Machine Learning System for Rainfall Estimates from Single Polarization Radar

Tinar Pamuji Waskita, Adhi Harmoko Saputro, Ardhasena Sopaheluwakan, Muhammad Ryan

2019

Abstract

Rainfall becomes one of the weather parameters that is most widely considered because the phenomenon of its occurrence can significantly affect human activities, including in agriculture, plantations, fisheries, transportation and others. In addition, rainfall information is very important to do weather analysis, especially in analyzing the occurrence of floods caused by heavy rains so there is a need for accurate rainfall information. This study aims to obtain an optimal rainfall estimation system at locations where there is no direct rainfall observation data. Machine learning is one branch of artificial intelligence that provides a learning system for machines to learn automatically without explicit instruction. The machine learning used in this study is Multi layer perceptron (MLP), with backpropagation as a gradient value search algorithm and adam optimizer as an optimization function. The structure of the MLP used is 2 hidden layers which in the first hidden layer uses 7 neurons with a hyperbolic tangent activation function and the second hidden layer contains 3 neurons and the activation function is sigmoid and finally the output layer, the activation function used is pure linear. MLP system input data is radar data, reflectivity, radial velocity, spectrum width and radar rain estimation data which are validated with automatic rain observation data around the Single Polarization Radar observation in Yogyakarta. The results using MLP can improve rain detection accuracy by 79% and reduce the error value in the estimated rainfall.

Paper Citation

in Harvard Style

Pamuji Waskita T., Harmoko Saputro A., Sopaheluwakan A. and Ryan M. (2019). Machine Learning System for Rainfall Estimates from Single Polarization Radar.In Proceedings of the International Conferences on Information System and Technology - Volume 1: CONRIST, ISBN 978-989-758-453-4, pages 41-48. DOI: 10.5220/0009409400410048

in Bibtex Style

@conference{conrist19,
author={Tinar Pamuji Waskita and Adhi Harmoko Saputro and Ardhasena Sopaheluwakan and Muhammad Ryan},
title={Machine Learning System for Rainfall Estimates from Single Polarization Radar},
booktitle={Proceedings of the International Conferences on Information System and Technology - Volume 1: CONRIST,},
year={2019},
pages={41-48},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0009409400410048},
isbn={978-989-758-453-4},
}

in EndNote Style

TY - CONF

JO - Proceedings of the International Conferences on Information System and Technology - Volume 1: CONRIST,
TI - Machine Learning System for Rainfall Estimates from Single Polarization Radar
SN - 978-989-758-453-4
AU - Pamuji Waskita T.
AU - Harmoko Saputro A.
AU - Sopaheluwakan A.
AU - Ryan M.
PY - 2019
SP - 41
EP - 48
DO - 10.5220/0009409400410048