Model Peramalan Muka Air Tanah Pada Lahan Gambut Menggunakan Pendekatan Artificial Neural Network (ANN) (Studi Kasus Kabupaten Bengkalis Provinsi Riau)

Fajri Rahmatullah, Imam Suprayogi, Ari Sandhyavitri


The Artificial Neural Network (ANN) method is a computationally soft method that can predict water levels, one of them being for early warning of peatland fires. It is necessary to prove the performance of Artificial Neural Network (ANN) to predict the water level in Bengkalis Island, Riau Province. In this research, we use observation data using HOBOwater Logger level for 2014 data. Data usage was conducted trought time series of 1 day period with the maximum value, test and validation with Artificial Neural Network (ANN) Model is Backpropagation Algorithm. The result of simulation of water level in Artificial Neural Network (ANN) method in MATLAB software, yields statistical value of correlation coefficient (R) with very strong category. In the time series scheme the period of 1 day in maximum time to 24 hours (H1t + 24) produces correlation coefficient (R) of 0.995929. Keywords : peatlands, HOBOwater level logger, ANN Method, MATLAB, forecasting.

Full Text:



  • There are currently no refbacks.