Application of Rainfall Prediction Model on Two Rice Production Centers, in West Java

Abstract

<div style="left: 99.2127px; top: 907.281px; font-size: 16.9238px; font-family: serif; transform: scaleX(1.00096);" data-canvas-width="788.525">The primary objective of this study was to develop rainfall prediction model using artificial neural network analysis</div><div style="left: 85.0393px; top: 928.948px; font-size: 16.9238px; font-family: serif; transform: scaleX(1.01114);" data-canvas-width="802.7116666666665">techniques, the second objective was to apply the prediction model in rice production centers, and the third objective</div><div style="left: 85.0393px; top: 950.615px; font-size: 16.9238px; font-family: serif; transform: scaleX(0.984825);" data-canvas-width="802.6816666666666">was to compare the model predictions in rice production centers. Research is a desk study with case study in Indramayu</div><div style="left: 85.0393px; top: 972.281px; font-size: 16.9238px; font-family: serif; transform: scaleX(0.991622);" data-canvas-width="802.7016666666663">and Cianjur districts, West Java. The primary step of this study was collection of rainfall data and map information and</div><div style="left: 85.0393px; top: 993.948px; font-size: 16.9238px; font-family: serif; transform: scaleX(1.00627);" data-canvas-width="802.6666666666667">Climatology of Rainfall Stations in each district using a combination of input SST Anomaly Nino3.4 and DMI, using</div><div style="left: 85.0393px; top: 1015.61px; font-size: 16.9238px; font-family: serif; transform: scaleX(0.979254);" data-canvas-width="802.6783333333336">data from 1990 to 2010, the second step was preparation of rainfall prediction models using network analysis techniques</div><div style="left: 85.0393px; top: 1037.28px; font-size: 16.9238px; font-family: serif; transform: scaleX(0.982665);" data-canvas-width="802.7066666666665">nerve propagation, the third step was validation the model by comparing the output that has been formed with the actual</div><div style="left: 85.0393px; top: 1058.95px; font-size: 16.9238px; font-family: serif; transform: scaleX(1.0111);" data-canvas-width="802.7116666666666">rainfall data, and the fourth step was comparing rainfall prediction with the results of global climate predictions. The</div><div style="left: 85.0393px; top: 1080.61px; font-size: 16.9238px; font-family: serif; transform: scaleX(0.969247);" data-canvas-width="802.6983333333333">results showed that formulation and Validation of the model using input anomalies in the sea surface temperature Nino3.4</div><div style="left: 85.0393px; top: 1102.28px; font-size: 16.9238px; font-family: serif; transform: scaleX(0.981492);" data-canvas-width="802.7133333333333">and DMI applied in Indramayu district was able to follow the actual value of the variability of rainfall, especially during</div><div style="left: 85.0393px; top: 1123.95px; font-size: 16.9238px; font-family: serif; transform: scaleX(1.0249);" data-canvas-width="802.6883333333335">the dry season, while in Cianjur district the model was less able to describe it well. The resulting model for Cianjur</div><div style="left: 85.0393px; top: 1147.95px; font-size: 16.9238px; font-family: serif; transform: scaleX(0.9854);" data-canvas-width="570.3333333333331">district validation was low value so it is advisable not to use the model for prediction.</div>