Computers are no longer getting faster, they are just getting bigger. So in order to be able to do more detailed and accurate weather and climate simulations, models must be able to run efficiently on massively parallel computers. There has been alot of effort recently optimising the Met Office Unified Model (UM) (unified for weather and climate) which can now run efficiently on 14,000 processing cores. But this is still not enough. More detailed climate predictions may need 100,000 or even a million cores. The GungHo project is joint between the Met Office, a number of NERC funded academics and STFC to design a new model without the bottlenecks that are currently limiting model performance.

Why won’t the existing model scale to more processing cores?

In order to simulate the climate on a large number of cores, the atmosphere and ocean are decomposed into lots of regions and each region is simulated on a different core. These must all communicate so that the weather in one region influences neighbouring regions. A model scales well if, when you decompose the model domain into, say, 10,000 regions and run on 10,000 cores, the model runs nearly 10,000 times as fast as if you ran on one core. One of the first bottlenecks is input and output (IO). When the model finishes a period of its simulation and needs to write out the results, all the data is sent to one processor to be written out to disk. This can be overcome using temporary storage local to each processor. There are also bottlenecks that are more fundamental to the design of the climate model and these are the subject of research during the GungHo project.

Problems with the Latitude-Longitude Grid

The weather in one location is influenced by its surroundings. The size of the “surroundings” is dependent on the model time-step. Towards the poles of the latitude-longitude grid many grid-points are clustered together so if I live at 89 degrees north, many grid-points will influence my weather. These grid-points may be distributed over many different processing cores and so each must communicate their weather to my core in order for my weather to be predicted. This inevitably leads to a lot of communication and so the forecasting models on each core end up waiting for data from their surroundings. Consequently, various quasi-uniform grids are being considered and studied for the next generation Unified Model (see figure).

Semi-implicit time-stepping

The efficiency of a model is related to the largest possible time-step that can be taken. If a very large time-step is used then the model will finish its forecast more quickly. If it doesn’t crash in the process. Large time-steps can make models blow up if you are trying to propagate a wave or move some mass by more than one grid box in a time-step. The fastest waves simulated by climate models are sound waves. We do not want the model time-step to be restricted by the time taken for a sound-wave to travel from one grid-box to the next. So instead of solving the sound-waves explicitly, moving them by just one grid-box every time step, they are solved implicitly, which requires a simultaneous solution over every grid-box of the equations governing their motion. This removes the time-step restriction based on sound waves which enables a model to take much longer time-steps and therefore complete its forecast more quickly. But the simultaneous solution over all grid-points means global communication which will limit parallel scaling. In GungHo we are investigating linear equation solvers which are more efficient on many cores (to solve the simultaneous equations) and we are also considering only solving the equations simultaneously in the vertical direction and not the horizontal direction. The vertical grid-spacing is much smaller than the horizontal grid-spacing (by perhaps a factor of 1000) so if we can remove the time-step restrictions based on the speed of sound in just the vertical direction then this might be sufficient to allow long enough time steps to finish the forecast in time while not requiring global communication which slows down a model running on 100,000 cores.

Semi-Lagrangian advection

To be able to take even longer time-steps and thus finish a forecast even more quickly, semi-Lagrangian advection is used in many advanced weather forecasting models. Rather than propagating information from one grid-box to the next, the properties of the atmosphere (temperature, velocity, moisture content etc) are moved a long distance in one time-step. In order to predict the conditions at any grid-box at the new time-step, the winds from the old time-step are used to calculate an approximate departure point. Ie where did the atmosphere come from to arrive at this grid-box at the new time step. The conditions at the old time-step are then interpolated onto the departure point and moved to the current grid-box. Semi-Lagrangian advection allows much longer, stable and accurate time-steps. But it seems that any advanced technique that allows longer time-steps will also limit parallel scaling. The departure point may be on a different processing core or even (near the pole) not the next core along but the one after that. Therefore alot of communication is needed for semi-Lagrangian advection and the advantages may be lost when moving to massively parallel computers. The GungHo project is therefore investigating alternatives, such as “Auxiliary semi-Lagrangian” which basically means advecting the properties one grid-box at a time but using lots of small sub time-steps.

It is possible that the next version of the UM will still use semi-implicit time-stepping, semi-Lagrangian advection but will use the Yin-Yang (overlapping) grid (see figure). The one after may look very different, perhaps on an icosahedral grid with an option for implicit time-stepping only in the vertical and without semi-Lagrangian advection.

by Keri Nicoll

Radiosondes play an important role in modern day atmospheric measurements, providing high vertical resolution data (typically 5m) that cannot be obtained by other means.  Data from radiosonde flights is assembled into numerical weather prediction models and used by forecasters to assess movement of weather systems, cloud heights and predict fog.  Nowadays data from the upper atmosphere is relatively easy to obtain, but the development of technology to get us to this point has been long and arduous….

The first measurements of atmospheric parameters above the surface were made by kites in the mid 1750’s, starting with the work of Alexander Wilson, who carried a train of mercury thermometers, mounted at different heights on a kite string up to height of ~1km.   Kites were adopted as the standard method of measuring above the surface (see Figure 1) by several early meteorological services (including the US National Weather Service up until well into the 1800s), as the only alternative at the time was the manned balloon.   Manned balloons (which were first created in the 1780s by the Montgolfier brothers) provided a more stable measurement platform than kites, with the capacity to carry heavier equipment, as well as human observers.  However, the occupation of an aeronaut was a relatively dangerous one, with many meeting an early end through oxygen deprivation from flying too high (some have survived to altitudes as high as 10km without extra oxygen), or falling from the basket of the balloon.  Despite the dangers, many successful atmospheric measurements were made from these platforms, including those of Henry Coxwell and James Glaisher, who from 1862 to 1866 made 27 successful flights measuring pressure, temperature and humidity.  The expensive nature of manned balloons, as well as height limitations imposed by the presence of human observers did not make the manned balloon a particularly efficient platform for atmospheric research.    During the 1880s meteorologists had begun to experiment with free flying paper balloons, using their movement to determine wind speed and direction.  This idea was further developed by Hermite and Besacon (in 1892), who launched the first instrumented unmanned meteorological balloon made of gold beater’s skin, carrying a meteorograph to measure temperature.

