Floods-Atmospheric rivers: When the sky falls

The following information is used for educational purposes only.






Atmospheric rivers: When the sky falls

03 April 2013 by Dana Mackenzie







Extreme floods around the world could have a common cause – mysterious great rivers of water that gush through the atmosphere

IN THE south-west toe of the UK, 2012 was the year that the weather played Scrooge to everybody's festive plans. In the five days leading up to Christmas, the seaside city of Plymouth got more rain than it usually gets in the whole of December. In Braunton, 80 kilometres to the north, the river Caen overwhelmed a recently completed flood-control project, inundating the town with water instead of shoppers. The main rail link connecting the region to the rest of the UK was cut off for six days. Even for an area more accustomed to wet than white Christmases, it was out of the ordinary.

That is nothing on the Christmas California endured 150 years ago. Starting on Christmas Eve 1861, Sacramento experienced a biblical 43 consecutive days of rain that left it submerged under 3 metres of water. The surrounding Central Valley became a lake 30 kilometres wide that did not recede for months.

Different times, different places. But there are similarities between the two cases beyond unusually soggy and cheerless Yuletides. California and the UK are both mid-latitude regions with an ocean-facing west coast. And the chances are the floods had a common cause: an atmospheric river.

Atmospheric rivers are vast, unbroken streams of water-laden air that can snake thousands of kilometres through the sky. Only recently identified and named, they are huge not just in geographical extent. "In terms of the water they dump as precipitation, atmospheric rivers are every bit as big and bad as hurricanes," says Michael Dettinger of the United States Geological Survey in La Jolla, California. Unlike hurricanes, they do not generate massive publicity, evacuations and early-warning efforts. Dettinger and others say this must change.

The effects of atmospheric rivers are nothing new. Residents of California have long talked about the "Pineapple Express", winter storms laden with warm water that originate around Hawaii. But atmospheric rivers were officially discovered on the other side of the country – in a computer print-out. In 1998, Yong Zhu and Reginald Newell of the Massachusetts Institute of Technology were running a model of Earth's climate when they noticed that it showed that almost all of the water vapour travelling between the tropics and mid latitudes was contained in narrow, intense bands.

This went against the grain. Severe weather was generally associated with the low-pressure centre of a storm system, an assumption reinforced by the satellite images available at the time. These images were recorded by monitoring Earth's infrared emissions, which are absorbed by water and other molecules on their passage through the atmosphere. They generally show blobby weather systems speckled with areas of more or less moisture. From this perspective, temperate zones were watered by a diffuse system of sprinklers – not the fire hose the model suggested.

As it turned out, 1998 was an El Niño year, and so an ideal time to settle the issue. This Pacific-wide phenomenon tends to bring unusually wet winters to the US west coast, and the US National Oceanic and Atmospheric Administration (NOAA) was planning to fly several sorties into storms over the Pacific, releasing expendable instruments called dropsondes. Like weather balloons in reverse, these beam back measurements of wind speed and water vapour as they fall.

They saw exactly what the model predicted: warm "conveyor belts" of moist air a few hundred kilometres across not at the centre of storm systems, but moving rapidly along their peripheries. The real surprise was how much water they transported – and how far it got. "One storm was conducting something like 20 per cent of all the water vapour transport from the tropics to the poles for the whole northern hemisphere," says Dettinger. "That's the sort of thing that makes you stop and say, 'Whoa! What is that all about?'"

Rivers in the sky

The clincher came from weather satellites equipped with microwave imagers. Unlike infrared radiation, microwaves are not absorbed to the same extent by water vapour in the atmosphere, so they can punch through all the way from Earth's surface to the imaging satellite. The images revealed that, summed vertically through the atmosphere, the greatest quantities of water were found not in blobs, but long, thin ribbons extending thousands of kilometres. Atmospheric rivers had simply been hidden: looking for them using infrared was like using your eyes to discern the bottommost layer of water in a steaming bathtub.

