Bedload transport in SE Asian streams—Uncertainties and implications for reservoir management

Abstract

This paper reviews the current state of knowledge regarding bedload transport in SE Asian streams and presents the results from a case study on bedload transport in a mountain stream in northern Thailand. Together, the review and new data help contextualize the paucity of work done in the region in the face of a rapid increase in development and reservoir building throughout SE Asia. Data from both the reviewed studies and the case study indicate that bedload transport in many SE Asian streams (e.g. catchment areas<100km2) is often much higher than is commonly assumed for tropical streams (i.e., about 10% of the total sediment load). Estimated annual bedload proportion was 18% of the total annual sediment load in the 74-km2 Mae Sa Catchment in northern Thailand. Bedload transport rates ranged from 0.001 to 1.1kgs−1; and measured total suspended solid (TSS) rates ranged from 0.01 to 39kgs−1, equivalent to TSS concentrations of 20 to14,000mgl−1 (associated with flows ranging from 0.4 to 30m3s−1). Event and annual loads of bedload and TSS were determined from rating curves based on automated measurements of discharge and turbidity (for TSS only). When taking uncertainty into account, the estimated range for the bedload proportion of total sediment load was 9–25% (equivalent to a yield of 81–279Mgkm−2y−1). The corresponding TSS yield estimate ranged from 649 to 1037Mgkm−2; and the total sediment load is an estimated 730–1313Mgkm−2 y−1. The proportion of bedload was lower than that reported in some other Asia streams, probably due to the occurrence of extended periods with high TSS that dampened the bedload signal, which was sand-dominated during the low-energy events that were sampled. Nevertheless, the bedload rate was generally higher than for most SE Asian locations, likely due to the occurrence of several road-related landslides the previous year. Although we were not able to measure bedload transport for high energy flows (discharges>4.5m3s−1), we believe our upper estimates for bedload variables (25% of the total sediment load; and a yield of 279Mg Km−2 y−1) provide reasonable upper bounds. Finally, the bulk of bedload transport is episodic in nature, with a higher proportion moved during high energy tropical storms that occur late in the monsoon rainy season, as well as in response to both natural and anthropogenic landscape disturbances. The possibility that bedload proportion could exceed 20–40% for rivers and streams of various sizes reinforces the need for accurate estimates of both bedload and suspended solid loads prior to building dams in the region. Past examples of reservoir closure following rapid infilling possibly stem from underestimating sediment loads, particularly the bedload component, and failing to factor in the very high sediment loads associated with large storm events.