Figure1.  An early pioneer of airborne meteorological instrumentation, W.H. Dines from the UK Met Office, waits to launch a kite (1907). Taken from W. Pike, Weather, 60, 308–315, November 2005.

By the turn of the 20^th century, unmanned meteorological balloons had become the standard tool for upper air measurements, which lead to the discovery of the stratosphere ( e.g Teisserence de Bort  and others).  The main limitation with these measurements was that since there was no known method to transmit the data back to the observer, the instrumentation had to be recovered after each flight, therefore real time analysis was impossible.  This changed with the invention of radio, enabling  French scientists Idrac and Bureau to successfully obtain radio signals from an unmanned balloon that reached the stratosphere.  Thus the term “Radiosonde” was coined, where “sonde” is French  for “probe”.  The early definition of a radiosonde was a balloon-borne payload that transmitted atmospheric parameters to a ground receiver via radio, and this definition has not changed to this day.  This paved the way for a spate of early radiosonde flights, including Bureau, who flew a bimetallic thermometer and barometer in 1929, and Molchanov, who launched the first operational radiosonde in 1930, where the data was relayed back to a Russian forecast center.  Many worldwide meteorological services developed their own radiosonde systems following this, including the UK Met Office, who created the Kew MK1, MK2 and MK3 radiosondes, which were used in the UK between 1939 and 1988.  The most successful design of radiosonde has been that of Vaisala from Finland, who’s radiosondes have been in use since the early 1930s (see Figure 2), and are now the most widely used radiosondes worldwide.

Figure 2.  One of the first meteorological radiosondes (Vaisala 1931)

Nowadays there are many commercially available radiosondes (Figure 3), all of which measure pressure, temperature and relative humidity as standard.  The position of the sonde can also be measured, either by LORAN, or more recently by GPS.  A typical maximum height for a radiosonde ascent is ~30km, although this depends strongly on the size of balloon, volume of gas used (either helium or hydrogen), and the weight of the payload.  Balloons drift with the wind and can often cover much larger horizontal distances than they travel vertically – typical horizontal distances can be around 100km.  Data is often available on the descent stage of a flight (which is slowed by parachute after the balloon bursts), provided the radio signal is strong enough to detect. Thus a radiosonde is capable of providing two profiles in separate locations on the same flight.

Figure 3.  Selection of modern day commercially available radiosondes.

Although radiosondes only measure pressure, temperature and humidity as standard, some can be configured to accept data from additional research sensors.  Such additional sensors have included ozone sensors, which are widely used throughout the world for ozone monitoring – the UK Met Office launch one ozone sensor per week from their Lerwick station.  In addition, considerable research has been carried out into thunderstorm electrification in the US via additionally instrumented radiosondes.  The group of Marshall and Stolzenberg at the Univeristy of Mississippi has developed an electric field sensor, and charged precipitation sensor, which have made numerous flights into the heart of thunderstorms in an effort to understand the typical charge structure, as well as charge generation mechanisms in thunderstorms.  A balloon platform is particularly useful in this situation due to the substantial hazards to aircraft.

Over the past few years we have been developing at suite of additional science sensors at the University of Reading, as part of the MORSE project (More Operational Radiosonde SEnsors) with Prof Giles Harrison.  These sensors are designed to attach easily to the side of a standard meteorological radiosonde (see Figure 4).  Hundreds of radiosondes are launched around the world everyday, therefore an additional sensor package that fits easily onto an existing radiosonde flight can provide a wealth of extra data for little extra cost.  There are several constraints involved when designing instrumentation for radiosonde use including low cost, as the instrument is disposable, low weight (few hundred grams), low power (<30mA), and temperature stability as the typical temperature range encountered on a flight is +20 to -70°C.  In addition, a data acquisition system is required to interface the extra science sensors to the radiosonde, to allow the data to be transmitted via the radio link synchronously with the standard met data.  We have developed such a data acquisition system, which has been flown many times, with a variety of different science sensors.  These include sensors for turbulence, electrical charge, solar radiation, ionisation, aerosol particles and cloud droplets, some of which are now described.

Figure 4. Photo of Vaisala RS92 radiosonde instrumented with extra sensors developed at the University of Reading.  The additional sensor package includes a data acquisition board, solar radiation and turbulence sensors.

The disposable turbulence sensor uses orthogonal geomagnetic field sensors (magnetometers), to detect rapid motion during a balloon flight. The variability in each magnetometer channel is used to detect motion (see Figure 5 for a flight through a cloud layer). As well as sensing potential atmospheric hazards, information obtained from this technique has already been applied in understanding the turbulent motion of the Huygens planetary probe descending to Saturn’s moon, Titan (see http://www.nasa.gov/mission_pages/cassini/media/cassinif-20070828.html for more details).

