Model Prakiraan Cuaca Maritim Menggunakan Copula Bayesian Network

Wita, Feby Fara (2024) Model Prakiraan Cuaca Maritim Menggunakan Copula Bayesian Network. Diploma thesis, Institut Teknologi Sepuluh Nopember.

Text 2043211088_Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 April 2027. Download (3MB) | Request a copy

Abstract

Perubahan cuaca yang tidak menentu, seperti badai tropis, gelombang tinggi, dan perubahan arah angin, menjadi salah satu penyebab utama kecelakaan pelayaran. Dalam dekade terakhir, kecelakaan pelayaran didominasi oleh kesalahan teknis dan manusia, yang sering kali dipengaruhi oleh kondisi cuaca maritim yang buruk. Cuaca ekstrem dapat mengganggu peralatan navigasi dan sistem komunikasi, serta memengaruhi pengambilan keputusan oleh awak kapal. Oleh karena itu, informasi cuaca yang akurat dan tepat waktu sangat penting untuk mendukung perencanaan rute pelayaran yang lebih aman. Penelitian ini menggunakan Copula Bayesian Network (CBN) data yang dianalisis mencakup berbagai parameter cuaca dari seluruh perairan Indonesia. Dari berbagai jenis Copula yang diuji, Copula Gaussian terbukti sebagai yang terbaik dalam memodelkan ketergantungan antar variabel cuaca. Hasil penelitian menunjukkan bahwa beberapa variabel cuaca, seperti komponen horizontal angin pada ketinggian 10 meter, rata-rata tekanan permukaan laut, suhu permukaan laut, rata-rata arah gelombang, rata-rata periode gelombang, total curah hujan, dan kerapatan udara, memiliki pengaruh langsung terhadap tinggi gelombang signifikan. Hasil prediksi tinggi gelombang menunjukkan konsistensi dengan data asli, namun terdapat beberapa perbedaan seperti estimasi berlebih, di mana prediksi model menghasilkan nilai lebih tinggi dibandingkan data asli. Tingkat kesalahan prediksi mencapai 0,277 meter yang cukup baik karena mendekati nilai aktual, 72,231% data aktual berada dalam interval prediksi. Prediksi yang lebih akurat, kesiapan dan respons awak kapal terhadap kondisi cuaca buruk diharapkan dapat ditingkatkan, sehingga risiko kecelakaan pelayaran akibat cuaca ekstrem dapat diminimalkan. Dengan demikian, prakiraan cuaca maritim khususnya tinggi gelombang yang lebih akurat diharapkan dapat meningkatkan kesiapan dan respons awak kapal terhadap kondisi cuaca buruk.
=================================================================================================================================
Erratic weather changes, such as tropical storms, high waves, and changes in wind direction, are one of the main causes of shipping accidents. In the past decade, shipping accidents have been dominated by technical and human errors, which are often affected by poor maritime weather conditions. Extreme weather can disrupt navigation equipment and communication systems, as well as affect decision-making by the crew. Therefore, accurate and timely weather information is essential to support safer shipping route planning. This study uses the Copula Bayesian Network (CBN) data that is analyzed covering various weather parameters from all Indonesian waters. Of the various types of copula tested, Gaussian copula proved to be the best at modeling the dependence between weather variables. The results showed that several weather variables, such as the horizontal component of the wind at an altitude of 10 meters, the average sea surface pressure, the sea surface temperature, the average wave direction, the average wave period, the total rainfall, and the air density, had a direct influence on the significant wave height. The waveheight prediction results show consistency with the original data, but there are some differences such as overestimation, where the model prediction produces a higher value than the original data. The prediction error rate reached 0.277 meters which is quite good because it is close to the actual value, 72.231% of the actual data is in the prediction interval. More accurate predictions, crew readiness and response to adverse weather conditions are expected to be improved, so that the risk of shipping accidents due to extreme weather can be minimized. Thus, maritime weather forecasts, especially more accurate wave heights, are expected to improve the crew's readiness and response to adverse weather conditions.

Item Type:	Thesis (Diploma)
Uncontrolled Keywords:	Copula Bayesian Network, Cuaca Maritim, Copula Bayesian Network, Maritime Weather
Subjects:	Q Science Q Science > QA Mathematics > QA279.5 Bayesian statistical decision theory.
Divisions:	Faculty of Vocational > 49501-Business Statistics
Depositing User:	Feby Fara Wita
Date Deposited:	19 Dec 2024 08:38
Last Modified:	19 Dec 2024 08:38
URI:	http://repository.its.ac.id/id/eprint/116005

Actions (login required)

View Item