Hydrodynamic Model Calibration for Mangrove Replanting using Maximum Absolute Error and Brier Skill Score
Published 12-09-2024
Keywords
- Mangroves Replanting, Hydrodynamic Model, Maximum Absolute Error, Brier Skill Score
How to Cite
Copyright (c) 2024 Journal of Water Resources Management
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
Abstract
Hydraulic numerical models are now widely used as a prediction tool in decision making
because they are convenient and practical for projects that involved wide area, thus more cost-saving compared to physical models. Numerical modeling normally applies complicated mathematical equations, which have coefficients that are site specific. Therefore, model calibration and validation are important to facilitate accurate representation of the study area and provide high level of confidence in the model output. The Indian Ocean tsunami disaster in 2004 has increased recognition of the importance of mangroves for coastal protection; hence encouraged more replanting projects along the coast of Malaysia. Since mangrove growth is governed by many coastal processes such as tides, waves, currents, type of sediment, nutrient availability, sediment transport, erosion and deposition; careful assessment of the factors is necessary to facilitate successful replanting activities. Numerical modeling was applied for the study area at Sungai Haji Dorani, Malaysia to investigate the performance of the developed model in assessing the suitability of the study area chosen for mangrove replanting. Calibration and validation of water levels, current speed and directions at Sungai Selangor, Pintu Gedong and Sungai Haji Dorani were carried out using Maximum Absolute Error (MAE) and Brier Skill Score (BSS) for time series data. Model results show that the BSS values are between 0.73 and 0.95 for water levels and current directions while BSS for current speeds are between 0.6 and 0.84. These values indicate that the model results can be accepted since the best prediction model is when the BSS value is 1. Analysis of model results show that the average velocity over Haji Dorani coast is 0.14 m/s and peak velocities varied from 0.1 to 0.4 m/s, higher than the particle velocities for fine sand, silt and clay, which will not encourage deposition of fine sediments. Even if flocculation processes do occur, the flocculants will disintegrate when the velocity is higher and the mud will be suspended and re-suspended in the water column as the substrate is not stable enough to consolidate. If mangroves are planted in the study area, the young mangroves will be washed away by the high tidal currents since the mangrove roots are not long enough to hold on to the unstable substrate. A temporary structure would likely assist to reduce the tidal current velocities in the study area if mangrove replanting is still desirable.