Pemetaan Chl-A Perairan Lima Danau Prioritas Di Indonesia Menggunakan Regresi Linier (Study Kasus : Danau Toba, Maninjau, Singkarak, Matano, Dan Towuti)

Pangestu, Martanti Aji (2023) Pemetaan Chl-A Perairan Lima Danau Prioritas Di Indonesia Menggunakan Regresi Linier (Study Kasus : Danau Toba, Maninjau, Singkarak, Matano, Dan Towuti). Masters thesis, Institut Teknologi Sepuluh Nopember.

Text 6016221006-Master_thesis.pdf - Accepted Version Restricted to Repository staff only until 1 September 2025. Download (3MB) | Request a copy

Abstract

Berkembangnya teknologi menuntut manusia untuk melakukan segala hal dengan cepat. Salah satunya adalah pemantauan kesuburan perairan, utamanya yaitu danau yang sesuai dengan RPJMN 2020-2024. Dari ke-15 danau prioritas yang ada, penelitian ini menggunakan 5 danau prioritas yaitu Danau Toba, Maninjau, Singkarak, Matano, dan Towuti. Pengembangan algoritma model adalah jawaban dalam mempercepat pemantauan kesuburan danau melalui pemantauan konsentrasi Chl-a. Penelitian ini bertujuan untuk mengevaluasi, mengembangkan, serta mencari algoritma model untuk pemetaan konsentrasi klorofil-a (Chl-a). Pencarian algoritma dapat dilakukan dengan memasukkan hasil extract pixel value kedalam band ratio algorithm. Kemudian, 22 band ratio dikorelasikan dengan data in-situ. Pemilihan algoritma dilakukan dengan evaluasi model menggunakan validasi Leave One Out Cross Validation (LOOCV), pada evaluasi ini menyajikan hasil terbaik dari model algoritma yang dapat dilihat melalui nilai akurasi R, R2, NMAE, dan RMSE. Menggunakan 42 titik koordinat sampel dari data in-situ pengambilan Chl-a, untuk ekstraksi nilai reflektan citra satelit Landsat 8 dan 9 didapatkan hasil bahwa tidak semua algoritma band ratio yang ada pada penelitian sebelumnya, menjadi algoritma yang sesuai pada studi kasus kelima danau tersebut. Setiap danau memiliki kesesuaian algoritma yang berbeda. Danau Toba di musim kemarau memiliki RMSE = 0,7 mg/m3, NMAE = 22,5%, 0,7, dan R2= 0,3 dan Danau Singkarak memiliki RMSE = 0,2 mg/m3, NMAE = 9%, R=0,7 dan R2= 0,2. yang lebih cocok menggunakan C6, Danau Matano memiliki RMSE = 0,0 mg/m3, NMAE = 4,2%, R=1,0, dan R2= 1 lebih sesuai menggunakan C10, Danau Towuti memiliki RMSE = 0,0 mg/m3, NMAE = 3,4%, R=1,0, dan R2= 1 lebih sesuai menggunakan B3, Danau Toba musim hujan memiliki RMSE = 0,1 mg/m3, NMAE = 7%, R=0,5, dan R2= 0,2 lebih sesuai menggunakan C2, serta Danau Maninjau memiliki RMSE = 0,5 mg/m3, NMAE = 32,5%, R=0,5dan R2= 0,3 lebih sesuai dengan A2.
=================================================================================================================================
Technology development requires humans to do everything quickly. One of them is monitoring the fertility of the waters, especially the lakes, following the RPJMN 2020-2024. Of the 15 existing priority lakes, this study used 5 priority lakes, namely Lake Toba, Maninjau, Singkarak, Matano, and Towuti. The development of a model algorithm is the answer to accelerating the monitoring of the lake through monitoring chlorophyll-a (Chl-a) concentrations. This study aims to evaluate, develop, and search for model algorithms for mapping Chl-a concentrations. An algorithm search can be done by entering the results of the extracted pixel value into the band ratio algorithm. Then, 22 band ratios were correlated with in-situ data. Algorithm selection is done by evaluating the model using Leave One Out Cross Validation (LOOCV) validation. This evaluation presents the best results from the algorithm model, which can be seen through the accuracy values of R, R2, NMAE, and RMSE. Using 42 sample coordinate points from in-situ data taking Chl-a to extract the reflectance value of Landsat 8 and 9 satellite imagery, it was found that not all band ratio algorithms in previous studies were suitable for the case studies of the five lakes. Each lake has a different suitability algorithm. Lake Toba in the dry season has RMSE = 0.7 mg/m3, NMAE = 22.5%, 0.7, and R2 = 0.3 and Lake Singkarak has RMSE = 0.2 mg/m3, NMAE = 9%, R=0.7 and R2= 0.2. which is more suitable using C6, Lake Matano has RMSE = 0.0 mg/m3, NMAE = 4.2%, R=1.0, and R2= 1 is more suitable using C10, Lake Towuti has RMSE = 0.0 mg/ m3, NMAE = 3.4%, R=1.0, and R2= 1 are more appropriate to use B3, Lake Toba in the rainy season has RMSE = 0.1 mg/m3, NMAE = 7%, R=0.5, and R2 = 0.2 is more suitable for using C2, and Lake Maninjau has RMSE = 0.5 mg/m3, NMAE = 32.5%, R = 0.5 and R2 = 0.3 which is more suitable for A2.

Item Type:	Thesis (Masters)
Uncontrolled Keywords:	Band Ratio Algorithm, Chl-a, Landsat, LOOCV, Model Global; Band Ratio Algorithm, Chl-a, Global Model, Landsat, LOOCV
Subjects:	G Geography. Anthropology. Recreation > G Geography (General) > G70.5.I4 Remote sensing
Divisions:	Faculty of Civil, Planning, and Geo Engineering (CIVPLAN) > Geomatics Engineering > 29101-(S2) Master Thesis
Depositing User:	Martanti Aji Pangestu
Date Deposited:	18 Sep 2023 00:54
Last Modified:	18 Sep 2023 02:08
URI:	http://repository.its.ac.id/id/eprint/104365

Actions (login required)

View Item