Prediction of Floods in Kerala using Hybrid Model of CNN and LSTM

Abstract

Floods are a widespread natural calamity that have far-reaching effects on human society, economics, and culture. Researchers have been hard at work for years on flood prediction models. Predicting the condition of flooding is a difficult task that requires in-depth research of the causes of floods. Changes in LULC are of paramount importance to decision planners and environmentalists because of unsustainable development's impact on ecological systems. Therefore, the purpose of this research is to determine the LULC for the Vembanad Lake System (VLS), Kerala. In this research, we combine the CNN and the LSTM network to create a hybrid CNN-–LSTM classical for flood prognostication, which takes full advantage of the fact that the CNN can efficiently extract features while the LSTM can accurately reflect the longer-term historical procedure of the time series data. Four models are then constructed for flood prediction: a LSTM perfect, a LSTM perfect, a CNN-–LSTM perfect, and a CNN-LSTM model. Finally, these models are evaluated using a variety of metrics, and the suggested CNN-LSTM model is exposed to be the most effective because to its low error and fast training time