Water shortages arise for reasons that are partly natural and partly man-made. In the humid south, there is abundant water - 80 per cent of the mainland's supply - but relatively inhospitable terrain.

The dry north, on the other hand, boasts 70 per cent of the country's arable land that must be irrigated with only 20 per cent of available water.

As the population has increased from less than 500 million to more than 1.3 billion in the past 50 years, farming in the four key northern river basins - the Yellow, Hai, Huai, and Liao - has tapped ever-increasing amounts of water for irrigation.

Industrial usage, while proportionately lower, also has seen rapid expansion.

The result is demand outgrowing supply. Throughout northern China, 550 million people now have insufficient water, while 400 of the nation's 600 cities have water shortages.

In 1972, the Yellow River ran dry for the first time. Since then, increasing rates of extraction upstream have caused the river to dry up with monotonous, and alarming, frequency for 133 days in 1996, and 226 days last year.

As the upper reaches of the Yellow River undergo rapid industrialisation, conditions can only get worse.

Meanwhile, the Huai River ran dry for the first time last year; while out-take from the Hai, which feeds Beijing and Tianjin, is projected to soon exceed sustainable supply.

The implications for areas downstream of key out-take points, most notably Shandong province, are ominous.

According to the World Watch Institute, Shandong, which provides 20 per cent of the mainland's corn and almost 15 per cent of its wheat, depends on the Yellow River for half of its irrigation water. With falling water supplies and rising consumption for agricultural purposes - from the present 85 per cent of supply, or 400 billion cubic meters, to a projected 665 billion cubic metres in 2030 - it is clear that extraction patterns will not allow the province a sustainable supply of water.

There are only two possible outcomes. Either agricultural production declines, or the mainland finds an immediate and effective way to address the shortfall.

The former outcome, propounded by World Watch Institute's Lester Brown and endorsed by a study commissioned by the US National Intelligence Council, could have serious global ramifications.

As Mr Brown has written in a doomsday forecast, emerging water shortages could raise the country's demand for grain imports sharply, pushing the world's import needs beyond exportable supplies.

The reality, however bleak, is unlikely to be catastrophic.

As World Bank water expert Daniel Gunaratnum said: 'If they don't continue to moderate according to their proposed plan things are going to get worse, but I don't think the authorities are going to let that happen.' According to Mr Gunaratnum, the government has three basic strategies to address the problem.

Firstly, water management will be improved. This involves reducing upstream extraction by enforcing water allocation plans in areas where out-take is especially flagrant, such as Inner Mongolia.

Authorities will rationalise the activities of existing dams, which are being run as power stations, rather than flow regulation points. The new emphasis will be on efficient release of water instead of the maximisation of hydro-electric power output.

New damming along the Yellow River also will allow better overall water management, and a means to target supplies to specific areas. These projects will include the Lijiaxia hydro-electric power station and the $2.2 billion Wanjiazhai dam, which will allow the pumping of water via a network of several hundred kilometres of underground tunnels to Taiyuan, the capital of Shanxi province.

Still, these measures are, at best, a stopgap. In the end, effective relief of the north's water woes is likely to be possible only after completion of the biggest project of all - one that in financial and engineering terms will dwarf the challenges of the Three Gorges Dam, and have environmental impacts that are impossible to predict.

The quaintly named South Waters North project involves the diversion of water from the Yangtze, the only major watershed with a surplus, to the north. Three schemes have been mooted.

The first, known as the western route, will draw about 19.5 billion cubic metres of water from remote Qinghai province on the Tibetan Plateau, and pipe it through a network of tunnels and aqueducts to the upper reaches of the Yellow River.

Although the shortest of the three, it also is the most technically difficult.

The second, or middle route, will draw 15 billion cubic metres of water from the Danjiangkou Reservoir in Hubei province 1,230 kilometres north to Beijing. It would involve the resettlement of 220,000 people and the construction of more than 1,500 buildings and bridges.

The third, or eastern, route involves diverting water from the Yangtze near Shanghai and shifting it north via the remains of the original Grand Canal - China's first large adventure in aquatic infrastructure, built in the 7th century - to Tianjin.

As if distance were not enough of a challenge, the eastern route also suffers from originating at a lower elevation, by about 40 metres, than its ultimate destination. This means that the billions of cubic metres of water involved must be pumped uphill along its 1,100km journey.

Although debate has raged for years over which, if any, of the projects will be built and when, Mr Gunaratnum believes that ultimately all three routes can and will be completed.

'The first two will be the middle and the eastern routes, which may come in simultaneously - the western route may come in rather later,' he said, adding that construction was likely to start by 2000 and continue for between seven and 10 years.

Cost projections vary, but all run to tens of billions of dollars. Construction, in any event, is likely to cost at least as much as the Three Gorges Dam.

The final plank in the water conservation campaign is rationalisation of water-pricing. Although more prosaic than its heroic socialist equivalent, this type of scheme probably is more cost effective than simply carting water from one part of the country to another.

Reform is needed, but has been patchy in implementation. Water prices in urban areas are on average a mere 12 US cents per cubic meter for domestic users, while industrial and commercial enterprises pay 16 cents and hotels 29 cents. In rural areas, the average price drops to four cents, or about 10 per cent of the cost of supplying the water. Pricing at this level promotes waste and deprives utilities of the means to improve infrastructure and services.

As water shortages worsen, these various solutions drift toward completion.

Although both the scale of the problem and the complexity of the proposed solutions prevent any reliable analysis of the ultimate outcome, one thing is sure - water scarcity, and the means of overcoming it, is set to become one of the most dominant political and social issues for decades.

Water shortages could sharply raise the country's demand for grain imports