To an extent, it is: Surat never finished building embankment walls along the Tapi, nor did it have the river desilted, nor did it lay storm-water drains to take away rain water. But to give due credit to the administration, the flood warning was issued in time, and evacuation began promptly.

Work on the embankment walls never really began in earnest: the state government had suspended construction saying it was expensive and that embankment walls would be of little help during heavy flood.

The fact is this year flood water entered the city from the same spots -- near Ved Road and Katargam -- where the embankments are breached or do not exist. Officials of the Surat Municipal Corporation (SMC) and those at the collectorate argue that so much water was released that the embankments would not have helped, but others says they could have at least delayed the entry of water into the city.

Even after the 1994 experience, no study has been initiated to measure the extent of silting in the river. Waste from the city, and the three bridges that are close together, have caused a levelling of the river, which means it will flood more easily. SMC's consulting surat city engineer P.C. Shah admits this, and says the corporation has now initiated a desilting scheme. ``But even that (desilting) wouldn't have helped in the face of nature's fury,'' he said.

Without an effective network of storm water drains in place, Surat has another problem at hand: the flood water will take more time to recede. In 1994, it had taken five days, creating such unhygienic conditions in the city, then famed for its squalour, that people were not really surprised at the outbreak of the plague epidemic.

Another factor contributing to the flooding is that the mushrooming of slums on the riverbed has narrowed the river in its passage through the city, reducing its capacity to take away water from the overflowing dam.

An expert Narmada & water resource department in Gandhinagar said the Tapi, which used to carry away 3.66 lakh cusecs of flood water, could now drain only some 2.5 lakh cusecs.

``It is high time that local authorities discourage slums on riverbeds, not only on the Tapi, but also other rivers flowing through big towns and cities,'' he warned. ``In Ahmedabad, for instance, the Sabarmati riverbed has two or three slums''.

Some things, however, went right for Surat. Sources in the Central Water Commission (CWC), which warns of floods developing in the catchments of Tapi in Madhya Pradesh, said warnings had been passed on in advance and that they had been acted upon promptly. CWC official Ashok Kumar said, ``By 7:30 a.m. on September 15 the Ukai authorities were informed that a flood of 17,500 cusecs is developing near Burhanpur and would bring in 2,500 million cubic metres of water to the dam after 40 to 50 hours''.

He said release of water from Ukai had started within two hours of information reaching them and that it had been released gradually to avoid a flood.

``This time, if they allowed the dam level to go up to 346 feet, one feet over its full reservoir level (FRL), it was only a little less in 1994. But what worsened the situation this time was the heavy rain in the catchments of Ukai dam, which only added to the fury,'' he said.

Kumar said there could not be any system to gauge the extent of a sudden cloudburst in the dam's catchments. So, at one point of time, when CWC told Ukai to expect 6 lakh cusecs, what came was much more.