Glidemet

Tomorrow (Wednesday) will see rain, thundery at times, push up from the south through the day. Quite hard to say where, if anywhere, will see any soaring weather—-maybe the west midlands and Wales, maybe East Anglia for a while in the morning. I’m grasping really—-it’s not possible to make an accurate forecast in such a dynamic situation.

Thursday isn’t much clearer—-probably “showery”. Friday might be OK in the east but the low pressure system I mentioned yesterday will be winding up by now and beginning to affect the west.

The weekend looks awful, Saturday looks like this:

White isobars are sea level pressure and the grey blobs clouds.

Ladder thoughts
Interesting comments on last Saturday’s post if you haven’t seen them.

NB Yes, the bottom of this post (and some previous posts) is missing in IE6. I’ve tried to fix the issue but admit defeat. If you have the option, you really should use a modern browser (e.g. IE7, Firefox, Opera (my preference), or Safari) rather than IE6.

Stand-out flights from today have to be Dan Pitman’s and A.C. Wright’s 300s. It sounds like a couple of people were flying with Dan, and Mr Wright (Alistair?) did another 100 km after for good measure! Next longest flights on the ladder are around 160 km, which is a bit… odd.

The weather was very interesting as the north-south running front over western England unexpectedly began to wave (i.e. formed a wave-shape in plan view, not orographic wave) and then formed a cyclonic circulation through the day. This was cyclogenesis happening on our doorstep and was fascinating to watch on the sat loops, though it did mean that frontal cloud moved further east in a few places than expected, and the southwest was brighter. The low center formed over Carmarthenshire.

The reason for this was something of a kink forming in the tropopause, with dry stratospheric air being brought down to lower levels than usual. The following IDV image uses the 00z GFS visualised at 06z to illustrate the situation:

The grey isosurface is something called “isentropic potential vorticity”, with a value of 2 (representative of the “dynamical tropopause”, where the troposphere meets the stratosphere). You can see a finger of IPV poking down into the lower levels. The cross-section shows vertical velocity—-ahead of the IPV anomaly is red rising air (deepening the nascent depression); behind is blue sinking air. (You can also see more blue sinking air further east which was forms the capping inversion—-as the air sinks it compresses and warms (like how a bike pump becomes hot when inflating a tyre), so that rising thermals meet warmer air at height and can’t rise any further.)

The reason why IPV leads to vertical motion is fairly horrible but you can read it here if you want. Jo, if you’re reading this, didn’t you study this recently? Perhaps you can explain it  .

Tomorrow (Friday), unfortunately, the large mass of cloud just east of England at the moment is forecast to move westwards bringing cloud and patchy rain to eastern England. Showers are likely to develop ahead of the front over the northwest and Midlands into Wales. Best soaring is likely to be south of the Thames—Bristol Channel line and into southwestern England, where bases should go up to 4,000 ft. Winds surface 5 knots from west through to north, at height 10-15 knots from northwest.

There’s still mixed messages for Saturday, and ironically the models have reversed their positions somewhat (i.e. those forecasting bad now looking good and vice versa). It basically depends on whether or not the front which will hamper the east on Friday moves back eastwards or not, but either way western parts should have a good day. Sunday isn’t a certainty either as although the surface pressure pattern will be a ridge, a frontal system comes rather close to the west. Hopefully it won’t get too far.

Monday looks fairly reasonable.

Tuesday morning
I was fed duff info about what happened over central southern England on Tuesday morning, so want to set the record straight. While a circulation did form it a) wasn’t at the surface and b) was an end-product of another process, not the cause of the persistent rain itself. The reason why there was no surface low on the analysis chart was because it didn’t exist! What the actual cause was this:

That’s a drawing of the height of the 500 hPa pressure surface on Tuesday morning. The big blue dip in the background is an “upper low”, but running towards the viewpoint, over southern England, is an “upper-level trough”. Below on the surface you can see the rainfall rate (no scale I know, I’ll come back to that) and MSL pressure.

The next adds a PV isosurface:

As I had to stretch the vertical scale to illustrate the topography of the pressure surface, the PV isosurfaces are massively exaggerated! But you can see one bit going straight through the area of rain over southern England. The problem with the forecasts that morning was that the models—-all of them—-totally under-estimated the strength of the PVA. The result was that the warm, moist air in the front lying over England was lifted much more than forecast, so much more rain fell than models predicted (so it doesn’t matter that there’s no scale on the image above!).

Monday’s storm is an excellent example of what has several names in meteorology: rapid cyclogenesis (“RACY”), explosive cyclogenesis, or even “bombing”. All refer to the much faster and deeper development of a depression than is normal.

I spent a bit of time this evening creating some videos with the IDV to illustrate this event, but when I uploaded them to Youtube the results were, not to put too fine a point on it, crap. Instead I’ve uploaded them to box.net and will link to them directly—-you’ll be downloading Quicktime videos if you choose to view.

