Ramdani, Wira Satria (2023) Analisa Deteksi dan Prediksi Titik Api dengan Menggunakan Metode LSTN di Provinsi Riau. Other thesis, Universitas Islam Riau.

Text 183510052.pdf - Submitted Version Restricted to Registered users only Download (3MB) | Request a copy

Abstract

Global warming is increasing the earth's temperature, further increasing the number of hotspots in some forest areas, especially in tropical areas where there is a high risk of forest fires and wildlife. Indonesia is one of the countries in Southeast Asia that has experienced a large number of forest fires which have had a dangerous impact on neighboring countries due to haze and carbon emissions. This study aims to plot and identify locations with a high number of hotspots and then predict the potential number of hotspots in the future based on previously collected historical data. The detection data obtained is very important and useful for those in power because it can be used as a reference for preventive measures and to prevent the spread of forest fires. The long shot-term memory (LSTM) algorithm is implemented in this study to detect and predict the number of hotspots, while Python programming is used to plot hotspots. The hotspot dataset was obtained from the National Aeronautics and Space Administration (NASA) Moderate Resolution Imaging Spectroradiometer (MODIS) which was recorded from 2017 to 2022, recording a total of around 101,000 hotspots in Riau Province. To prove that the proposed algorithm works well, simulations have been carried out using training data from 2017 to 2022 and data testing from 2022 to 2023, then forecasting results are obtained with the same pattern of number of hotspots compared to data available in 2022.

Actions (login required)

View Item