Figure 5. Vertical profile of variability in magnetic field from University of Reading turbulence sensor, plotted alongside RH measured by a  Vaisala radiosonde.  The increased RH at 600hPa denotes a cloud layer in which substantial turbulence was present.  Taken from Harrison et al 2007, Rev. Sci. Instrum. 78 (12). p. 124501.

Aerosol particles in the atmosphere are of ever increasing importance to atmospheric researchers, and was emphasised by the eruption of the Eyjafjallajökull volcano in Iceland during April and May 2010.  A balloon borne aerosol particle counter, developed at the Universities of Hertfordshire and Reading, was launched through the volcanic plume on 19^th April, from Stranraer, Scotland.  The aerosol particle counter measures particle size and concentration using light scattered from particles drawn into a sampling chamber by an air pump.  Particles are detected in five different size bins from 0.6 to 10.6µm.  The balloon flight took place during the flight ban when UK airspace was closed, and demonstrates the usefulness of instrumented balloons in locations deemed too dangerous for manned aircraft to fly.  A 600m thick layer of ash was detected at 4km altitude, with a maximum particle concentration of 100 cm^-3, and mean particle diameter of 1.4 µm (see Figure 6).

Figure 6. Vertical profiles through the ash plume of  Eyjafjallajökull, over Scotland.   (a) aerosol particle concentration   and (b) electric charge from balloon borne particle counter and charge sensor respectively.  Adapted from Harrison et al, Environ. Res. Lett., 5, 024004, 2010.

The characterisation of atmospheric dust is another area that has been investigated using instrumented radiosondes.  Dust particles contribute to the radiative balance of the atmosphere by scattering and absorbing radiation, therefore it is important to understand the transport of dust throughout the atmosphere.  Observations hint that dust particles can sometimes become vertically aligned, possibly due to electrical effects, which can alter the transfer of solar and terrestrial radiation through a dust layer.  It is therefore useful to determine whether elevated layers of dust are substantially electrically charged, for which balloons provide a suitable measurement platform.  The Saharan desert is the largest producer of mineral dust on Earth, therefore a good location to investigate atmospheric dust is the Cape Verde Isles, just off the west coast of Africa, which are frequently afflicted by Saharan dust outbreaks.  In the summer of 2009, several instrumented balloon ascents were made from Sal, Cape Verde, through elevated layers of dust.   These research radiosondes were instrumented with the data acquisition system, charge sensor and aerosol particle counter.  Two of the balloon flights measured large concentrations of dust particles in layers up to 4km altitude, which coincided with increased levels of charge, showing that the elevated dust layers were electrically charged.

Figure 7. Vertical profile of (a) temperature (grey) and RH (black) measured by a Vaisala radiosonde, (b) solar radiation measured by the University of Reading solar radiation sensor.  The solar radiation is seen to increase markedly as the sonde passes through the cloud layer, and the transition from cloudy to clear air at the cloud top is seen in much finer detail from solar radiation measurements than from the standard RH measurements.

Recently we have focused on developing a range of optical sensors for use alongside standard radiosondes, comprising a solar radiation sensor and cloud droplet sensor.  These sensors provide finer detail about the transition from clear to cloudy air that occurs near cloud edges than thermodynamic measurements made by conventional radiosondes can provide (Figure 7).  The continued use of the long-established radiosonde as a research tool is limited only by the availability of suitable inexpensive sensors.  Consequently the development of novel types of balloon borne meteorological sensors will continue here at the University of Reading, with the aim to provide the existing operational radiosonde network with a range of specialised sensors, opening a new network of scientific measurements.

In January this year a new measure to tackle climate change came into force.  Designed by the European Union, it targets the 220 million tonnes of CO₂ emitted on flights departing from or arriving to a European airport annually (figure from 2006).  The measure is to include aviation CO₂ emissions in the EU’s emissions trading scheme; the aim is to achieve real reductions in the CO₂ emitted by this fast-growing industry.  Put simply: the cost of a flight, for both airline and passenger, now includes CO₂.

Why is there a need for such a scheme in the first place?  In 2006, globally, aircraft emitted around 700 million tonnes of CO2 into the atmosphere, 30% of which was from flights originating or departing Europe. Putting this into context, global aviation contributed approximately 2% of man-made CO2 emissions (in that year).  This is a small proportion; the UK’s share of aviation CO2 emissions are much less than the contribution from, for example, heating our homes (14.8%) or generating electricity (about 26%).  So, why all the fuss?  First, the aviation industry is growing by around 5% per year, meaning that its share of CO2 emissions could rapidly increase.  Second, the climate impact of aviation is not just from CO2 emissions; the non-CO2 climate effects of aviation (like water vapour, ozone creation and contrails) increase aviation’s total contribution to human-induced climate change to up to 15%.  The good news is that the aviation industry isn’t sticking its head in the sand.  It has set its own stringent targets that new aircraft entering service in 2020 should produce 50% less CO2 (per passenger kilometre) than aircraft in 2000. The question is, left to its own devices, will the aviation industry really be able to make a significant dent in its CO2 emissions?  The EU thinks not.

Now for the technical bit: the ETS limits (‘caps’) overall CO2 emissions in Europe from certain industries; under this cap and trade scheme, companies buy carbon credits to cover their CO2 emissions up to the cap, and may trade surplus carbon credits on the carbon market.  So how does this work for airlines?  The total amount of CO2 that can be emitted by aviation is now capped at 97% of what the annual average that was emitted in 2004-2006; this includes CO2 emissions from all flights which arrived at or departed from a European airport (i.e. it counts the total CO2 from the flight, not just from the portion of the flight that was in European airspace).  85% of this cap is distributed to airlines as free allowances, proportional to that airline’s share of the emissions in 2010.  The remaining 15% of the cap will be auctioned.  At the end of April 2013, each airline must surrender sufficient carbon credits to cover their 2012 CO2 emissions.  Therefore if an airline wishes to expand its operations in Europe, it must either buy carbon credits at auction (or trade from other airlines who have reduced their emissions and thus have surplus credits) or from another industry sector.  The implementation is fairly complicated but the rational behind the scheme is simple: the less CO2 you emit, the smaller your costs, and in a competitive market this should effectively drive CO2 emission reductions.