The first video shows the jetstream piling out of North America at over 140 knots (over 190 knots in the core) and crossing the Atlantic (time is from midnight Saturday until midday Monday), with sea level pressure below. If you look at the front left of the jet you’ll see, about half way through, the isobars south of Greenland start to tighten up. In the space of just 18 hours they go from a gentle U-shaped trough to a tightly-wound depression.

The second video shows the depression as the height of the 925 hPa level, this time viewing the North Atlantic from the northwest looking southeast (so the UK is in the corner closest to the viewer). You can see the deep depression suddenly appear half-way or so through and then drive towards Ireland.

The final video shows the results. This is a short series of four Meteosat IR images through Sunday showing the developing storm, overlaid with sea level pressure and surface wind barbs—-you can see the isobars tighten and become circular picking the wind up to over 60 knots. The effect can be dramatically seen in the data from an Irish bouy a few hundred miles out into the Atlantic:

While that graph is the combination of not just the deepening of the depression but also the depression moving in, the Met Office analysis charts show that the storm’s central pressure dropped 30 mb in just twelve hours.

Bernard Burton, who operates his own satellite image receiving station, has produced a couple of excellent visible-light views of the developing storm. The first is from a polar orbiter at 2pm on Sunday, while the second is from MSG1 at 6pm as it looks down on the storm in the evening light. Pretty dramatic!

The reason all this occurs is basically quite simple—-the front left of a jetstream (the “left exit) sucks air up, cooling it and making it become unstable, lowering the surface pressure. The opposite end (the “right entrance”) does the same thing, and you can actually see another deep depression coming off the eastern US seaboard behind the jet. (You can read a more detailed explanation in a previous post.) This particular jetstream is unusually fast, and so the depressions it creates are unusually deep.

The result is going to be a day with two very windy periods for southwestern and southern Britain, with occasional gusts of up to 80 mph and lots of 60 mph ones, no doubt doing a fair amount of damage. The first period will occur as the depression moves in and the pressure rapidly drops early Monday morning, while as you can see from the animations there’s a tight squeeze to the isobars on the rearward southwestern flank of the depression, which will bring another bout of windy weather as it crosses in the afternoon.

The low pressure is also arriving during some of the highest spring tides of year, and the resulting storm surge may well lead to flooding in a few susceptible places on the Cornwall and Devon coasts.

Perhaps the most amazing thing about this very-well-forecast storm is that from last Tuesday at the earliest until Sunday morning it only existed as the mathematical imaginings of a few big supercomputers. Now, in the space of less than a day, it’s become the very real violent movement of many millions of tonnes of air and water. And sadly, by this time tomorrow, it may well have killed several people.

Review and early outlook some time on Monday, hopefully.

I went out to observe it myself yesterday evening, and it was quite eerie to be looking up at the glowing halo above the horizon and know it was over 20 km up. Low-level mist obscured the view slightly and robbed it of any colour, but it was still quite a sight. However people who saw the glow both yesterday and on Monday report that last night’s was only 1/3rd the brightness.

If, like me, you missed Monday’s display a photographer on the south coast did a very good job of capturing the remarkable colours and glow.

The favoured explanation is Type 1 Polar Stratospheric Clouds (PCSs)—-nitric acid bound in ice crystals forming at -78C and lower at levels far into the stratosphere. The following image shows a “bubble” of air at or below -85C high over Britain (bear in mind that the IDV shows the vertical scale an order of magnitude larger than the horizontal—-the top of the box is 25 km up) at 6pm on Monday evening:

These low temperatures have been caused by the large high that was over us (now declining rapidly) pushing the tropopause (the boundary between the troposphere—-where our weather is—-and the stratosphere) and the stratosphere above upwards, like a dome. The following shows the 200 hPa level, roughly the tropopause (the top of the high):

Now, the tropopause is always higher towards the equator and lower at the poles, simply because the warmer air around the equator is “thicker”—-as air warms it expands. So in that image you can see the usual down-slope from the equator to the pole. What’s significant is the arrowed feature: this where the high pressure system has pushed itself under the tropopause from the southeast, lifting it upwards into a ridge-like structure. As the air above is lifted, it cools and if conditions are right hey presto, cloud forms. Cloud in this case being made up of very tiny ice crystals (only a millionth of a metre across).

For a look at the temperatures and humidities that were actually observed, here’s the very top of the Nottingham sounding that evening, at some 22 km (76,800 ft):

While the relative humidity is only 39%, at -89C and a pressure of just 1/30th of that at the surface water vapour can be deposited directly as ice crystals. In the very tiny space around any existing ice crystals, the relative humidity can reach 100% or more, allowing the crystal to spontaneously grow.

However ice alone can’t account for the incredible colours that were observed. That was, undoubtedly, the result of volcanic ash. Together the ice crystals and ash combined to produce the extraordinary sunsets and sunrises.

In fact, there were two processes at work. We’ve already established how the PSCs formed at some 20-25 km height, but these clouds reflect white light. What was causing the colours was a mass of Saharan dust that had been transported up from north Africa over the weekend, far below down in the troposphere. This dust alone would produce spectacular sunsets, but it was also causing only red light to pass through, resulting in the underside of the PSC cloud sheet to be lit up such spectacular colours.

