A Random Blog

By Peter Clark

As a young scientist I was introduced to turbulent flow in the traditional way – we consider an ‘infinite ensemble of realisations’ of a random flow, and split each realisation into the average over the ensemble and the ‘random’ fluctuations. I remember being unsatisfied by this approach. Classical physics is not random! What actually is this ‘ensemble’? Why treat the fluctuations as just random noise when any curious eye can see there is a rich structure to the flow?

Many of these questions have (at least partially) been answered by the revolution in mathematics and thinking that is chaos theory (and siblings such as ergodic theory). Perhaps the most remarkable result is that some systems in which the future state is perfectly predictable in terms of the current state (‘deterministic’), evolve to become indistinguishable from a random system. The system ‘forgets’ its initial state, in the sense that to track backwards to find it out requires increasingly accurate knowledge of the current state the further one goes back, to a degree which soon becomes beyond any kind of practicality. This is the converse of the problem of forecasting.

At the same time the computer revolution has enabled us to simulate the evolution of at least a finite sample of an ‘ensemble’ explicitly – a process in weather forecasting sampling the ‘ensemble of initial states’ pioneered with considerable success (and rigour) by ECMWF and now a standard methodology.

Ensemble techniques are now a widespread practice in expressing (often poorly defined) ‘uncertainty’.  This powerful approach has become so universal we often forget to ask the question ‘what ensemble?’ The mere use of an ensemble technique is sometimes taken to give credibility to a piece of work. Too often, arbitrary random perturbations, or worse, an arbitrary mixture of model configurations are used to express ‘uncertainty’, even though it is difficult to know exactly what the results actually mean. While all science is uncertain, perhaps unsurprisingly, some users reject ‘uncertain’ advice with the cry ‘I need to be sure!’

We can, however, return to real physical ensembles arising from the turbulent processes in the atmosphere as an example where uncertainty really matters. When we build weather and climate models, we have to approximate (‘parametrize’) small-scale aspects of the flow (which may be smaller than anything from a few km to several hundred km, depending on the model and application). We simply don’t know how to do this, and there is no reason to suppose it is even possible. However, we do know that, with some restrictions, we can accurately predict an ‘ensemble mean’ behaviour of the small-scale flow. So we use that instead.

The trouble is, we don’t live in an ‘ensemble mean’ world – we live in ‘one realisation’. However, by returning to the quite rigorously defined ensemble, we can also make predictions about the variability of realisations. Figure 1 illustrates this with a very simple model of a real turbulent system. In practical weather forecast models we have shown that using physically realistic random variability can significantly improve the performance of a model (even if the ensemble system we use remains a simplification of the real world) – for example, thunderstorms may form at a more realistic time and evolve more realistically. The downside is that so-called ‘deterministic’ forecasts are an impossibility. Behaving like the real world means behaving, to a certain extent, randomly. Physical realism and not being sure go hand in hand.

Figure 1a

Figure 1a

Figure 1b

Figure 1b

Figure 1c

Figure 1c

 

Figure 1. Results using an ensemble of 10000 realisations of the Lorenz (1963) simple model of Rayleigh-Bénard convection
Top, Figure 1a)     Two realisations of the rate of heating at z=0.75 the height of the system. The ensemble mean must be zero.
1b)     The position of each realization in phase space – the ensemble is randomly distributed over the ‘Lorenz attractor’ – see animation 
1c)      The standard deviation of the time averaged heating rate as a function of averaging time. The red line varies as 1/averaging time.

Reference

Lorenz , E.N., 1963, Deterministic Nonperiodic Flow. Journal of the Atmospheric Sciences 20 (2): 130–141. doi:10.1175/1520-0469(1963)020<0130:DNF>2.0.CO;2.

Posted in Numerical modelling, Weather forecasting | Tagged , | Leave a comment

Characterising extreme event occurrence

By Reinhard Schiemann

When presented with a new data sample, the first thing many of us scientists do is to characterise it in terms of two numbers: the average or mean value of the sample, and the spread or variance of the sample values around the mean. This has become second nature and we rarely stop to think twice about it. Yet it is indeed quite remarkable that data as different as Reading summer temperatures, the chest circumference of Scottish soldiers, or the sum of points obtained by rolling several identical dice can all be characterised by just these two numbers. Essentially, this is a consequence of the Central Limit Theorem in statistics, which states that in the examples above and many other situations, where the data arise as an average of more elementary data (for example tossing individual dice, averaging temperature throughout a season), the samples will tend to follow a Gaussian or normal distribution. The bell-shaped curve of this distribution is ubiquitous in all areas of quantitative science and may be the only mathematical function that has made it onto a bank note (Figure 1). The curve is described by two numbers, the mean determining the location of the bell, and the variance determining the width of the bell.

2016 05 12 Reinhard Schiemann Figure 1

Figure 1. Carl-Friedrich Gauß (1777-1855) and the distribution named after him on the former 10 Deutsche Mark note (source: Wikipedia).