So what causes atmospheric rivers? The short answer is we still do not know. To weather modellers they are simply an "emergent" phenomenon. Program in a few basic physical facts about the atmosphere, such as the conservation of matter and momentum, the distribution of incoming solar radiation, Earth's rotation and the thermal properties of water, and out they pop.

In the northern hemisphere, we generally become aware of an atmospheric river when a cyclone – an anticlockwise-rotating low-pressure system – sweeps warm, moist air on to a coastline from the south and south-west. If winds are particularly strong about a kilometre up – a layer NOAA researcher Marty Ralph calls the river's "controlling layer" – vast quantities of water-laden air can pass over an area in a short time. If this stream hits mountainous coastal terrain, such as the Coast Range or the Sierra Nevada in California, it cools as it rises over the range, and its water vapour condenses into rain (see diagram). "That's where we get our big precipitation from," says Dettinger.

For the UK it is a similar, though less extreme, story: because the country is further from the equator, the air is usually already cooler and holds less water vapour by the time it gets there. In 2011, David Lavers, then at the University of Reading, studied the 10 largest floods in four UK river basins over the past 40 years, including particularly devastating floods that hit Cumbria in November 2009. "We decided to reverse engineer the floods," says Lavers. "We looked at the largest impacts and asked what caused them." In almost every case, archived wind-speed measurements and water vapour data suggested the presence of an atmospheric river (Geophysical Research Letters, vol 38, p L23803).

If you know what to look for, it is easy to spot an atmospheric river in satellite microwave images: typically there are half a dozen of them snaking above Earth at any given time. Many rain themselves out over the ocean without ever making landfall, and a typical river conveys as much moisture as seven to 15 Mississippis – or one Amazon. In the most part they are unproblematic: winds move the river about like a garden sprinkler head, allowing it to distribute its moisture over a large area. California receives one-third to a half of its precipitation in this way. Things get dicier when surrounding weather systems cause a river to stall in one place. "Then you see a real problem," says Ralph, and one not just confined to mountainous coastlines, either (see "Tennessee blues").

It is a problem that seems likely to grow. As far as we can tell, climate change has two opposing effects on atmospheric rivers. The temperature difference between the poles and the equator provides the ultimate energy source for mid-latitude storms. As the poles are warming quicker than mid-latitudes this temperature difference is getting smaller, and storms should weaken. But warmer air holds more water vapour, which could make atmospheric rivers even moister.

Receding snow line


Dettinger has used the same sort of general climate model that first exposed atmospheric rivers to evaluate which effect will be stronger in the western US. It suggests that atmospheric rivers will in fact form as often or perhaps slightly more frequently than today – but they will be moister. The peak season for atmospheric rivers might also lengthen. Because the air will be warmer, the snow line will be higher, and some precipitation that would today fall as snow in the Sierra Nevada will fall as rain instead, increasing the immediate flood risk downstream in places like Sacramento (Journal of the American Water Resources Association, vol 47, p 514).

Harsh experience suggests we should take note. The 1861-2 California flood killed thousands of people in an era when the state was much less densely populated than today, and it was by no means unique. Sediment deposits in the Sacramento river valley, near Santa Barbara on the Pacific coast and around San Francisco Bay provide evidence of comparable floods occurring in California roughly every 200 years.

Our ability to respond to a coming storm is currently limited. To react appropriately, we need to know how much rain will fall and in what river basins. But to create maps of water-vapour distribution and wind speed from microwave images, you need the background signal to be uniform – otherwise it is difficult to isolate which fluctuations are caused by atmospheric effects. While this is true over the wide expanses of the ocean, the varied nature of land cover currently makes reliable microwave sounding over land impossible.

