆盖范围,由图2可知,t1时刻和t2时刻异常事件空间覆盖范围有重叠,且事件类型相同,则t1时刻的异常事件和t2时刻的异常事件是同一个事件相邻2个时刻的状态。海洋异常事件的生命周期包括：产生阶段、发展阶段、稳定阶段、减弱阶段、消亡阶段[18]。如图2所示,t1为产生阶段,t2为发展阶段,t3为稳定阶段,t4为减弱阶段,t5为消亡阶段,且每个时刻不同区域可能会发生不同的异常事件。例如,t5时刻发生2个异常事件,即降低事件和升高事件。本文提出的异常提取算法是针对一个海洋异常事件,使用拓扑关系（拓扑相交,相并）来追踪海洋异常事件的整个发生过程。
Fig. 2 Spatial and temporal information of MAE

图2 海洋异常事件的空间和时间信息

通过海洋异常事件的生命周期过程（产生-发展-稳定-减弱-消亡）可知,海洋异常事件至少持续一定的时间段,假设为MAET,具体算法如下：
（1）提取T时刻的海洋异常事件preMAE假设为MAE_1,以及异常事件的空间范围、时间信息、事件类型等信息,存到相应的数据库,则MAE_1的空间范围就是PreMAE的空间范围;
（2）提取T-1时刻的所有海洋异常事件,以及异常事件的空间范围等信息;
（3）判断PreMAE空间范围和T时刻的海洋异常事件空间范围是否有交集：如果没有交集,说明是另一个事件的起始,执行步骤(4);如果有交集,并且T时刻的所有异常事件都已经遍历,T=T+1,执行步骤(1),否则直接执行步骤(1);
（4）提取T+1时刻的海洋异常事件,以及异常事件的空间范围、时间信息、事件类型等信息;
（5）判断T+1时刻的异常事件postMAE和T时刻的异常事件preMAE是否是同一个事件：
① 如果postMAE的空间范围和preMAE空间范围有交集,则认为postMAE和preMAE是一个事件,postMAE和preMAE空间范围的并集为异常事件MAE_1的空间范围,则事件持续时间N=N+1,更新数据库,执行步骤(4);
② 如果preMAE空间范围和任何的postMAE的空间范围都没有交集,且如果事件持续时间N小于给定的阈值MAET,那么认为PreMAE不是合格的异常事件,将PreMAE及其信息从数据库中删除,更新数据库。如果事件持续时间N大于阈值MAET,则是合格的异常事件,保存数据库。如果T时刻的所有异常事件已经遍历,T=T+1,执行步 骤(1);
（6）重复步骤（1）-（5）,直到遍历所有时刻,追踪所有异常事件。

3 实例验证

太平洋海域是海洋环境要素和ENSO（El Niño Southern Oscillation）的多互动区域,在全球气候变化和区域海气相互作用中扮演重要的角色[19],因此,本文选择太平洋海域作为研究区域（100° E~60° W,50° S~50° N）,1993年1月至2012年12月的月均海表面高度异常（Sea Level Anomaly,SLA）作为实验数据。SLA数据来源于AVISO分发中心（http://www.aviso.oceanobs.com/duacs）,空间分辨率为0.25º,时间分辨率为月,空间覆盖为全球。根据ENSO的定义,海洋尼诺指数持续5个月以上高于0.5º（或低于-0.5º）定义为一次El Niño（或La Niña）事件,所以时间阈值设定为5个月,即TCT=5 和MAET=5。根据经验统计和正态分布理论,k=1,提取的海洋异常事件空间覆盖范围过大,区域差异性很难体现,k=3,提取的海洋异常事件较少,时间上的连续性很难保证,所以比例因子k设定为2,即k=2;空间邻域采用3×3邻域,即w=3,栅格数据集的空间分辨率为0.25 º,所以空间邻域窗口大小设定为0.75 º。基于上述阈值,1993年1月至2012年12月的月均SLA共有37个海洋异常降低事件,29个海洋异常升高事件。
ENSO是全球气候变化的典型信号之一,和SLA之间存在相互“响应”和“驱动”关系,且1997-1999年的ENSO事件是130年来最强的一次气候异常事件[20],因此选择该期间的SLA海洋异常事件作为分析实例。图3显示了1997-1999年的一次El Niño事件和一次La Niña事件,MEI大于0.5的区间定义为El Niño事件,MEI小于-0.5的区间定义为La Niña事件[21]。图4为1997年3月至1999年8月的SLA海洋异常事件,其中蓝色代表异常降低事件,红色代表异常升高事件。从图4中可以看出,1997-1999年发生2个异常降低事件和一个海洋异常升高事件。在西太平洋发生的海洋异常降低事件定义为WPOMAE（Western Pacific Ocean Marine Abnormal Event）,1997年4月产生,1999年8月消失。在北太平洋的海洋异常降低事件定义为NPCMAE（North Pacific Ocean Marine Abnormal Event）,1997年11月产生,1999年1月消失。东太平洋发生的海洋异常升高事件定义为EPCMAE（Eastern Pacific Ocean Marine Abnormal Event）,1997年3月产生,1998年11月消失。
结合图3、图4分析可知,El Niño事件1997年3月开始发生,WPOMAE在El Niño之后的4月产生,随着El Niño的增强,WPOMAE覆盖范围不断扩大,位置不断由西太平洋向中太平洋移动;1998年6月El Niño事件消失,1998年8月La Niña事件产生,WPOMAE覆盖范围不断衰减分裂,1999年8月在中太平洋消失。
Fig. 3 ENSO index from 1997 to 1999

