Urgency for sediment control

The loess plateau in China's Yellow River middle basin is one of the world's mostly badly eroded regions in the modern time. Extending for an areas of 430,000 sq km, the loess plateau contributes over 90% of the 1.6 billion tons of the total sediment carried by Yellow River annually. To control sediment production from the region we might first target areas identified as "severe eroded areas," whose annual erosion rates exceed 5,000 tons per sq km, for a total area of 156,000 sq km, contributing to 83% of the river's sediment load. Some areas have erosion rate reaching as high as 30,000 tons per sq km. Targeting the severe eroded areas is just a stop-gap measure attempting to reverse the continued deterioration of the conditions of the river, and to prevent flooding in the downstream densely population region. This paper once again emphasizes the importance to control erosion in the loess plateau, since new opportunities to finance such an effort opens up as the nation enters into a new economic environment.

Let us recall that when the Sanmenxia dam in the river's lower reaches was completed in 1960, sediment accumulation in its reservoir threatened to deplete its storage capacity in a few years. Consequently, the Sanmenxia dam had to be redesigned to include sediment-flushing facility to clear its reservoir of sediment accumulation, delaying its normal operation schedule till 1974. Presently, roughly 1/3 of Yellow River's annual flow is used for sediment flushing, and in so doing, 3/4 of the total sediment flow is flushed out to sea, leaving behind 1/4 of it, the coarse-grain sediment, to silt up the downstream river channel steadily. After thirty-some years of sediment accumulation, Sanmenxia dam's reservoir no longer has the capacity to temper a flood peak to protect safely the downstream diked region. Consequently, the Xiaolangdi dam is being built upstream of the Sanmenxia dam to afford safety to the densely populated downstream region. Even with sediment-flushing proviso, the Xiaolangdi dam is expected to have a very limited useful life, if sediment deposition in the river remains unchecked. What then is the long-term prospect for Yellow River?

Returning to the strategy for erosion and sediment control in the loess plateau, opinions on how best to carry it out differ widely. On the one hand, it is advocated to implement an extensive and comprehensive water-soil conservation program over the entire land surface with terracing of cultivated hill slopes, planting of trees, shrubs, and grass to retard surface erosion, as well as legal ordinance limiting human activities to prevent the destruction of forests and herding of sheep and goats. On the other hand, others argue in favor of the relatively simple proposals of just stopping sediment flows coming out of the gullies and minor river tributaries, and preventing them from reaching the major rivers.

In my opinion, while it is perfectly sensible to implement a well-planned water-soil conservation program just for environmental protection reasons if nothing else, the program will have accomplished very little, in the next few decades, in solving the sediment accumulation problem afflicting Yellow River's downstream region. The climate in the loess plateau is far too dry to support a reasonable vegetation cover, and the loess soil is too loose to be protected from water erosion due to severe rainstorms during the monsoon season by such a sparsely covered surface. As many Chinese engineers have already advocated, a workable sediment-control program should best be based on just the limited goal of preventing eroded soil from leaving the gullies before it enters any drainage system. I tend to share the latter point of view.

Gully development

The loess plateau is badly dissected by erosion into tens of thousands of steep gullies. Surface runoff drains very rapidly from the region, creating flood peaks to the downstream rivers during the flood season. Recognizing the importance to arrest the surface runoff while it is still in the gullies before entering the rivers, check dams are built there to retain sediment near its source. Funded by the government and with the sole purpose of retaining sediment, the program suffers from the lack of funds due to its limited goal. On the other hand, farmers by their own initiative also build silt trappers, or warping dams, in the gullies, wherever they can, back filling them with sediment to create new farmland. Unfortunately, runoff through the gullies after a major rainstorm is sufficiently rapid that a 10-year storm can cause such destruction to the warping structures that the farmers are discouraged from utilization of the gully floors. Many of the gullies are left barren, and oftentimes eroded to the bedrock, one or two hundred meters deep.

