PUTRI, MUTIARA ADHISTYA (2026) PREDIKSI SPATIOTEMPORAL KONSENTRASI KARBON MONOKSIDA BERBASIS MACHINE LEARNING. S1 thesis, Universitas Mercu Buana Jakarta.
|
Text (HAL COVER)
Cover.pdf Download (401kB) | Preview |
|
![[img]](https://repository.mercubuana.ac.id/style/images/fileicons/text.png) |
Text (BAB I)
Bab 1.pdf Restricted to Registered users only Download (43kB) |
|
|
Text (BAB II)
Bab 2.pdf Restricted to Registered users only Download (259kB) |
|
|
Text (BAB III)
Bab 3.pdf Restricted to Registered users only Download (475kB) |
|
|
Text (BAB IV)
Bab 4.pdf Restricted to Registered users only Download (893kB) |
|
|
Text (BAB V)
Bab 5.pdf Restricted to Registered users only Download (27kB) |
|
|
Text (DAFTAR PUSTAKA)
Daftar Pustaka.pdf Restricted to Registered users only Download (97kB) |
|
|
Text (LAMPIRAN)
Lampiran.pdf Restricted to Registered users only Download (525kB) |
Abstract
Forest and land fires in the Kalimantan region have led to increased concentrations of carbon monoxide (CO) in the atmosphere, adversely affecting air quality and public health. This study aims to develop a spatiotemporal prediction model for CO concentrations using machine learning approaches. The dataset was obtained through Google Earth Engine, utilizing Sentinel-5P TROPOMI data to measure CO concentrations, MODIS data to identify hotspots and land surface temperature (LST), CHIRPS data to estimate precipitation, and ERA5-Land data to measure wind speed during the period from January 2019 to December 2025. The research methodology consisted of data preprocessing, feature engineering, relevant feature selection, the construction of prediction horizons for one-day�ahead (H+1) and seven-day-ahead (H+7) forecasting, model development using the Random Forest and XGBoost algorithms, and model performance evaluation using the Root Mean Square Error (RMSE), Mean Absolute Error (MAE), and coefficient of determination (R²) metrics. The results demonstrate that both algorithms were able to accurately predict CO concentrations, with the best�performing model selected based on the evaluation results. This study is expected to provide reliable information for predicting the spatial and temporal distribution of carbon monoxide (CO) concentrations across the Kalimantan region, thereby serving as a valuable reference for air quality monitoring and research on forest and land fires. Keywords: Carbon Monoxide, Machine Learning, Random Forest, XGBoost, Spatiotemporal, Forest and Land Fires. Kebakaran hutan dan lahan di wilayah Kalimantan menyebabkan meningkatnya kadar karbon monoksida (CO) di udara, yang berdampak negatif terhadap kualitas udara serta kesehatan masyarakat. Penelitian ini bertujuan mengembangkan model prediksi konsentrasi CO yang bersifat spatiotemporal menggunakan pendekatan machine learning. Data tersebut diperoleh menggunakan Google Earth Engine dengan sumber data Sentinel-5P TROPOMI untuk mengukur konsentrasi CO, MODIS untuk mengidentifikasi titik panas dan suhu permukaan daratan (LST), CHIRPS untuk menghitung curah hujan, serta ERA5-Land untuk mengukur kecepatan angin selama periode Januari 2019 hingga Desember 2025. Tahapan dalam penelitian mencakup proses pra pemrosesan data, pembuatan fitur, pemilihan fitur yang relevan, pembentukan horizon prediksi untuk H+1 dan H+7, pembuatan model menggunakan metode Random Forest dan XGBoost, serta penilaian kinerja model dengan menggunakan metrik RMSE, MAE, dan R². Penelitian menunjukkan bahwa kedua algoritma berhasil memprediksi kadar CO secara akurat, dan model yang paling baik dipilih berdasarkan hasil pengujian yang dilakukan. Penelitian ini diharapkan dapat menyediakan informasi yang memprediksi distribusi spasial dan temporal konsentrasi karbon monoksida (CO) di wilayah Kalimantan, sehingga dapat dijadikan acuan dalam pemantauan kualitas udara serta penelitian mengenai kebakaran hutan dan lahan. Kata kunci: Karbon Monoksida, Machine Learning, Random Forest, XGBoost, Spatiotemporal, Kebakaran Hutan dan Lahan
Actions (login required)
 |
View Item |