This year NOAA, together with the California Department of Water Resources and the Scripps Institution of Oceanography in La Jolla, is hoping to fill those gaps with four dedicated atmospheric river observatories positioned along the coast of California. The first observatory, at Bodega Bay north of San Francisco, is scheduled for completion this month. Roughly the size of a dump truck, it contains a suite of standard weather instruments plus a "wind profiler" and a reconfigured GPS receiver. The profiler reflects radar off turbulence in the atmosphere to measure wind speeds at various altitudes, while the receiver mathematically inverts errors introduced into GPS signals by atmospheric water vapour to infer the amount of vapour the signal has passed through. "The instrumentation will provide real-time conditions," says California state climatologist Michael Anderson.

Each observatory costs roughly $750,000 – peanuts compared with the cost of flood damage. Lavers would like to see something similar in the UK. "If the observatory takes off, that will be a great motivation to push them through here," he says.

But the observatories will give only a few hours' warning – enough to open dam gates or issue flood warnings, but not much more. Dettinger thinks California needs to be doing more offshore reconnaissance to give earlier warning, allowing the organisation of evacuations, for example. "We live off satellite imagery to a ridiculous extent," he says. "We've got the whole of the Pacific Ocean covered by only a couple of weather ships", plus a few permanent weather stations around Hawaii. The US east coast faces a similar problem to track hurricanes that approach from offshore, but in this case satellite images are supplemented by a fleet of "hurricane hunter" aircraft, which measure the intensity of the storm and enable detailed predictions of its likely track.

A first step towards something similar in the Pacific was taken by NOAA in the winter of 2011. In collaboration with NASA, a retired spy drone was flown into three storms, including one atmospheric river. The project was part research mission and part technology demonstration: NASA was looking for useful things to do with the drones, and wanted to prove that they could deploy dropsondes. The main obstacle to more regular flights is money – and that means convincing the authorities that the technology helps. "We may have some more work to do, clearly demonstrating the impact of the observations on forecasts," says Gary Wick at NOAA.

One thing can be done without financial investment: raising public awareness. We now know the extent to which atmospheric rivers are responsible for the most extreme rainfall and the most severe floods, and are gradually getting a handle on how to spot them. When weather forecasters see a storm coming that is fuelled by an atmospheric river, they should warn the public of the flood danger, or "rattle some cages", as Dettinger puts it. It won't put an end to wet Christmases, but at least it will help Santa decide if he needs to put pontoons on his sleigh.


This article appeared in print under the headline "Skyfall"


Tennessee blues


On 1 and 2 May 2010, the city of Nashville, Tennessee, experienced its rainiest and third-rainiest days since weather records began there. This storm of storms dumped more than 30 centimetres of rain on the city itself, causing the Cumberland river to overflow into the streets. Almost 50 centimetres fell in some outlying areas. The cost of the damage totalled $2 billion, with 11 deaths in Nashville.

According to work done by Benjamin Moore of the US National Oceanic and Atmospheric Administration, this event was caused by an atmospheric river (see main story). The combination of a strong "Bermuda high" in the Atlantic Ocean and a low pressure trough along the east coast of Mexico funnelled a jet of moist air from the Caribbean Sea. On hitting the south-eastern US, uplift was provided not by coastal mountains, but by a squall line of thunderstorms parked inland over Tennessee and Kentucky. The warmer tropical air was forced to rise over the top and release its massive load of water (Monthly Weather Review, vol 140, p 358).

It seems to be an atypical atmospheric river: it did not dump its cargo on a hilly west coast, it was not winter, and there was no associated low-pressure cyclone. However, its combination of circumstances may be more common than we realise. Paul Dirmeyer and James Kinter of the Center for Ocean-Land-Atmosphere Studies in Calverton, Maryland, have even dubbed it the "Maya Express" in homage to California's "Pineapple Express". Similar conditions, according to Moore, may have contributed to large-scale floods in the central US in 1993 and 2008 – and perhaps elsewhere.

Dana Mackenzie is a freelance writer based in Santa Cruz, California



Source: www.newscientist.com