Web-based geographic information system for flood disaster using spatial autoregressive method

Abstract

The territory of Indonesia is classified as one of the countries prone to disasters, both natural disasters and disasters caused by human activities. Flood disaster is one example of a disaster that often occurs, especially in the DKI Jakarta area. The common cause of the flooding is usually caused by high rainfall exceeding the average in general so that water sources such as rivers, reservoirs, and lakes cannot accommodate the amount of rainwater. This is also supported because the water absorption system is minimal so that it is unable to absorb excess water, causing flooding. The method used to define the relationship between various factors and the relationship between regions is to use the Spatial Autoregressive (SAR) method. In order to get a better relationship between regions, the existing spatial data will be visualized with the Geographical Information System. This research focuses on creating a web-based Geographic Information System application. Applications created using the programming language PHP and Javascript with the Codeigniter framework. The results obtained by the SAR equation to predict the height of flooding in the DKI Jakarta Province and the test results using the Mean Absolute Deviation (MAD) were 1.354656% while using the Mean Absolute Percentage Error (MAPE) was 2.157288%. Testing applications using the black box testing method was obtained by 93.35%. This shows that the SAR Model is suitable for predicting flood heights in the DKI Jakarta Province. This flood prediction can be used one year from the prediction model made.