So far so good.  But that’s not the end of the story.  Internationally, the inclusion of aviation into the EU ETS has been highly controversial; the EU’s decision to introduce a regional scheme is seen by many as taking unilateral action that is both unfair and counterproductive. Legal action, brought by US airlines, was defeated by the European courts in December, however the US government may still make it illegal for US airlines to comply with this new EU law (interestingly, Delta have coincidentally introduced an unspecified $3 per passenger surcharge, ‘just in case’?).  China doesn’t want its airlines to pay, and has variously threatened to cancel a large order of (the European-based) Airbus aircraft, or to refuse to allow its airlines to comply.  UK airlines, with the noted exception of Ryanair, have been mostly supportive; BA have previously voluntarily participated in a UK-based ETS.   It is expected that airlines will pass on the costs of participating in the ETS to its passengers in the form of surcharges; so far surcharges have been at the bottom end of the range of estimates of 2 Euros to 3% of the ticket price, and in any case will be insignificant compared to UK air passenger duty (currently £60 for a flight to New York).

It is unclear how this story will end.  At a debate I attended at British Airways last December, the overwhelming majority agreed that market-based measures are the best way to tackle CO2 emissions, even if few people believed that the EU would manage to implement the aviation ETS without some concessions or modifications.  Some concessions seem likely as international opposition to the scheme is not only increasing but becoming more organised: a coalition of countries (including the US, China, India and Russia) will meet next week to decide on ‘retaliatory action’ against the EU, threatening to escalate the situation into a full-scale trade war.  Meanwhile, aviation’s governing body ICAO says it is accelerating efforts to design a global ETS-style measure for aviation and will present its proposals by summer (full story here).  Even the EU acknowledge that a global solution is clearly the best way forward.  The EU ETS could just be the catalyst to make that happen.

Geoengineering is not going away any time soon. It is beginning to move from a fringe topic into the mainstream, and a genuine research community is starting to set the agenda for future research. ‘Climate engineering’ or ‘geoengineering’  is an umbrella term covering methods by which humans can modify Earth’s climate: carbon dioxide removal (CDR) or decreasing the amount of incident solar radiation (solar radiation management; SRM). In August 2011 the Banff Centre in Canada hosted the Second Transdisciplinary Summer School on Climate Engineering. The aim of the summer school was to develop interdisciplinary research skills and to bring together the younger members of the nascent research community.

Banff from Sulphur Mountain, Alberta, Canada

Currently the geoengineering discourse is monopolised by a few. In the words of the school’s organiser David Keith, ‘the same people keep popping up’. It is never healthy to have a field dominated by a small group of people, no matter how individually brilliant and well-meaning they may be. The summer school sought to provide students with the tools to get involved in the discourse. A strong research community helps us get our research right; that means it must be timely, transparent and relevant. This requires information exchange between natural scientists, social scientists, policymakers and the general public.

The central problem of geoengineering (especially by SRM) is that of risk analysis. Conventionally we might weigh cost against benefit. However, according to David Keith there are some geoengineering schemes which can be viewed as essentially costless. For example, injecting aerosols into the stratosphere would cost (Robock et al 2009) tens of billions of US dollars per year. In this case monetary cost is not the limiting factor in decision-making and the trade-off is between the risks of action and the risks of inaction. These risks can operate on different spatial and temporal scales, and also on different groups. For the sake of argument, imagine that not injecting aerosols poses a low risk (from climate change) to economically developed countries through loss of key port cities; and that injecting aerosols poses a high risk to regions of Africa relying on monsoon rainfall for their agriculture. The rational risk-risk decision would be not to act, but the more powerful influences could decide to act and maximise their self-interested benefits. The challenge of geoengineering governance is to ensure decisions reflect the interests of all parties and not just influential minorities.

Prof. Granger Morgan also spoke about analysis of climate risk. Conventional risk analysis involves putting a value on positive and negative effects and making a decision based on the balance of these effects. This involves certain assumptions:

• There is a single decision-maker.
• The values we hold are known and static.
• The decision-maker maximises utility.
• There are manageable impacts which can be valued.
• The discount rate (the rate at which the future is considered to be ‘irrelevant’) is exponential – that is, the future becomes increasingly ‘worthless’.
• The system is linear and predictable.

Morgan emphasised that none of these assumptions were satisfactory when considering either geoengineering or climate change. He asked whether we were deluding ourselves into certainty using cost-benefit analysis on a problem where it was not appropriate. The question still stands.

So what are the risks? Here is David Keith’s offering, in order of severity:

1. Conflict. Our technology for developing reflective aerosols is advancing rapidly and the cooling potential is very high. SRM also acts very quickly, with cooling happening within months. Finally, the impacts will be felt more in some areas of the globe than others. These three factors suggest a high risk of conflict over SRM technology. SRM was in an ideal world is relatively safe, but adding in human conflict and folly the worst-case SRM scenario is at least as bad – or perhaps worse – than unmitigated anthropogenic climate change.
2. Accelerating the end of nature. Once humanity has the ability to control the amount of solar radiation incident on the planet, it is on a slippery slope towards planetary management. The temptation to attempt to fine-tune aspects of the Earth system to our own profit will be strong. The evidence from case studies of human management of planetary systems suggests we might not do a good job.
3. Moral hazard. Having the option of fast-acting SRM may reduce the public and political will to take the long, slow path of emissions reduction. It does not matter that SRM is an incomplete solution as long as the perception is that it would offer an easy way out.
4. Environmental risks. These include ozone depletion, impacts on tropospheric clouds from aerosol particles settling out of the stratosphere and impacts on the global hydrological cycle. Keith did not dismiss these risks, but claimed they were relatively small and could be dealt with. Hence they were his least important concern.

