Effects of Erosion Control Measures on Mountain Floods： A Case Study of the Censhui River South Branch Watershed

To investigate the effects of various erosion control measures on mountain floods, a case study was conducted in Censhui River South Branch Watershed using scenario analysis and soil conservation service （SCS） methods. A distributed hydrological model was developed, and watershed parameters were determined based on satellite imagery, digital terrain models, digital maps and field investigations. Two types of erosion control measures were investigated： the variation of vegetation covers and the change of cultivation techniques. Seven scenarios were considered for the test watershed. The results show： （1） while the de-vegetation results in the increase of peak discharge, the improve of vegetation covers decreases peak discharge at watershed scale; （2） by both improving vegetation cover and enhancing terrace-cultivation technology, the peak discharge is reduced and the peak flow arrival time is delayed; （3） attention should be attached to both early warning system and measures changing the underlying surface and conveyance systems.

Influence of Flow Regime on the Vegetation Zonation along Mountain Streams in the Western Cape, South Africa

Zonation patterns of riparian vegetation have been sampled and described in mountain streams in two catchments in the Hottentots-Holland Mountains, Western Cape, South Africa. Six main vegetation types that differ in structure and species composition, are dominant along these river banks: Aquatic vegetation, Wetbanks, Palmiet, Scrub, Forest and Shrubland(Fynbos). The study aims to correlate the vegetation patterns to flooding patterns, in particular the inundation frequency and stream power. A problem arises: because these catchments are ungauged, like most mountain catchments, with the only weirs at the downstream end of the catchment. Discharge data at the weirs are extrapolated to the sites upstream by multiplication with a factor based on the size of the subcatchment that drains through a sample site. In this way, recurrence intervals for floods in mountain streams are derived. Discharges at sites are also calculated using bed roughness(Manning's n) and slope in straight sections with uniform flow conditions. Stream power is derived from the discharges calculated in this manner. The combination of stream power and recurrence intervals explains the occurrence of most vegetation types occurring on the banks, except for one type: Afromontane Forest. This type is probably more dependent on other factors, such as protection from fire and the depth of the groundwater table.

Application and comparison of coaxial correlation diagram and hydrological model for reconstructing flood series under human disturbance

Intense human activities have greatly changed the flood generation conditions in most areas of the world, and have destroyed the consistency in the annual flood peak and volume series. For design flood estimation, coaxial correlation diagram and conceptual hydrological model are two frequently used tools to adjust and reconstruct the flood series under human disturbance. This study took a typical mountain catchment of the Haihe River Basin as an example to investigate the effects of human activities on flood regime and to compare and assess the two adjustment methods. The main purpose is to construct a conceptual hydrological model which can incorporate the effects of human activities. The results show that the coaxial correlation diagram is simple and widely-used, but can only adjust the time series of total flood volumes. Therefore, it is only applicable under certain conditions(e.g. There is a strong link between the flood peaks and volumes and the link is not significantly affected by human activities). The conceptual model is a powerful tool to adjust the time series of both flood peak flows and flood volumes over different durations provided that it is closely related to the catchment hydrological characteristics, specifically accounting for the effects of human activities, and incorporating expert knowledge when estimating or calibrating parameters. It is suggested that the two methods should be used together to cross check each other.