图3 1997-1999年ENSO指数图

Fig. 4 Tracking MAE from 1997 to 1999

图4 1997-1999年海洋异常事件追踪过程

由图3、图4可知,El Niño事件在1997年3月产生的同时EPCMAE在中太平洋突然产生,并随着El Niño事件的增强,EPCMAE覆盖范围不断扩大,位置不断向东太平洋移动;1998年2月随着El Niño事件强度的减弱,EPCMAE覆盖范围不断衰减分裂;1998年8月开始产生La Niña事件,EPCMAE基本消失;1998年11月EPCMAE在东太平洋完全消失。
众所周知,ENSO事件与太平洋海域SLA存在相互“响应”与“驱动”的关系。Casey和Adamec对1993-1999年的SLA数据进行EOF分析和SVD分析,结果显示El Niño发生期间（1997-1998年）,赤道西太平洋海面高度降低,赤道东太平洋海面高度升高[22],这与EPCMAE和WPOMAE期间的SLA变化相吻合。此外,通过分析WPOMAE和EPCMAE的时空变化,还可获得SLA在西太平洋和东太平洋的空间分布及在时间上的动态变化。

4 结论

面向长时间序列的卫星遥感影像,本文从时间维度、空间维度和事件追踪3个层次设计了一种全新的海洋异常事件提取算法（MAESTEM）,并利用1993年1月至2012年12月的月均SLA数据,验证了算法的有效性和实用性。本文集成空间维度和时间维度算法来提取海洋异常事件,并利用空间拓扑关系追踪海洋异常事件。海洋异常事件的提取结果能够充分地体现海洋异常事件在发生过程中的空间区域分布,及其随时间的动态变化,这对科学研究人员更好地了解海洋区域特征具有重要意义。例如,海洋异常事件的持续周期及空间覆盖范围的演变,可为区域海气相互作用提供事实特征。本算法也存在不足：当事件在前后时刻空间覆盖范围变化比较大时（即前后时刻的空间覆盖范围没有重叠）,则不能有效地进行海洋异常事件的追踪。这有待进一步完善与修改,是今后工作的重要内容之一。

The authors have declared that no competing interests exist.

References

Publishing order | Descend order by publishing year | Descend order by cited within
[1]
Wu T S, Song G J, Ma X J, et al.Mining geographic episode association patterns of abnormal events in global earth science data[J]. Science in China Series E: Technological Sciences, 2008,51(1):155-164.Abnormal events in earth science have great influence on both the natural envi-ronment and the human society. Finding association patterns among these events has great significance. Because data in earth science has characteristics of mass,high dimension,spatial autocorrelation and time delay,existing mining technolo-gies cannot be directly used on it. We propose a RSNN (range-based searching nearest neighbors) spatial clustering algorithm to reduce the data size and auto-correlation. Based on the clustered data,we propose a GEAM (geographic episode association pattern mining) algorithm which can deal with events time lags and find interesting patterns with specific constraints,to mine the association patterns. We carried out experiments on global climate datasets and found many interesting association patterns. Some of the patterns are coincident with known knowledge in climate science,which indicates the correctness and feasibilities of our methods,and the others are unknown to us before,which will give new information to this research field.