These risks are at the core of decision making regarding geoengineering, and researchers from a wide range of disciplines are beginning to unpick them. The process of guiding scientific research may soon become more formalised as a result of a programme governing geoengineering research. This was a common item on the ‘wish list’ of many at the summer school – a global programme managing research goals. We might also expect ethical norms to be developed, much like there have been in stem cell research.

Such a programme would be able to co-ordinate research efforts to answer key scientific questions. Two questions I noticed being asked repeatedly concerned the identification of climate tipping points (could SRM be implemented in a ‘climate emergency’, and would be know it if we were in one?) and the regional climate impacts of SRM technology. These are the two areas where climate scientists can bring their expertise to bear. There is, however, one more important role scientists can play: that of the rational sceptic. There are many people out there who are losing their heads over geoengineering. Either they think it is fundamentally evil, or they think it is something we should be doing right now. It is easy for voices of caution and reason to get lost in the hubbub.

What the future holds

Geoengineering is an expression of a problem humanity has struggled with since it developed organised societies. How do we manage our impact on the Earth system? Initially it started out as small-scale resource management, such as crop rotation to avoid soil degradation. Then we effected regional environmental change through deforestation. Now we have the technology to affect the entire planet. The question for the future is: what is our place on this Earth?

A population of humans inevitably shapes its environment to suit its needs, but until now this has been on a relatively small scale (sub-global) or an unintentional side-effect. This might change in the future as we understand the magnitude of our potential impacts on the Earth system. This Pandora’s Box, once opened, will never be shut. We must learn to deal with that which is contained within.

Welcome to a new term for the WCD blog with our usual mix of recent weather and climate and some of the research and other work that other members of the department are working on. Blogs are usually published on Friday.

Week 1 (20th Jan) – Angus Ferraro
Week 2 (27th Jan) – Emma Irvine
Week 3 (3rd Feb) – Keri Nicholl
Week 4 (10th Feb) – Hilary Weller
Week 5 (17th Feb) – David Lavers
Week 6 (24th Feb) – Jon Robson
Week 7 (2nd Mar) – Andy Turner
Week 8 (9th Mar) – Jon Robson
Week 9 (16th Mar) -
Week 10 (23rd Mar) -

As part of their final practical session for the course MT11C Introduction to Weather and Climate, a group of first year undergraduate students conducted a synoptic analysis of the extra-tropical cyclone Friedhelm. This storm brought strong winds and disruption to Scotland and Northern England over the 8th and 9th December, 2011. You can read more about the research flight that some members of the department took through the storm as part of the DIAMET campaign here.  The analysis was conducted in small groups by the students listed below, assisted by demonstrators Angus Ferraro, Kirsty Hanley and Kaah Menang

Modis visible composite of Friedhelm as it crossed the UK (courtesy University of Dundee)

Surface synoptic development (Jon Baker, Sara Jackson and Liana Lucatniece)

A strong surface low pressure system developed over the Atlantic to the north-west of the UK. The surface analysis shown below, at 00UTC on 8th December shows a system with a central low pressure of 985hPa. The system moved south-eastwards towards Northern Scotland and strengthened with central low pressures dropping by 20hPa to 965hPa by 06UTC. By 12UTC, the system developed further and moved over the tip of the UK. At this point the pressure reached it’s lowest analysed value of 960hPa, a deepening of 25hPa in 12 hours. The centre of the low pressure passed over the UK by 18UTC, however a significant region of high surface pressure gradient remained over the UK well into the 9th Decemeber, with very strong wind gusts (up to 165 knots) observed at Cairngorm in Northern Scotland.

00UTC Thursday 9th December

12UTC Thursday 8th December

Surface wind observations (Jonathan Beverley, Emma Camp and Amadou Jawara)

Analysis of the 12UTC surface weather reports showed that the areas with the strongest average winds and highest wind gusts were in southern and western Scotland. The maximum gust was over 70 knots and the highest average wind speed of 54 knots for the same station. The area of highest wind speed was to the south of the centre of the low and corresponded to the area on the surface pressure charts where the isobars were closely compressed giving a high pressure gradient. The gusts seen at 18UTC were still up to about 70 knots but were much less concentrated. This means that between 12UTC and 18UTC there were recurring gusts of around 70 knots which would have caused the damage to power lines that was seen.

Observed surface wind gusts 12UTC Thursday 8th December

Air Masses associated with the storm (Cassie Elswood, Jack Marshall, Luke Storer, Harriet Turner)

The weak high pressure to the south of the storm is associated with a flow of warm Tropical Maritime air into the warm sector of the cyclone. The warm air (with temperatures up to 14 degC) meets very dry, cold arctic maritime air at the surface cold front (temperatures in the cold air mass range from -2 degC to -17 degC). When this very cold air encounters a warm land surface it can lead to unstable conditions. There is also a flow of dry cold air over the occluded front. The causes the ice clouds at the occlusion to evaporate, creating a region of dense air and bringing very strong winds to the surface.

