Rainfall Prediction Based on Himawari-8 IR Enhanced Image Using Backpropagation
Abstract
The wider sea area causes greater evaporation of water in Indonesia. In addition, these conditions have an impact on the season that Indonesia has. Indonesia’s high rainfall disrupts human activities. As a result, it is very important to detect cumulonimbus clouds using satellite imagery. The satellite image used is intended to be taken two values of the characteristics possessed. Characteristics taken are average cover and average cloud temperature. Previous studies predicting rain were only done using observational data taken at the height of 10 meters. This research predicts using satellite imagery that represents the cloud peak temperature value. Furthermore, the classification of data is done using backpropagation. The results of the classification process using backpropagation obtained the best results on the distribution of 80% training data and 20% testing data, with the activation function logging in the hidden layer and that the output layer. The results obtained indicate the accuration rate of 88,283%.
- Publication:
-
Journal of Physics Conference Series
- Pub Date:
- March 2020
- DOI:
- 10.1088/1742-6596/1501/1/012011
- Bibcode:
- 2020JPhCS1501a2011N