Chaos in the waters: investigating non-linear dynamics of rainfall in river basin

Floods are a perennial natural disaster in Malaysia, having significant impacts on urban and rural society. It is essential to recognize the random nature of rainfall to enhance flood prediction and water resource planning. This research explores the evidence of chaos in rainfall patterns in the Ampang (urban) and Hulu Langat (rural) river systems through the 0-1 Test and Lyapunov Exponent (LE) techniques. A ten-year record (2013–2023) from ten chosen rainfall stations was used to identify the evidence of non-linear rainfall time series. The imputation of missing values and reduction in noise through the Single Exponential Smoothing Technique improved the quality of the data, and the chaos test was applied to the improved data. The findings affirm the evidence of chaos in rainfall in urban and rural river catchments, displaying sensitivity to the initial condition and non-linear. Against anticipation, the magnitude of chaos in the urban and rural catchments remained close to each other, indicating the influence of urbanization and land use alteration did not impact the chaos in rainfall. The findings suggest the limitation in conventional hydrological models since they are linear in nature and fail to capture the unpredictability of extreme rainfall events. Through the integration of chaos theory in the study of hydrology, the research offers useful contributions to the improvement of flood forecasting models and disaster readiness. The research underscores the necessity to develop adaptive, non-linear flood forecasting models to accommodate the innate chaos in rainfall patterns. The research findings could support the policymaker and urban planner in the formulation of effective flood protection strategies to build flood resilience to extremeweather events.