The odds on these two events coinciding are… long! So if you saw it, count yourselves lucky. What isn’t so good is that the process that forms the nitric acid also starts a cascade of chemical reactions that ends with the destruction of ozone. The “ozone layer” is also at around 25 km, so that beautiful sunset you saw was also eating away the ozone layer…

Today was very difficult to forecast with the Met Office getting it quite wrong, along with pretty much everyone else! The rain extended much further northwest, was heavier, and there’s more snow around than there looked like being yesterday afternoon. The only model to get it right—-I’m told, as there’s no public access to it—-was the ECMWF. The Met Office’s UK mesoscale performed poorly, which probably explains why no weather warnings were issued until past five o’clock—-long after flooding had already occured in the southwest.

The GFS didn’t do a great job either, but enough to draw some pictures with the IDV showing what’s happening. This is the +6 hours forecast from the 12z run, so it was fairly well dialled in (though still underestimating the snow). The first image shows the height of the 850 hPa surface, showing the slightly elongated low firmly rooted over southeastern England. I could have just drawn a surface pressure chart, but that would have been boring  .

This next image adds a few more things. On the surface is total precip for the last six hours, and the heavy rain is clearly visible. Wind arrows at 925 hPa show the wind swirling around the low. The isosurface looks like something out of The Day After Tomorrow, doesn’t it? It is, in fact, theta-e showing the frontal boundaries—-warm to the right, cold to the left.

The weird stuff in the lower right is part of the occulsion process, with one airmass pushing in and over another.

The next image removes the theta-e isosurface and adds one for 5 deg C. You can see the cold air being driven in around the low from the northeast, meeting the rain and turning it to snow on it’s northern side. Obviously it needs to be colder than 5 deg C for snow (and a few other things), but the GFS isn’t handling that part too well. It shows the colder air a treat though.

So that’s tonight, what about tomorrow’s (Saturday) forecast? A ridge will move in (visible as the red area in the near left corner of the top picture), as forecast all week, giving a bright mostly sunny day in the east. Some high cloud from the next front will be spilling into the west, although it will stay dry except for the far west later in the afternoon. Wales and the west still look good for some wave.

Sunday will be a cloudy damp day with increasingly heavy rain spreading from the west. Monday will be briefly brighter again (maybe showers about), while Tuesday looks very wet again.

I thought it would be interesting to take a look at what happened last night, both down here in the south, and up in Scotland where there were severe gales.

All the images were produced using the IDV and GFS data for midnight.

This image shows some of what was happening. Up above is an orange isosurface of the jetstream. All the air within that surface is travelling at 90 knots or more. It shows the deep U-shape of an upper-level trough, and the speed shear on the outside of the bend is increasing the voriticity of the air. On the leading side it is adding to the planet’s rotation, and on the trailing side cancelling it out, so on the leading side it’s increasing the spin of the air (and thus drawing air in and up). The red isosurface is air that’s rising rapidly. It’s not possible to say how much this is due to the increased vorticity or actual convection within the front—-it’s the sum. The grey area is cloud, and on the surface is total rainfall over the last six hours (heaviest over the western hills) and surface pressure.

Zooming in and adding a cross-section through the front shows the drier colder air behind the front running into the warmer damper air ahead. This is the basic frontal process—-like a battle front between two armies. The jet stream above is enhancing the effect and making the weather along the front more severe.

That’s all well and good for the south but what’s all that in the background over Scotland? Lots of red rising air and grey cloud, plus those surface isobars are quite closely packed (i.e. it’s windy) and it’s obviously raining heavily. What’s causing the air to rise over Scotland? Once again it’s vorticity. The following is a plan view of relative vorticity with red showing where the air is being spun up and drawn in:

To be clear—-the vorticity is resulting in the general uplift of the air over a wide area, which cools as it rises. As it cools it becomes unstable, allowing convection. The red rising air in the 3D pictures is the sum of both processes. Looking at voriticity helps us to seperate the two.

Where’s the vorticity coming from? If you look back at the first 3D image you can see that some of this area is in the “left exit” of the jetstream—-always the point where uplift is strongest. On the left side of the exit the air diverges (pulling air up from below) and on the right side converges (forcing it down). However the jetstream is also doing something really weird to the tropopause, the boundary between the troposphere (where all our weather is) and the stratosphere (dry air above). The tropopause is shown in the next image by the magenta surface:

You can see that the tropopause has folded back on itself, and that the fold is colocated over the area of highest relative vorticity. The dry stratospheric air is intruding down into the troposphere, and air drawn down from the stratosphere really wants to spin. The reason why can only be explained mathematically (Jolien, if you’re reading this, you should be able to work it through ). Sometimes the dry cold (and thus dense) stratospheric air begins to move extremely quickly and a small but powerful “sting jet” of hurricane-speed air can develop in the heart of the storm. If this reaches the surface the winds can be very damaging. The 1987 Great Storm was actually a sting jet, and that was first time one had ever been observed. Sting jets are still an area of active research today.