The Analysis of Tropical Cyclone Tracks in the Western North Pacific through Data Mining. Part II: Tropical Cyclone Landfall

Abstract

This is the second paper of a two-part series of papers on the analysis of tropical cyclone (TC) tracks in the western North Pacific Ocean. In this paper, TC landfalls in the South China Sea and western North Pacific basins are investigated through the data-mining approach. On the basis of historical TC archives, the C4.5 algorithm, a classic tree algorithm for classification, has been employed to quantitatively discover rules governing TC landfall. A classification tree, with 14 leaf nodes, has been built. The path from the root node to each leaf node forms a rule. Fourteen rules governing TC landfall across the Chinese coast have been unraveled with respect to the selected attributes having potential influence on TC landfall. The rules are derived by the attributes and splitting values. From the classification tree, split values, such as 27°N latitude, 130°E longitude, 141°E in the west extension index, and 0.289 in the monsoon index have been shown to be useful for TC forecasting. The rules have been justified from the perspective of meteorology and knowledge of TC movement and recurvature (e.g., deep-layer mean winds and large-scale circulation). The research findings are also consistent with existing results concerning TC movement and landfall. Both the unraveled rules and the associated splitting values can provide useful references for the prediction of TC landfall over China. .J structure, mtenstty, and motion of the TC will change two-part paper, troptcal cyclone (TC) recurvature and ,

A Systematic Classification Investigation of Rapid Intensification of Atlantic Tropical Cyclones with the SHIPS Database

“…The most widely used feature selection method is based on information theory such as information gain and gain ratio (Tan et al 2006;Li et al 2009;Zhang et al 2013a). The information gain is defined as the decrease in entropy (impurity) of the classes after discretizing values of an attribute.…”

“…Identifying new predictors requires mining among high-dimensional spatiotemporal datasets in their original formats and is difficult. Therefore, most of the TC-related data mining work follows the classical data structure and attribute tables as in relational databases (Yang et al 2007(Yang et al , 2008(Yang et al , 2011Li et al 2009;Zhang et al 2013a). Data in SHIPS databases essentially follow the same data structure.…”

“…Special treatment, such as unbalanced samples and other data mining techniques such as feature selection, can also be easily performed in WEKA. Examples of using WEKA in meteorology studies include the multivariate time series data mining application to severe weather prediction performed by Gagne et al (2009), estimation of rainfall rate from radar data with an artificial neural network by Root et al (2010), and movement of western North Pacific TCs with the classification and regression tree method by Zhang et al (2013b). The version of WEKA used for this work is WEKA 3.6.9 in a Windows desktop environment.…”

“…One way to decrease the complexities of the classification is to reduce the number of features (parameters or attributes). This technique is actually used in most studies of TC intensity with statistical methods or data mining (KD03; Rozoff and Kossin 2011;Li et al 2009;Yang et al 2007Yang et al , 2008Yang et al , 2011Zhang et al 2013a). In those studies, the commonly used scheme for feature selection is the t test with two means, means of RI cases and non-RI cases for RI studies.…”

Climate patterns of political division units obtained using automatic classification trees

“…It is more robust than simple percent agreement (between the machine classifier and the reality) because it takes into account the agreement occurring by chance. Other measurements, like Zhang et al, 2013b) that classifies a cyclon into 0 ("does not make a landfall along the Chinese coast") or 1 ("makes a landfall…"). These numbers are seen to the left of the left parenthesis in each leaf node (rectangles in the tree).…”