DOI

[2]
Xue C J, Dong Q, Fan X.Spatiotemporal association patterns of multiple parameters in the northwestern Pacific Ocean and their relationships with ENSO[J]. International Journal of Remote Sensing, 2014,35(11-12):4467-4483.Not Available

DOI

[3]
Yang J, Gong P, Fu R, et al.The role of satellite remote sensing in climate change studies[J]. Nature Climate Change, 2013,3(10):875-883.Satellite remote sensing has provided major advances in understanding the climate system and its changes, by quantifying processes and spatio-temporal states of the atmosphere, land and oceans. In this Review, we highlight some important discoveries about the climate system that have not been detected by climate models and conventional observations; for example, the spatial pattern of sea-level rise and the cooling effects of increased stratospheric aerosols. New insights are made feasible by the unparalleled global-and fine-scale spatial coverage of satellite observations. Nevertheless, the short duration of observation series and their uncertainties still pose challenges for capturing the robust long-term trends of many climate variables. We point out the need for future work and future systems to make better use of remote sensing in climate change studies.

DOI

[4]
Cheng T, Li Z.A multiscale approach for spatio-temporal outlier detection[J]. Transactions in GIS, 2006,10(2):253-263.Abstract A spatial outlier is a spatially referenced object whose thematic attribute values are significantly different from those of other spatially referenced objects in its spatial neighborhood. It represents an object that is significantly different from its neighbourhoods even though it may not be significantly different from the entire population. Here we extend this concept to the spatio-temporal domain and define a spatial-temporal outlier (ST-outlier) to be a spatial-temporal object whose thematic attribute values are significantly different from those of other spatially and temporally referenced objects in its spatial or/and temporal neighbourhoods. Identification of ST-outliers can lead to the discovery of unexpected, interesting, and implicit knowledge, such as local instability or deformation. Many methods have been recently proposed to detect spatial outliers, but how to detect the temporal outliers or spatial-temporal outliers has been seldom discussed. In this paper we propose a multiscale approach to detect ST-outliers by evaluating the change between consecutive spatial and temporal scales. A four-step procedure consisting of classification, aggregation, comparison and verification is put forward to address the semantic and dynamic properties of geographic phenomena for ST-outlier detection. The effectiveness of the approach is illustrated by a practical coastal geomorphic study.

DOI

[5]
Lu C T, Kou Y, Zhao J, et al.Detecting and tracking regional outliers in meteorological data[J]. Information Sciences, 2007,177(7):1609-1632.Detecting spatial outliers can help identify significant anomalies in spatial data sequences. In the field of meteorological data processing, spatial outliers are frequently associated with natural disasters such as tornadoes and hurricanes. Previous studies on spatial outliers mainly focused on identifying single location points over a static data frame. In this paper, we propose and implement a systematic methodology to detect and track regional outliers in a sequence of meteorological data frames. First, a wavelet transformation such as the Mexican Hat or Morlet is used to filter noise and enhance the data variation. Second, an image segmentation method,-connected segmentation, is employed to identify the outlier regions. Finally, a regression technique is applied to track the center movement of the outlying regions for consecutive frames. In addition, we conducted experimental evaluations using real-world meteorological data and events such as Hurricane Isabel to demonstrate the effectiveness of our proposed approach.

DOI

[6]
Kut A, Birant D.Spatio-temporal outlier detection in large databases[J]. Journal of Computing and Information Technology, 2006,14(4):291-297.Outlier detection is one of the major data mining methods. This paper proposes a three-step approach to detect spatio-temporal outliers in large databases. These steps are clustering, checking spatial neighbors, and checking temporal neighbors. In this paper, we introduce a new outlier detection algorithm to find small groups of data objects that are exceptional when compared with rest large amount of data. In contrast to the existing outlier detection algorithms, new algorithm has the ability of discovering outliers according to the non-spatial, spatial and temporal values of the objects. In order to demonstrate the new algorithm, this paper also presents an example application using a data warehouse

DOI

[7]
Das M, Parthasarathy S.Anomaly detection and spatio-temporal analysis of global climate system[C]. Proceedings of the Third International Workshop on Knowledge Discovery from Sensor Data, ACM, 2009:142-150.