Schematic of surface features and air masses associated with storm

Fronts associated with the storm (Peter McAward, Ben Powrie and Jess Welford)

Five day forecasts of the storm showed snow as far south as Birmingham, however this did not happen; Scotland did see snow in the mountains but any snow that fell in England did not settle. Instead it fell as a mixture of rain and snow. The analysis thickness chart below, for 00Z on the 9th (about 12 hours after the frontal passage) showed that the relative position of the occluded front and the 528dm line was not conducive for snow ahead of the occlusion. This makes sense, because according to the Met Office analysis, the occlusion is a cold occlusion with the coldest air behind the surface front. Earlier forecasts showed the 528dm line behind the occlusion, leading to the forecast of snow. Later forecasts changed the position of the 528dm line, moving it further north ahead of the occlusion, more in line with the observed conditions.

UK Met Office thickness analysis for 00UTC Friday 9th December

Vertical structure of the storm and development (Martin Baumber, Ty Buckingham and Luke Poundall)

The cyclone followed well the standard textbook example. At 12UTC on December 8th, the surface low was situated in Northern Scotland, with the upper trough at 500hPa located to the west of the surface low. The upper low (or short-wave trough) shows a typical patern of convergence and divergence in relation to the upper-level jet stream. The region of divergence at 200hPa is directly above the surface low and is one of the reasons that the mid-latitude cyclone strengthens. At 500hPa, isotherms show strong cold advection behind the surface low, further acting to deepen the system. The radiosonde ascent from Castor Bay shows moist air up to around 650hPa with a cold dry region above, associated with the dry, cold airstream behind the surface cold front.

Analysed upper level structure at 200hPa (top) and 500hPa (bottom) 12UTC Thursday 8th December

Schematic showing vertical structure of Friedhelm

Forecast Validation (Ian Rollins, Matt Rout and Ed Williamson)

The Met Office forecast was accurate up to around 84 hours before the event on Thursday the 8th and Friday 9th Decemeber. The forecasts were accurate in positioning the surface low. The 120 hour forecast did not represent what acutually happened, but forecast skill at five days is typically limited. The 84 hour forecast for Friday predicted that the surface low would have a central pressure of 964hPa and the analysed value was 970hPa. The only major difference between the forecasts and the analysis charts is the placement of fronts, this could be due to limited observations of the surface fronts while the system was over the Atlantic.

At the beginning of term, Tom Frame contributed this excellent post about the ongoing DIAMET field campaign.  In this blog entry, I’ll provide an update on some of the more exciting and stomach-turning developments from the past few weeks of flying into storms around the British Isles.

Backing up for just a moment, you may recall that the purpose of DIAMET (DIAbatic influences on Mesoscale structures in ExtraTtropical storms) is to improve our understanding of and ability to model the interactions between cloud microphysical processes and mesoscale flow structures in midlatitude cyclones. What does this mean? In and around  clouds, water changes phase between ice, liquid, and vapour. The phase transitions are accompanied by heating and cooling of the air, sometimes over very large regions such as along a synoptic scale front. All of this heating and cooling leads to changes in temperature and pressure, which in turn can influence the flow of air in and around clouds. Many questions remain unanswered concerning how the interaction between diabatic heating and mesoscale flow influence the evolution of heavy precipitation and severe wind events. In a numerical weather prediction model, this interaction involves communication between subgridscale processes and the explicitly resolved flow. Does a model represent this interaction correctly? This is a difficult question to answer, but a field campaign such as DIAMET can help us to find out.

Over the past three weeks, the FAAM (Facility for Airborne Atmospheric Measurement) BAe-146 research aircraft has been flying into storms and collecting data that will be analysed to determine microphysical heating and cooling rates as well as the dynamical structure of the storms. When the campaign was planned over a year ago, we couldn’t possibly know if the weather regime would lend itself to supporting the kinds of storms that we wanted.  Fortunately for everyone who has invested their time and energy in this project, the weather did indeed comply. For a complete run down on all of the field campaign happenings, visit the DIAMET blog.

On 8th December , our very own John “The Intrepid”  Methven led a mission into the centre of an explosively developing cyclone that appears to have produced a sting jet. This was quite a rare opportunity to sample one of the most violent types of cold season wind phenomena. The Scottish Sun has written an article about the flight here, and the Met Office issued this press release. Rumor has it that the 160+mph winds and associated turbulence were too much for one Reading scientist who flew on this mission. Some lunch may have been lost, but we won’t be telling the Sun about that. Before you pass judgment, have a look at this video posted to Youtube by a member of the flight crew on the 1st December mission north of Scotland :

Flying at 150 ft above 70 ft waves

The purpose of this particular sortie was to collect enough data to infer the heat and moisture flux from the ocean surface into the boundary layer of an intense cyclone.

In summary, both the September and December flying periods were a big success. The weather gifted us more cases than we could have hoped for. Now begins the real heart-pounding phase of work when we scatter back to our desks and analyse the data!

By Ray Bell

Wednesday 30th November saw the end of an exceptionally active Atlantic hurricane season. This season produced 19 tropical storms (wind speeds > 38 mph), from which 7 reached hurricane strength (>73 mph) and 3 major hurricane strength (>110mph) (see figure below – or you can watch this animation of the hurricane season produced by NOAA). Most notable storms of the season were Hurricane Irene, which caused an estimated $4.3bn damage and resulted in 40 deaths (iii, 2011) and post-tropical storm Katia made headlines in the UK (BBC, 9th September 2011).