Although it is relatively inexpensive to build warping dams to create fertile farmland, in many areas this cannot be done until the occasional severe runoffs down the gullies are blocked. Various schemes have been proposed to coordinate the placement of the warping dams so as to distribute the burden of flood retention and to minimize destruction. Such schemes are however only applicable in the more densely populated areas where there is enough labor force to carry out such extensive and well-coordinated program. For the majority of areas, many of which still uninhabited, the straight-forward approach of placing flood-prevention structures, called the key dams, one for each subwatershed, providing a flood-free environment for agricultural development would seem to work best. Typically, key dams are earth dams 30-50 m high with reservoir capacities to withstand a 300-year storm even after 20 years of silting.

Up to a thousand key dams have been built with government funding in the past thirty-some years, and a great deal of engineering expertise has been acquired in the process. However, in order for the sediment control program to have an impact on the conditions of Yellow River, we anticipate building around 20,000 of such key dams. To do so the sources of funding must be greatly expanded, and an economic environment must be created to make gully development a sustainable enterprise. Looking over some of the existing key dams, it is apparent that these projects are designed mainly with sediment retention in mind. The projects pay little attention to either the recovery of the investment cost or the projects' agricultural potential. Usually after 10-15 years of sediment retention, the dam's reservoir will be filled with sediment 20-30 m deep, to nearly the top of the dams, creating 10-15 hectare (ha) of high-quality farmland. To utilize the sediment plots created in the reservoir, more earth structures will be added upstream of the reservoirs so that the flood water will not inundate the crops before they are harvested. The land is turned over to the local villages, and the original investor in the projects, namely the government, recovers no direct payment for its investment, and new projects, or just strengthening of the existing dams, must rely on new allocations of fund. The lack of direct return to the funding source explains partially why the projects of this type could not move ahead with a sustained pace, while the national economy as a whole is making rapid progress in other sectors.

To advocate the utilization of gullies, it is necessary to think of sediment not as destructive entities to be arrested, but as resources to be utilized and developed. Trapping sediment in reservoirs to a depth of 20-30 m is simply too wasteful a scheme to utilize such valuable resources. Instead, the sediment should be allowed to pass through the dam to fill the gully floor downstream. Each year 2-3 hectare of sediment plots can be created for each key dam. The plots are topped by no more than 1 m in depth of the fertile sediment over prepared beds. With flood water under control by the key dam and sediment passage regulated, the warping structure built downstream can be relatively inexpensive. Such a scheme would maximize the creation of sediment plots with irrigation potential. Under the present policy of land auctioning to the farmers for multi-decade user rights, and collecting reasonable annual rents on the new land, land development can actually become a profitable investment.

Take for example a small watershed of 60 sq km in area located in a remote site not yet densely populated. It will have a number of branch gullies each draining a subwatershed of 3-10 sq km in size. Lets take for the sake of discussion a subwatershed of 5 sq km. A key dam can be placed roughly in the middle of the branch gully controlling a catchment basin of 3.5 sq km in size. The key dam will be 35-40 m in height meeting all standards of current key dam specifications. After each major rain burst it intercepts on average 70,000 cu m of storm runoff together with 15,000 tons of sediment. The drainage facility of the dam must be such as to discharge the entire sediment intake within a few hours, before the sediment settles in the reservoir, to create sediment plots at the foot of the dam. (This is an added requirement over the current design of key dams, and would increase their construction cost by roughly 20%.) If the gully floor downstream from the dam is already prepared with low warping dams and partially leveled with earth (using tractor operation as in terracing), 15,000 tons or 11,000 cu m of sediment can lay down 1 ha of new surface 1 m thick. Each summer, 2 to 3 such rain bursts are to be expected, and we anticipate the creation of 2-3 ha of new land each year. The construction cost of the key dam is estimated to be 600,000 yuan (Y), and the preparation cost of the sediment plots downstream from the dam to be 10,000Y per ha. [Approximately, 8 Y per each US dollar.]

The construction cost will now be recovered through rental of the sediment plots to the farmers, which can be assigned a reasonable annual rent of 3,000Y per ha, since the annual productivity of the land is over 11,000Y per ha, (and this is just considering raising staple crops, while cash crops can bring in much higher income).

Continued in Part 2 : Social factors - Organizational requisites - Sediment retention subsidy - Final remarks

Picture credits: Photos by George Leung, except for the first picture which is taken from Soil and Water Conservation in the Huanghe River Valley, edited by Huanghe River Conservancy Commission, published by Shanghai Educational Pub House, 1988.