[8]
Benezeth Y, Jodoin P M, Saligrama V, et al.Abnormal events detection based on spatio-temporal co-occurences[C]. IEEE Conference on Computer Vision and Pattern Recognition, 2009:2458-2465.

[9]
Coles S, Bawa J, Trenner L, et al.An introduction to statistical modeling of extreme values[M]. London: Springer, 2001.

[10]
Yu H, Zhang L, Dauwels J.Spatio-temporal graphical models for extreme events[C]. 2014 IEEE International Symposium on Information Theory (ISIT), 2014:2032-2036.

[11]
Huang Y, Zhang L, Zhang P.A framework for mining sequential patterns from spatio-temporal event data sets[J]. IEEE Transactions on Knowledge and Data Engineering, 2008,20(4):433-448.Given a large spatio-temporal database of events, where each event consists of the following fields: event-ID, time, location, event-type, mining spatio-temporal sequential patterns is to identify significant event type sequences. Such spatio-temporal sequential patterns are crucial to investigate spatial and temporal evolutions of phenomena in many application domains. Recent literatures have explored the sequential patterns on transaction data and trajectory analysis on moving objects. However, these methods can not be directly applied to mining sequential patterns from a large number of spatio-temporal events. Two major research challenges are still remaining: (i) the definition of significance measures for spatio-temporal sequential patterns to avoid spurious ones; (ii) the algorithmic design under the significance measures which may not guarantee the downward closure property. In this paper, we propose a sequence index as the significance measure for spatio-temporal sequential patterns, which is meaningful due to its interpretability using spatial statistics. We propose a novel algorithm called Slicing-STS-Miner to tackle the algorithmic design challenges using the spatial sequence index which does not preserve the downward closure property. We compare the proposed algorithm with a simple algorithm called STS-Miner that utilizes the weak monotone property of the sequence index. Performance evaluations using both synthetic and real world datasets shows that the Slicing-STS-Miner is an order of magnitude faster than STS-Miner for large datasets.

DOI

[12]
Tan P, Steinbach M, Kumar V, et al.Finding spatio-temporal patterns in earth science data[C]. KDD 2001 Workshop on Temporal Data Mining, 2001:19.

[13]
Xue C, Dong Q, Xie J.Marine spatio-temporal process semantics and its applications-taking the El Niño Southern Oscilation process and Chinese rainfall anomaly as an example[J]. Acta Oceanologica Sinica, 2012,31(2):16-24.

DOI

[14]
Zhang P, Steinbach M, Kumar V, et al.Discovery of patterns of earth science data using data mining[J]. New Generation of Data Mining Applications, 2005.Publication » Discovery of Patterns of Earth Science Data Using Data Mining.

[15]
Xue C, Song W, Qin L, et al.A spatiotemporal mining framework for abnormal association patterns in marine environments with a time series of remote sensing images[J]. International Journal of Applied Earth Observation and Geoinformation, 2015,38:105-114.A spatiotemporal mining framework is a novel tool for the analysis of marine association patterns using multiple remote sensing images. From data pretreatment, to algorithm design, to association rule mining and pattern visualization, this paper outlines a spatiotemporal mining framework for abnormal association patterns in marine environments, including pixel-based and object-based mining models. Within this framework, some key issues are also addressed. In the data pretreatment phase, we propose an algorithm for extracting abnormal objects or pixels over marine surfaces, and construct a mining transaction table with object-based and pixel-based strategies. In the mining algorithm phase, a recursion method to construct a direct association pattern tree is addressed with an asymmetric mutual information table, and a recursive mining algorithm to find frequent items. In the knowledge visualization phase, a “ Dimension–Attributes ” visualization framework is used to display spatiotemporal association patterns. Finally, spatiotemporal association patterns for marine environmental parameters in the Pacific Ocean are identified, and the results prove the effectiveness and the efficiency of the proposed mining framework.

