Free Flood by Stephen Baxter Page A

couple kilometers thataway,” Thandie called, pointing east. “Sanj, how’s the data? You got GPS?”
    “I got that,” said Sanjay, staring at his screen. “Climate sensors nominal, though that wind gauge is going to rip clean off at this rate. And the pressure’s dropping. Nine seventy. Nine sixty-five . . . The radar’s working, the sonar not so well, you’d expect that. It would help if this tub wasn’t bucking like a fairground ride.”
    “Doing what I can, brainbox.”
    Gary had had no idea that all this industry was out here. “It’s like a city in itself. And kind of vulnerable, isn’t it?”
    “When it comes to fuel London’s a big and thirsty monster, Gary. But they’re prepared for floods, they drill for them.” She snapped a switch, and the radio cut into a feed from a refinery crew going through shutdown procedures, working through checklists of pumps, furnaces, compressors, valves, catalytic crackers.
    “Leaving it late,” Gary said.“The storm’s been tracked since Scotland.”
    “A flood warning itself is an expensive event,” Thandie said. “With more than a million people living on the Thames flood plain, you don’t raise the alarm unless you have to. The river traffic is a problem too. The Barrier seems to be raised more often than it’s lowered nowadays. And shutting down those refineries is no joke, you don’t just throw a switch. It costs to abandon the processes they put their materials through. False alarms are unpopular. People are terrified of liability, legal claims.”
    “And in this case,” Sanjay said, “the error bars around the storm’s probable track and effects were just too wide to be sure. I told you, our modeling is breaking down. What’s worse is that the interfaces between different models aren’t working so well either . . .”
    Gary understood the principle. Mathematical models of the weather were generally based on dividing up land, air and sea into discrete elements and tracing the progress of variables like pressure, temperature and wind speed through from one element to the next. You might run a coarse model for the whole of the North Sea, and as a storm passed the Wash or the Thames estuary you would feed predicted conditions from the ocean model to finer-grained models to see what happened in there. But if all the models were suffering because of some underlying change in the physical condition in the planet’s weather systems, it would be at the edges and interfaces that errors would particularly multiply.
    Sanjay said, “The last great London flood was back in 1953. That event led to the construction of the Barrier, eventually. Much of Canvey is below sea level; people died here. But that flood was a convergence of high tide with a big storm surge.”
    The low-pressure air at the heart of a storm could lift the level of the sea below it, physically sucking it up into a hump that could be hundreds of kilometers across. And then the winds could drive the high water against the coast or into a river estuary. That was a storm surge.
    “So is this a surge? Are we hitting a high tide?”
    Sanjay said,“The storm is driving waves ahead of itself, but I wouldn’t call it a significant surge. And as for the tide, the predictions now are all over the place.”
    Gary said, “So this event doesn’t have those key features that characterized the 1953 event. And yet we’re getting a flood even so.”
    “Looks like it,” Sanjay said.“It’s not even a particularly severe storm.” He sounded unhappy, as if the real world were a bit of grit in the oyster-shell of his science.
    And Thandie called, “Oh shit. Here it comes.” The chopper dipped and bucked as she hauled them back out over the river for a better view.
    Gary, peering through a rain-streaked window, saw the wave coming, water raised and driven by the North Sea storm and bottled up in the narrowing, shallowing estuary. As it advanced it spilled almost casually over flood barriers and