War in the Balkans: How land assault could be made

The timing

IN ONE sense, it is already too late. Apocalypse has struck Kosovo. Half the 1.8m ethnic Albanian population has been uprooted. The rest could follow in two or three weeks, before any serious Nato ground attack could be launched. It took three months to mass the 500,000 allied troops who drove Saddam Hussein from Kuwait in 1991. This time distances are shorter and fewer troops will be needed.

About 40,000 Yugoslav troops and police are in and around Kosovo, apart from unknown numbers in militias. The conventional wisdom is that, in a ground war, an attacker must outnumber the defender by three to one to prevail. However, Nato's aerial dominance and the shortages of fuel, parts and even food soon to confront the Serbs mean the force could be only 50,000. Currently, the alliance has about 12,000 men in Macedonia, and a smaller force in Albania. But in both countries Nato has been distracted by the colossal humanitarian disaster of the refugees. It could take a month or more to build the force up to the minimum needed to take Kosovo.

The logistics

HOW DO you get the men there and supply them? Had it been able to choose, Nato could hardly have picked a trickier target than Kosovo for its first military offensive. You can't fly in a ground army overnight; it takes, for example, 500 sorties by giant C-141 transport aircraft to move just one light division with minimal supplies, and without armour. This sort of exercise requires decent ports and decent highways.

Desperate to avoid being sucked into the conflict, Macedonia, the most convenient springboard for any ground assault, has said it will not allow its territory to be used as a base for an invasion. And what if Macedonia's Slav-dominated government asked the Nato forces to leave? On the map, the Yugoslav republic of Montenegro, sympathetic to the West and anxious to escape the grip of Milosevic, looks an option. But it has no adequate port and, more importantly, its government might well be toppled in a coup by Serbian hardliners if it looked likely to become Nato's gateway into Kosovo.

This leaves Albania. Here, obviously, support for Nato is not a problem, but strategically it's a different story. The country has only four inadequate ports (Durres is the best of them), and bad roads up to the Yugoslav/Kosovo frontier.

The terrain

KOSOVO'S BORDER with Albania, like the one with Macedonia, favours the defender. Moreover, Belgrade - which has flexed its muscles frequently there with "hot pursuit" raids against Kosovo Liberation Army (KLA) bases - is preparing anti-tank defences, including mines and trenches at the most obvious entry points.

Once into Kosovo, Nato ground troops could also find the going tough. The opposition will be well hidden, and scattered across rugged, broken country. The ethnic Albanians left in Kosovo are more likely to be human shields for Yugoslav tanks than anti-Serb partisans. A Nato force of 50,000 to 100,000, with its vastly superior firepower, would surely prevail. But not before casualties running into the high hundreds or low thousands, military experts predict, and perhaps not before the war has sucked in Macedonia, Albania and maybe others. All of which could stretch to breaking point the strains within the alliance.

Nato Unity

SO FAR the alliance has held together far better than expected. Even Germany, with its historical qualms about taking part in an offensive war in Europe, has not wavered during the air phase. Whether it would participate in a land invasion is another matter.

Britain and France, Europe's two biggest military powers and co-sponsors of the failed Rambouillet peace conference, would be expected to play the major part. But America's involvement would be essential: if it refused to act while others went ahead, Nato's entire future would be in doubt.

The biggest risk is Greece. The only way of getting heavy armour close to Kosovo quickly is via the port of Thessaloniki, which has good rail and road links north to the Macedonian capital, Skopje. But that needs Greek approval and Athens, a friend of the Serbs and Nato's most reluctant warrior, is adamant that it will not become involved in military operations. To do so would almost certainly bring down the government of the Prime Minister, Costas Simitis, and destabilise Nato's south-eastern flank.

The options

WITH ALL these difficulties, therefore, a ground war will be no walkover. Nato's choices are limited. The most obvious option is to use forces in Macedonia and Albania to secure a safe zone in south-east Kosovo, which could serve as shelter for the refugees and thereafter as a platform for a wider campaign to drive the Serbs from the province - or as the basis for its future partition. If presented as a humanitarian operation, Greek and Macedonian objections could be overcome, while Nato's air superiority means the operation is feasible, even with a relatively limited force. Probably something along these lines will happen even if - for the reasons already outlined - supplying the troops could be a problem.

The extreme option is a ground attack, probably from Hungary, aimed not at Kosovo but at Belgrade, where the root of the problem lies. This would truly elevate Kosovo to the last great war of the century, and Mr Milosevic to a Hitlerian plateau of evil. It would also be a war with even less predictable consequences, and massive casualties, which even an emotional, morally driven Western public would find hard to accept.