DOI

[16]
McGuire M P, Janeja V P, Gangopadhyay A. Mining trajectories of moving dynamic spatio-temporal regions in sensor datasets[J]. Data Mining and Knowledge Discovery, 2014,28(4):961-1003.When mining large spatio-temporal datasets, interesting patterns typically emerge where the dataset is most dynamic. These dynamic regions can be characterized by a location or set of locations that exhibit different behaviors from their neighbors and the time periods where these differences are most pronounced. Examples include locally intense areas of precipitation, anomalous sea surface temperature (SST) readings, and locally high levels of water pollution, to name a few. The focus of this paper is to find and analyze the pattern of moving dynamic spatio-temporal regions in large sensor datasets. The approach presented in this paper uses a measure of local spatial autocorrelation over time to determine how pronounced the difference in measurements taken at a spatial location is with those taken at neighboring locations. Dynamic regions are analyzed both globally, in the form of spatial locations and time periods that have the largest difference in local spatial autocorrelation, and locally, in the form of dynamic spatial locations for a particular time period or dynamic time periods for a particular spatial node. Then, moving dynamic regions are identified by determining the spatio-temporal connectivity, extent, and trajectory for groups of locally dynamic spatial locations whose position has shifted from one time period to the next. The efficacy of the approach is demonstrated on two real-world spatio-temporal datasets (a) NEXRAD precipitation and (b) SST. Promising results were found in discovering highly dynamic regions in these datasets depicting several real environmental phenomenon which are validated as actual events of interest.

DOI

[17]
薛存金,周成虎,苏奋振,等.面向过程的时空数据模型研究[J].测绘学报,2010,39(1):95-101.根据近20年来时空数据模型的 研究现状、存在问题及其原因剖析,以连续渐变地理实体的表达、组织和存储为研究对象,提出面向过程的时空数据模型。根据连续渐变地理实体的内在特性,将其 分级抽象为过程对象系列,进一步探讨过程对象及过程对象间逻辑关系,并设计其UML模型结构及物理存储结构。通过抽象的过程对象隐式地记录地理实体动态变 化机制,及自定义的过程对象存储表提供演变机制的函数接口模式,实现连续渐变地理实体的过程化组织、存储与动态分析。最后,以海洋数据的过程化组织与分析 为例,构建时空过程模型原型系统(海洋过程对象—关系数据库系统与功能分析平台),验证和评价该模型的实用性。

[ Xue C J, Zhou C H, Su F Z, et al.Research on process-oriented spatio-temporal data model[J]. Acta Geodaetica et Cartographica Sinica, 2010,39(1):95-101. ]

[18]
薛存金,董庆.海洋时空过程数据模型及其原型系统构建研究[J].海洋通报,2012,31(6):667-674.从地球信息科学角度，分析了海洋现象的过程特性，并给出过程对象的BNF范式组织结构。把具 有过程特性的海洋现象分级抽象为海洋过程对象、阶段对象、序列对象和状态对象，讨论了海洋过程对象的语义分级抽象与顺序、包含关系及过程对象内的演变机 制，分析了通过过程对象隐式记录演变机制、过程对象存储表提供其程序接口，实现连续渐变表达机制在模型中的实现策略，发展了基于过程对象、过程对象关系表 的模型的过程化组织结构。在Geodatabase9．2基础上，扩展过程对象的存储机制，构建海洋时空过程数据模型的原型系统（海洋时空过程数据库系统 MarineSTPDMGDB和海洋系统功能平台MDMProtoTypeSystem），并给出其框架结构及关键技术实现流程。原型系统的构建及功能分 析结果证明了模型的科学性与适用性，为“数字海洋”战略的实施提供了理论基础。

DOI

[ Xue C J, Dong Q.Research on the marine spatio-temporal process data model and its prototype system construction[J]. Marine Science Bulletin, 2012,31(6):667-674. ]

[19]
McPhaden M J, Zebiak S E, Glantz M H. ENSO as an integrating concept in earth science[J]. Science, 2006,314(5806):1740-1745.Abstract The El Ni09o-Southern Oscillation (ENSO) cycle of alternating warm El Ni09o and cold La Ni09a events is the dominant year-to-year climate signal on Earth. ENSO originates in the tropical Pacific through interactions between the ocean and the atmosphere, but its environmental and socioeconomic impacts are felt worldwide. Spurred on by the powerful 1997-1998 El Ni09o, efforts to understand the causes and consequences of ENSO have greatly expanded in the past few years. These efforts reveal the breadth of ENSO's influence on the Earth system and the potential to exploit its predictability for societal benefit. However, many intertwined issues regarding ENSO dynamics, impacts, forecasting, and applications remain unresolved. Research to address these issues will not only lead to progress across a broad range of scientific disciplines but also provide an opportunity to educate the public and policy makers about the importance of climate variability and change in the modern world.

