Estimasi Curah Hujan Kota Banjarbaru Kalimantan Selatan Menggunakan Metode Jaringan Syaraf Tiruan

Abstract

Rain is weather phenomenon caused by climate’s physical conditions
of atmosphere variables such as temperature, air pressure, air density, humidity
and wind of velocity. Many researches on rain have been undertaken in
Indonesia. Many statistic models are resulted from those researches;
nevertheless, in statistic models climate’s physical conditions are not considered
as components which affect the rain occurrence. Besides, estimating the amount
of rain that is to fall-whether it is increasing, decreasing, or static- with statistic
models is still unsure. Physical and mathematical approaches of weather
variables only are not enough to estimate rainfall. More interpretations are
needed to apply. Therefore, this research implements a way to estimate daily
rainfall using a method called Artificial Neural Network. As parameter of the
inputs of JST program, then the data of daily temperature, air humidity and wind
speed play important roles, while the output which is manifested in daily rainfall
tested in data scale, a period of 5-9 years. The results obtained show that JST
method enables us to estimate daily rainfall of Kota Banjarbaru.