Open AccessHeight wave modelling using spatial extreme value with max stable process (MSP) brown-resnick model
Author(s) -
Arief Rachman Hakim,
Budi Warsito,
Hasbi Yasin
Publication year - 2021
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1943/1/012142
Subject(s) - generalized extreme value distribution , extreme value theory , gumbel distribution , mathematics , maxima , weibull distribution , gaussian process , block (permutation group theory) , gaussian , statistics , geometry , physics , art , quantum mechanics , performance art , art history
Analysis of extreme data containing spatial elements can be used the Spatial Extreme Value method, which is applied with a max-stable process approach. The data preprocessing process begins by identifying the extreme data of each location for each time period or block. This method is also called Block maxima. The data obtained from the Block Maxima process will follow the Generalized Extreme Value (GEV) distribution. GEV has three forms of distribution, namely Gumbel, Frechet and Weibull. In this study, the data will be transformed into the frechet distribution margin because the frechet has a heavy tail shape, this is required in MSP modeling. Smith model, Schlater model, and Brown Resnik model are the three main models in MSP. Brown technical model is used in this research because Brown Resnick form is a flexible ta of stationary max stable processes in Gaussian random fields. The best model is selecting based on the smallest TIC value from all combinations of Trend surface models. The selected model is then used to determine the prediction of the extreme wave height for each location with a certain time period. The location used in this study is a point where fishermen are widely used to find fish, namely Semarang, Pekalongan, and Rembang which are in Central Java Province.