DOI PMID

[20]
王绍武,龚道溢.近百年来的 ENSO 事件及其强度[J].气象,1999,25(1):9-13.利用Ｎｉｎｏ３区、ＮｉｎｏＣ区海表温度序列及两个ＳＯＩ序列，同时考虑ＳＳＴ和ＳＯＩ建立了１８６７～１９９８年季分辨率的ＥＮＳＯ指数序列。根据ＥＮＳＯ指数序列，并参考Ｗｒｉｇｈｔ的ＳＯＩ指数及其它资料。确认了１８６７～１９９８年ＥＮＳＯ事件，共确定出３２次暖事件（正ＳＳＴ、负ＳＯＩ）及３２次冷事件（负ＳＳＴ、正ＳＯＩ）。对每次事件的强度分强、中、弱三等进行了评估。虽然１９８２／１９８３年暖事件的峰值最高，但从整个事件的平均强度来看，１９９７／１９９８年的暖事件则是１３０多年来最强的一次。近２０年是暖事件的多发期。

DOI

[ Wang S W, Gong D Y.ENSO events and their Intensity during the past century[J]. Meteorological Monthly, 1999,25(1):9-13. ]

[21]
Wolter K, Timlin M S.El Niño/southern oscillation behaviour since 1871 as diagnosed in an extended multivariate ENSO index (MEI. ext)[J]. International Journal of Climatology, 2011,31(7):1074-1087.El Nino/Southern Oscillation (ENSO) remains the most important coupled ocean-atmosphere phenomenon to cause global climate variability on seasonal to interannual time scales. This paper addresses the need for a reliable ENSO index that allows for the historical definition of ENSO events in the instrumental record back to 1871. The Multivariate ENSO Index (MEI) was originally defined as the first seasonally varying principal component of six atmosphere-ocean (COADS) variable fields in the tropical Pacific basin. It provides for a more complete and flexible description of the ENSO phenomenon than single variable ENSO indices such as the SOI or Nino 3.4 SST. Here we describe our effort to boil the MEI concept down to its most essential components (based on SLP, SST) to enable historical analyses that more than double its period of record to 1871-2005. The new MEI. ext confirms that ENSO activity went through a lull in the early-to mid-20th century, but was just about as prevalent one century ago as in recent decades. We diagnose strong relationships between peak amplitudes of ENSO events and their duration, as well as between their peak amplitudes and their spacing (periodicity). Our effort is designed to help with the assessment of ENSO conditions through as long a record as possible to be able to differentiate between 'natural' ENSO behaviour in all its rich facets, and the 'Brave New World' of this phenomenon under evolving GHG-related climate conditions. So far, none of the behaviour of recent ENSO events appears unprecedented, including duration, onset timing, and spacing in the last few decades compared to a full century before then. Copyright (C) 2011 Royal Meteorological Society

DOI

[22]
Casey K S, Adamec D.Sea surface temperature and sea surface height variability in the North Pacific Ocean from 1993 to 1999[J]. Journal of Geophysical Research: Oceans (1978-2012), 2002,107(C8):1-12.1] Correlations between advanced very high resolution radiometer Pathfinder sea surface temperature (SST) and TOPEX sea surface height (SSH) anomalies are calculated for the 7 year period 1993鈥1999 in the equatorial and North Pacific. These correlations provide measures of SSH and SST covariability and are obtained using an empirical orthogonal function analysis separately on the two fields and a singular value decomposition (SVD) method on both fields simultaneously. These correlations reveal distinct SST-SSH states: a warm, elevated period from January 1993 to December 1994; a cooler, lower phase spanning January 1995 through December 1996; and the strongly variable El Ni帽o-Southern Oscillation (ENSO) period spanning 1997鈥1999. The relationships between these SST-SSH states and the dominant interannual and interdecadal modes of Pacific variability are investigated through correlations with indices for the Southern Oscillation, Pacific-North American (PNA) pattern, and Pacific Decadal Oscillation (PDO). When including the tropics, significant correlations between the first SVD mode of the SST-SSH covariability and the interannual ENSO pattern and interdecadal PDO pattern are evident. Restricting the focus to the extratropical North Pacific reveals significant correlations with only the interdecadal PDO pattern. While the first mode reveals connections to the ENSO and PDO patterns, the second SVD mode indicates significant correlations to the PNA pattern for both the tropical and extratropical Pacific.

DOI

Options
Outlines

/

〈 〉