Effect of bridge construction on floodplain hydrology - assessment by using monitored data and artificial neural network models

摘要

Study of floodplain hydrology of urban areas is important for its implications to flood and pollution control and ecology of the floodplain areas. In this study, the effect of bridge-pier-strengthening work over Tama River at Nagata district, Japan on the general hydrological condition of the floodplain was investigated by analysing monitored hydrogeological data and results of artificial neural network (ANN) models. A conceptualization of surface and subsurface of flow condition based on field observations was made and linked with the behavior of monitored data at various monitoring wells. The spatio-temporal variation of hydrological variables in the study area was found to have influenced by the morphological change of the river and its flood plain in the last 120 years. The old river-channels have served as preferential flow pathways in the floodplain, which caused rapid change in groundwater level in some groundwater monitoring wells. The presence of some form of organized heterogeneity in the electrical conductivity in the study area was indicative of one of the effects of historic changes of the floodplain and river channel-pathways. ANN-based models formulated using the pre-construction period data offered benchmarks to analyse the effect of construction on the floodplain hydrology. Based on the interpretation of monitored data and results of ANN modeling, it was clearly found that the effect of construction on the floodplain hydrology was significant. (C) 2004 Elsevier B.V. All rights reserved.