Please use this identifier to cite or link to this item: http://repository.ipb.ac.id/handle/123456789/158067
Title: Implementasi Algoritma Generative Adversarial Network (GAN) untuk Prediksi Sebaran Konsentrasi Klorofil-a di Perairan Pesisir Barat Sumatera
Other Titles: Implementation of Generative Adversarial Network (GAN) Algorithm for Prediction of Chlorophyll-a Concentration Distribution in West Coastal Waters of Sumatra
Authors: Jaya, Indra
Iqbal, Muhammad
Zul'ilmiwattaqwa, Amira
Issue Date: 2024
Publisher: IPB University
Abstract: Variabilitas konsentrasi klorofil-a di Perairan Pesisir Barat Sumatera dipengaruhi oleh pola angin musiman dan fenomena anomali iklim Indian Ocean Dipole (IOD). Fase IOD positif terjadi peningkatan konsentrasi klorofil-a dan fase IOD negatif terjadi penurunan konsentrasi klorofil-a. Teknik penginderaan jauh “Ocean Color” untuk pendugaan konsentrasi klorofil-a sebagai parameter tingkat kesuburan suatu perairan memiliki beberapa kelemahan, seperti terhalang oleh awan, kedalaman suatu perairan, dan pantulan gelombang laut yang mengaburkan citra. Oleh karena itu, metode deep learning dengan Generative Adversarial Network (GAN), khususnya Pix2PixHD dapat menjadi alternatif untuk prediksi sebaran klorofil-a. Pelatihan model menggunakan dataset bulanan dan harian, masing-masing berisi 348 dan 731 gambar plot sebaran klorofil-a. Model ini memprediksi konsentrasi klorofil-a untuk tahun 2022 secara bulanan dan tujuh hari kedepan untuk tahun 2024 secara harian. Hasil pelatihan menunjukkan model dapat memprediksi hingga empat hari ke depan dalam periode harian berdasarkan nilai koefisien korelasi pada regresi linear yang mencapai 0,7. Namun, akurasi prediksi model akan menurun seiring waktu.
Variability of chlorophyll-a concentration in the West Coastal Waters of Sumatra is affected by seasonal wind patterns and the Indian Ocean Dipole (IOD) climate anomaly phenomenon. Positive IOD phases increase chlorophyll-a concentrations and negative IOD phases decrease chlorophyll-a concentrations. The “Ocean Color” remote sensing technique for estimating chlorophyll-a concentration as a parameter of the water quality has several weaknesses, such as cloud obstruction, water depth, and wave reflections that obscure the image. Therefore, deep learning methods with Generative Adversarial Network (GAN), especially Pix2PixHD can be an alternative for predicting chlorophyll-a distribution. The model was trained using 348 chlorophyll-a plot images for monthly and 731 chlorophyll-a plot images for daily datasets. The model predicts chlorophyll-a concentrations for 2022 monthly and seven days ahead for 2024 daily. The training results show that the model can predict up to four days ahead in the daily period based on the correlation coefficient value in the linear regression that reaches 0.7. However, the prediction accuracy of the model will decrease over time.
URI: http://repository.ipb.ac.id/handle/123456789/158067
Appears in Collections:UT - Marine Science And Technology

Files in This Item:
File Description SizeFormat 
cover_C5401201028_20d671ee863e44deb8c4488a3051237a.pdfCover333.88 kBAdobe PDFView/Open
fulltext_C5401201028_0049de484e7045d1bb464f61386b39fe.pdf
  Restricted Access
Fulltext2.4 MBAdobe PDFView/Open
lampiran_C5401201028_352cfea516914ef68309c846be9a2b27.pdf
  Restricted Access
Lampiran237.49 kBAdobe PDFView/Open


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.