2011 has been the 3rd busiest Atlantic hurricane season tied with 1887, 1995, and 2010, and behind 1933 (21) and 2005 (28). Although the number of tropical storms was much greater than average (11, 1944-2010) the number and intensity of hurricanes was only slightly above average (6). The accumulated cyclone energy (ACE) index – a measure of the strength and duration of storms throughout the season – was 124, slightly above average (101).

Back in May I wrote an article saying what the Met Office and other institutes were predicting for the forthcoming Atlantic hurricane season. This table gives an ensemble forecast, combining all the forecast centres predictions. The Met Office predicted a most likely number of 13 tropical storms (70% range, 10-17) and an ACE index of 151. Overall, the forecast for ACE index was within their range predicted and the number of tropical storms was outside the predicted range. Although the number of tropical storms was outside their range, this is not unexpected as the observed number of tropical storms in a given year is expected to lie within the prediction interval (70%) approximately 7 out of 10 times (Met Office, 2011). Other forecast centres had a similar validation including NOAA and CSU. They all under-predicted tropical storms and over predicted hurricanes partly due to an unusual season.

Atmospheric/Oceanic Conditions

Phil Klotzbach and Bill Gray have produced a fantastic comprehensive document on the physical conditions that were present this season along with the local conditions for each tropical storm. Their summary for the season is as follows:

“The 2011 Atlantic basin hurricane season had activity at above-average levels, but not to the level that was expected in our pre-season forecasts. While anomalously warm tropical Atlantic SSTs and La Niña conditions likely provided conditions that were reasonably favourable for an active season, anomalously cool SSTs in the northeast subtropical Atlantic, anomalously dry mid levels in the tropical Atlantic and a positive Indian Ocean Dipole (IOD) event conspired to put a damper on this season’s more intense tropical cyclone activity. A weakness in the subtropical ridge located near the East Coast helped induce the recurvature of several systems that might otherwise have threatened the U.S. coastline. We have now gone six years since the last major hurricane made U.S. landfall (Wilma – October 2005).”

Anomalous vertical wind shear as observed across the Atlantic from August 17 – October 15, 2011. Note the anomalously strong vertical shear that dominated throughout the entire subtropical Atlantic (as highlighted by the blue ellipse). Klotzback and Gray, 2011

Dearth of Major hurricane landfalls

Roger Pielke Jr. in his blog was haste to acknowledge the fact that it is now the longest time since records began that the US has gone without a major hurricane landfall – causing headaches for the re-insurance industry. Klotzbach and Gray in their document look into this further and like how the number of tropical storms has huge multi-decadal variability so to does the timing of major hurricanes landfalls.

Whether or not a major hurricane makes landfall is governed by mid-level steering currents – a representation of the large scale atmospheric flow. The figure below shows the 500hPa geopotential height anomaly for 2006-2011 minus 2004-2005 (Klotzbach and Gray, 2011). Note the anomalous trough along the US East coast which has kept major hurricanes at bay for the last 6 years.

Short lived tropical storms

None of the first eight storms of the season developed into hurricanes which now means 2011 holds the record for ‘the longest string of tropical storms to start a season without a hurricane on record’. Landsea et al (2010) note there has been a large upswing  in the number of weak storms from one per year in early records to 5 per year since 2000 due to improved observational capabilities. It is likely many of the short-lived storms in 2011 may have gone undetected without modern satellite technology (NOAA, 2011).

These short lived storms are hard to detect in the Met Office’s forecasting system. Joanne Camp, a long-range hurricane forecaster at the Met Office says “The forecasting system automatically excludes any short-duration storms to help prevent detection of transient disturbances and depressions”. Joanne then goes on to mention “Hopefully a higher horizontal resolution (planned for the Met Office system summer 2012) will help us distinguish between weak tropical storms and tropical depressions so that unusual season’s (like this year) could be better forecast in the future”. Forecasts for the 2012 season will be available in May.

By Jane Strachan

Last Friday saw the release of the International Panel on Climate Change (IPCC) Special Report on Managing the Risk of Extreme Events and Disasters to Advance Climate Change Adaptation (SREX). This report has sparked much interest in the insurance sector, an industry I work with through a Knowledge Transfer Partnership – connecting climate research with insurance industry risk assessment.

The report provides a timely summary of how natural climate variability and human-generated climate change is affecting extreme weather and climate events. This summary isn’t the most reassuring of reads, yet the report then goes on to looks at how integrating expertise in climate science, disaster risk management and adaptation can inform discussions on how to reduce and manage the risks of extreme events and disasters in a changing climate.  It is good to see a report that contains not only a summary of the scientific findings, but also suggestions for action. As Chris Field, the other Co-chair of Working Group II, said: “We hope this report can be a scientific foundation for sound decisions on infrastructure, urban development, public health, and insurance, as well as for planning—from community organizations to international disaster risk management.”

So how can insurance play a role in managing these changes in climate-related hazards? Well, the insurance industry as we know it has been used as a mechanism to help people deal with risk and uncertainty for over 300 years.  The main concept of insurance is to spread risk, and insuring against damage to property can boost resilience in the event of a disaster. Risk sharing does this through providing a way to finance recovery of livelihoods and reconstruction, disaster relief and reducing vulnerability. Additionally, if insurance is priced correctly (i.e. by charging premiums that reflect the true level of risk) then it also can provide knowledge and incentives to encourage society to keep risks manageable, therefore lessening the impact of future extreme events. If premiums become too high, reflecting an unmanageable level of risk, then hopefully this will act to discourage the risk being taken.

However, managing risk in this way relies on an understanding of the underlying hazard. Which is why integrating knowledge from climate science into the industry is so important, particularly when faced with potential shifts in weather patterns and the behaviour of extreme events.

