This article assesses the risk of flooding and identifies efficient measures to reduce flood risk in Hat Yai Municipality, southern Thailand using GIS and satellite imagery. The center of commercial trade and administration in southern Thailand, Hat Yai is located in the downstream area of the Khlong U-Taphao Basin. From 1982 to 2002, forest resources in the basin's upstream were decreased from 48,281 to 26,781 hectares, equivalent to a reduction of 44.5 %. This was largely attributable to the expansion of rubber cultivation. Further analysis revealed that 17,116 ha, 42.7 % of the forest resources in the basin's headwater source areas mainly had been replaced by rubber plantations. As a consequence of extensive deforestation, particularly in headwater source areas, the municipality has become vulnerable to natural disasters; primarily floods.
Hat Yai Municipality experienced two flood events of catastrophic magnitudes in 1988 and 2000. These destructive floods were detrimental to its vulnerable social and economic development due to the loss of life and destruction of property. Since the 2000 flood, considerable structural mitigation measures have been undertaken to prevent and alleviate future flooding of the municipality. Recent flooding on 10 December 2003, caused by a storm like that experienced in 2000, indicates that structural mitigation efforts undertaken so far are inadequate to withstand natural threats. Hat Yai Municipality will continue to be extremely vulnerable to future flood disasters. This shows the compelling need to increase the municipality's resilience against flooding through adoption in the coming year of non-structural disaster reduction schemes to supplement existing efforts. Risk assessment based on hazard and vulnerability analysis is needed to identify and implement adequate and successful non-structural alternatives.
Analyses showed that 99.0, 400.0, 1110.0, 346.0, and 100.0 ha of the municipality's total land areas were subjected respectively to very low, low, moderate, high, and very high flood hazards. Further analysis revealed that 39.6, 654.6, 152.7, 664.5, and 543.6 ha of the land areas respectively faced very low, low, moderate, high, and very high flood risks. Analyses also showed that all the residential, commercial, industrial, and the public utilities and facilities areas, equivalent to 1,188.0 ha (57.8 %) of the municipality's total land area, faced high and very high flood risks. Measures to reduce the risk of flooding in Hat Yai Municipality are discussed.

View full abstract

Many slope failures have been observed to occur during or immediately after rainfall. Although conditions leading to these failures have been described as caused by a rapid rise in pore-water pressure as a result of rainwater infiltration, the important factors that influence the initiation of slope failures have not been adequately clarified. To investigate these factors, a series of laboratory experiments was conducted on model sandy slopes to determine the initiation process of rainfall-induced slope failure. In the tests, failures were induced in small-scale model slopes either by water percolation from the side upslope or by artificial rain falling on top of the slope. Besides monitoring pore-water pressure, changes in soil moisture contents and ground deformation were measured. Test results showed that slope failure was always induced when the soil moisture content within a certain region near the toe of the slope reached nearly full saturation, even though other parts of the sliding mass were still in a partially saturated state. In addition, minute deformations along the slope were shown to precede failure. The findings presented here show that by monitoring the soil moisture content of slopes and performing displacement measurements, it is possible to predict the occurrence of rainfall-induced slope failure.

View full abstract

In mid-January 2003 Tropical Cyclone Ami passed directly across the Fiji Islands in the southwest Pacific Ocean. The main northern island of Vanua Levu experienced torrential rainfall and consequent record-breaking flooding of its major rivers. The aims of this study were to document these record floods and compare them with previous floods on Vanua Levu Island. The Nasekawa River in southern Vanua Levu produced a phenomenal discharge of more than 6100 m3/s. Moreover, near the main town of Labasa on the north coast, simultaneous flooding of the Labasa, Qawa, and Wailevu rivers combined with storm surge to cause inundation of up to 4 m depths over a wide area of the floodplain. Tragically, seventeen people died, and there was extensive damage to farms, the infrastructure, homes, and commercial property.
Historically, the north coast of Vanua Levu island has suffered frequent severe floods, owing to several factors: 1. the approach of most tropical cyclones towards Fiji from the northwest; 2. Vanua Levu's steep volcanic topography, which rises in excess of 1000 m, and has strong orographic influence on rainfall generation during tropical storms, then rapidly transfers moisture into river channels, 3. the configuration of several drainage basins which deliver floodwaters to the same area of the coastal hinterland. Future regional ocean warming and more sustained El Niño conditions are projected to increase the intensity of tropical cyclones and thereby the potential for worse flood disasters. Disaster mitigation and adaptation options recently proposed by the World Bank and JICA need to be implemented to reduce flood impact in this vulnerable area of Fiji.

View full abstract

The focus of this paper is on 3-D adaptive analysis method of liquefiable soil improvement of the approximate quality of nonlinear numerical simulation of the liquefaction process with large deformation. This adaptive technique was applied to the 3-D non-linear FE analysis of liquefiable soil considering large deformation, including the liquefaction phenomenon. The cyclic elasto-plastic model and updated Lagrangian formulation were adopted in the three-dimensional FE analysis. The fission procedure belonging to the h-refinement indicated by the error estimator, a posteriori error estimate procedure depending on L₂-norm of strain and superconvergent patch recovery method, was used. The convergence of this error estimate scheme and the effectiveness of h-adaptive mesh refinement were shown by simple examples. The adaptive FE method was applied to the three-dimensional practical seismic analysis of an embankment constructed on liquefiable soil, and its efficacy demonstrated in detail. Application of the adaptive technique to the nonlinear analysis of saturated soil, including the liquefaction process, is a valuable step in the three-dimensional seismic analysis of liquefaction.

View full abstract