Siltation Risk and the Choice Between Upstream Treatment and Dredging
A catchment of 42 km² drains into a proposed irrigation reservoir at a gorge in Magway Region. Mean annual rainfall is 680 mm, about 80% of it falling between mid-July and mid-September, often in two or three intense storms. Since the 1970s the upper catchment has been cleared for shifting cultivation and fuelwood, and the measured sediment yield at the gauging station is 3,400 tonnes per km² per year. The reservoir is designed with 12 million m³ (Mm³) of storage and a 60-year life. A consultant's report states that without intervention siltation will remove 20% of storage by year 25, and recommends upstream terracing and reforestation while noting that these take about ten years to measurably reduce sediment delivery. Dredging is technically possible but has been costed at roughly four times the cost per m³ of storage recovered by upstream measures. Downstream, 1,200 ha of paddy and 400 ha of cotton depend on the scheme.
1. Using the sediment yield, an assumed trap efficiency of 85% and a deposited bulk density of 1.4 t/m³, test the consultant's claim that 20% of storage will be lost by year 25.
Read the worked answer
Work in consistent units and keep the arithmetic visible.
- Annual sediment delivered: 42 km² × 3,400 t/km²/yr = 142,800 t/yr.
- Convert to volume: 142,800 ÷ 1.4 t/m³ = 102,000 m³/yr, i.e. 0.102 Mm³/yr.
- Only part is trapped: 0.102 × 0.85 = 0.087 Mm³/yr.
- Over 25 years: 0.087 × 25 = 2.17 Mm³.
- As a share of the 12 Mm³ live storage: 2.17 ÷ 12 = 18%.
The consultant's 20% is therefore of the right order but slightly pessimistic, and the conclusion is credible. Two cautions must be stated: the calculation assumes a constant sediment yield, whereas clearing usually intensifies for some years before stabilising; and it ignores bank erosion and slumping of the exposed shoreline during drawdown, which adds sediment that the gauging station never measured. Adding those would push the figure toward or above 20%. Extending the same rate over the full 60-year design life gives about 5.2 Mm³, or 43% of storage, so the loss is not trivial even if the annual figure looks small.
2. Recommend how the budget should be divided between upstream measures and dredging, giving your reasoning about timing as well as cost.
Read the worked answer
A defensible answer must weigh cost per cubic metre against the lag in effectiveness.
Cost: upstream terracing and reforestation recover storage at roughly one quarter of the dredging cost per m³, so in pure cost terms they dominate. They also reduce the sediment supply rather than merely removing what has arrived, and they bring side benefits (reduced hillslope erosion, fuelwood and fodder, less road damage).
Timing: the benefit is delayed by about ten years, and during that decade roughly 0.87 Mm³ (about 7% of storage) still accumulates. That loss is tolerable but should be monitored with annual bathymetric surveys so the decision can be revisited.
Recommendation: fund the upstream package now and treat dredging as a contingency, not a substitute. Set a trigger - for example, if survey shows loss exceeding 12% of storage by year 12, or if the sediment yield has not begun to decline, then commission dredging. Retain check dams in the gully heads to trap coarse sediment cheaply while the vegetation establishes. Pure dredging-only is indefensible because it repeats a recurring cost without ever slowing the inflow.
3. Explain why the annual rainfall total is a weak predictor of sediment yield in this catchment.
Read the worked answer
Sediment yield depends on how the rain arrives, not just how much falls. In this catchment 680 mm is concentrated into a few high-intensity storms; a single 90 mm storm in an hour on bare, crusted ground can detach and move more soil than several weeks of gentle rain, because infiltration is exceeded and overland flow connects to the channel network.
Three further reasons should be given. First, ground cover matters more than rainfall: the same storm on terraced, vegetated slopes delivers far less sediment, which is exactly why the upstream measures are expected to work. Second, slope and connectivity control whether eroded soil reaches the channel or is redeposited on footslopes and in fields. Third, the dry-season conditions before the monsoon determine the antecedent soil moisture and whether crusting and cracking have prepared the surface for rapid runoff.
Consequently two years with identical totals can have sediment yields differing by a factor of two or more. A yield estimate for design should be built from storm intensity, land cover and slope, with rainfall total used only as context.