These principles are ambitious, but encouraging in a year that has been a challenging one for the global insurance industry. Economic instability alongside series of natural catastrophes has posed huge operational challenges. Yet the year’s events appear to have increased the urgency for action.

The department is already adding to the integration of climate science into the industry through its involvement in the Willis Research Network – a partnership between scientific institutions undertaking research relevant to catastrophe risk and Willis, a global reinsurance broker. Reinsurance, by the way, is effectively the insurance of insurance, and comes into play when a large disaster triggers multiple insurance claims. The majority of large disasters are natural catastrophes, and a high proportion of natural disasters are weather and climate related. Our work with the industry involves integrating climate hazard understanding into the sector, and exploring ways in which simulation of extreme events by high-resolution global climate models can be integrated into catastrophe assessment tools used by the the industry, to broaden their view of climate-related risk.

Climate scientists from around the world have spent the last few years preparing for, and now running and analysing the results for the fifth global climate model intercomparison project (CMIP5). The aim of CMIP5 is to get many (eventually in excess of twenty) modelling groups to run simulations of past and future climate that conform to a set of prescribed experiments – by comparing the results from this “multi-model” activity, one can both learn about the physics of climate and have more confidence in the simulations themselves. The experiments are broadly targeted at addressing outstanding scientific questions (especially those which arose from the last IPCC assessment report)  and providing estimates of future climate change. The latter include both traditional “hundred-year-plus” experiments like those analysed in previous intercomparison projects, and some new ”out there” experiments aimed at finding out how good (or bad) we are at prediction on decadal timescales (and how we might improve them).

The climate science results are still to come from CMIP5, but there have been a raft of computing problems solved (and yet to be solved) in delivering CMIP5.

Firstly, and most obviously, each modelling group has been improving their models so their simulations can both tell us more about climate mechanisms and be more useful for prediction.  These models have to target a huge range of experiments, the full details can be found here, but a flavour of the diversity can be seen in this figure (from Karl Taylor) showing just the long-term experiments:

Apart being an acronym-fest, the clear take home message from this figure is the diversity of activities being supported within CMIP5 – from what are essentially physics experiments (the lower half) such as the “Cloud Forecasting Model Intercomparison Project” (CFMIP) and the Stratospheric Processes and their Role in Climate (SPARC) activity – to the projections and hindcasts (the top half) which are aimed at projecting future climate (and evaluating the likely fidelity of those projections).

It’s been a challenge for nearly every modelling group: getting the models improved and evaluated, and then used to carry out many experiments on a very aggressive timescale (everyone wanted the main simulations done in time so that the analysis of key results can be done, and published, in time for incorporation in the next IPCC assessment report).   Most models have more complexity and higher resolution, and are being run with bigger ensembles, and while faster computers have helped with that, in practice, most groups have also been solving itty-bitty performance problems “under the hood” – resulting in hard won progress from sophisticated software engineering and computer science. They’ve also been dealing with stupid output data volumes : the Met Office alone is expecting to produce over 50 terabytes of data – roughly half as much data again than all the modelling groups in the world produced for the last global model intercomparison project (CMIP3 in 2006).

The often forgotten “ugly sister” of the climate computing problem is the data handling: in truth I don’t know how many folk are involved in the data pipeline from the Met Office supercomputer, but I’m sure there are at least five folks involved before the data hits a publicly visible website! Similar teams exist in every big modelling group. I sometimes think of climate computers and their running models as data hosepipes. It’s a useful analogy because it reminds us that if you leave the hose on, you need to worry about where all that water is going …

That’s the “modellers-eye-view”, but there’s another perspective, that of the “users” of all this data – some of whom are obviously the same folk who ran the experiments.  The users have to get at all this data, understand which model ran what (and what the model capabilities were), and then, and only then can they do the science. Stretching the hosepipe analogy means they’ll be faced with a swimming pool of data – many petabytes of the stuff! If you think of a dataset as being one variable, produced by one model, for one experiment, then we expect millions of such datasets!

It doesn’t take much thinking about that to realise that the “information management” problem for CMIP5 is just as big as the modelling problem: how to record and display information about the data, alongside the data? How to make the data easily downloadable around the planet? How to avoid those petabytes of data crossing the Atlantic more times than necessary?  How to get systems set up that cross national borders, and that have components built by teams all around the globe? Answering those questions has been a blend of research and both social and software engineering!

The resulting information systems are complex, and right now a wee bit more immature  than any of us would like (i.e. not yet as efficient and reliable as we expect) – but you can see the British gateway to CMIP5 here – just one node in a global federation of data nodes and gateways, the “Earth System Grid Federation”. Those information systems are backed up by teams of people working on the software to sustain it (more pipes) and adding and moving content (to “replicate” core content at key sites). There is even a quality control system in place aiming to eventually provide a formal publication process for the CMIP5 data.

While the eventual size of the CMIP5 data archive is expected to be many many petabytes, as of the time of writing (the current status is here), there are over 500 terabytes of data in over one million files from 17 modelling centres and 30 different models!  I like to think of that archive as a swimming pool of data being filled by many many hosepipes (ok, like all analogies, it’s flawed … maybe multiple swimming pools connected by pipes  … fed by hose pipes … oh stop it Bryan …)

I also like to think that, despite it’s flaws, the information system supporting the CMIP5 archive is an amazing achievement – both socially and technically. Now, like many of my colleagues, I hope to stop thinking about the CMIP5 process, and start thinking about the CMIP5 science (but yes, more data will appear, and the information systems will keep on improving)!