In meteorology we are often interested in extreme events such as strong windstorms, rain and flooding, heatwaves or drought. When we want to describe extreme behaviour, we have to change the way we collect data samples and characterise them. One option is to collect samples that comprise all strongest events in a block of data: the example I am presenting here is maximum daily winter precipitation (rain and snow) that falls over a river basin in each year. Unfortunately, such data samples can no longer be described by the tried and tested Gaussian distribution and its mean and variance. But mathematical statistics comes to the rescue in this situation too: there is an analogue of the Central Limit Theorem, called Extremal Types Theorem, telling us that we can replace the familiar Gaussian bell with a different function called the Generalized Extreme Value (GEV) distribution. We now need three numbers (or parameters) to characterise the GEV. They are called location μ, scale σ, and shape ξ, and their meaning is best illustrated graphically by so-called Gumbel diagrams shown in Figure 2. The vertical axis of these diagrams shows return values indicating the strength of an event (here daily river basin precipitation) and the horizontal axis shows return times, which tell us about the frequency of an event. The bold lines in the diagrams show different GEV distributions and they tell us how to relate a return time to an expected return value. For example, the brown curve in the top panel of Figure 2 shows that the expected return value for a return time of 20 years is 21 mm. We have to wait 20 years on average for a precipitation event of this amount to occur. The location parameter μ determines the vertical position of the GEV curve in the diagram – increasing it to μ=15 mm yields the green curve and the 20-year return value increases to 27 mm. The scale parameter σ determines the slope of the GEV curve in the Gumbel diagram as illustrated in the middle panel of Figure 2. The greater the σ, the more maximum precipitation will vary from year to year, and the more return values will increase with an increase in return time. Finally, the shape parameter ξ describes the curvature of the GEV curve (Figure 2, bottom panel).

2016 05 12 Reinhard Schiemann Figure 2

Figure 2. Illustrative Gumbel diagrams showing GEV distributions with different values for the location parameter (top), for the scale parameter (middle), and for the shape parameter (bottom).

What is all this good for? One application is model evaluation, the process where we assess how realistically numerical models simulate the observed weather and climate. Here, I am interested in how well two versions of a climate model, a low-resolution version (named N96 in Figure 3) and a high-resolution version (N512, also in Figure 3) simulate the extremes of daily winter precipitation over European river basins. To obtain a summary assessment of this performance, I estimate the three GEV parameters for each of the models (N96, N512) and for a reference dataset (E-OBS) based on observed precipitation data from rain gauges. The results are shown in Figure 3. The top row shows the location, scale and shape values for the observations, and the middle and bottom rows show differences between the two models and the observations. We see that both models tend to produce too high precipitation extremes over large parts of Europe, especially over the northern European plains from the Loire river basin in the west to the Vistula basin in the east (greenish colours for the model-observation differences for the location and scale parameters). We also see that this problem is alleviated in the high-resolution (N512) model, where these differences are smaller than in the coarse (N96) model.

The statistical summary assessment shown here is only the first step in model evaluation and many questions remain. How do our two models represent rain-producing Atlantic storms, and how do these storms interact with the European landmass and, in particular, major mountain chains, such as the Alps? Trying to answer such questions is called process-based model evaluation and is an important part of the meteorological research here at Reading. But we will have to leave that for another blog.

2016 05 12 Reinhard Schiemann Figure 3

Figure 3. Estimated GEV parameters for daily winter precipitation over European river basins. Top: precipitation observations (E-OBS), middle: difference between coarse model simulation (N96) and E-OBS, bottom: difference between high-resolution model simulation (N512) and E-OBS. Left: location parameter μ, centre: scale parameter σ, right: shape parameter ξ. Stippling shows statistically significant differences between N96 and E-OBS (middle row) and between N512 and N96 (bottom row).

 

Posted in Climate, Numerical modelling | Tagged , , , | Leave a comment

When Did Fronts First Appear in the Met Office’s Daily Weather Report?

By David Livings

One of the good things that can now be found on the web is a complete series of the Met Office’s Daily Weather Report going back to 1860. An overview of the early history of the report can be found on the Met Office’s web site. The reports themselves are available at the Met Office Digital Library and Archive.

On discovering this resource a few months ago, the first thing that I wanted to know (after finding out the weather on my birthday) was when fronts first appeared on the charts in the reports. Jon Shonk wrote about fronts on this blog in 2014. Although the concept of a weather front dates back almost a century, it still plays an important role in understanding mid-latitude weather systems.

The following images sample the evolution of the charts in the Daily Weather Report from the first charts in 1872 to a time when fronts had become an established part of the Met Office’s operations. Click on an image to see a larger version. The images show how successive generations of meteorologists have tackled the problem of presenting multiple meteorological variables in a compact, easily assimilated form. The reports also include pages of purely tabular or textual information, which are not shown here.

DWR-1872-03-12-p3m-2

Figure 1. These charts are for 12 March 1872. Charts first appeared in the Daily Weather Report the previous day, but high pressure dominated then and it is doubtful that any fronts would have been shown, even if the concept had been available. On the 12th, there was low pressure to the NW of Scotland. A mixture of graphical, numerical, and verbal presentation is used, described in the keys to each of the four charts. Isobars are labelled in inches of mercury (inHg). The contour interval for isobars is not constant, being 0.1 inHg (about 3.4 hPa) in some cases, 0.2 inHg in others.

 

DWR-1897-03-03-p2m-2

Figure 2. Twenty five years later (3 March 1897) and the four charts of 1872 have been merged into two. This has been achieved partly by omitting the attempt to show the general motion of the air and partly by replacing words with shading (for sea disturbance) or with Beaufort letters (for weather). Isobars are now regularly spaced every 0.1 inHg (about 3.4 hPa).

 

DWR-1922-03-01-p2m-2 DWR-1922-03-01-p3m-2

Figure 3. 1 March 1922. There is now one large main chart and three smaller ones. Since 1897, quantitative rainfall has been dropped from the variables plotted, but isallobars (contours of constant pressure change) and low cloud have been added. Isobars are now labelled in millibars and spaced every 2 mbar (1 mbar = 1 hPa). There is redundancy in this way of plotting the data: temperature, weather, and wind speed are all plotted twice. Fronts have not appeared yet, but the concept is still new.

 

DWR-1932-03-06-p2m-2 DWR-1932-03-06-p3m-2

Figure 4. Ten years later (6 March 1932) and the multiple UK charts have been merged into one. Some information has been lost (such as low cloud), but in compensation there is now a full-page chart covering much of the extratropical northern hemisphere. The representation of observations on the UK chart is converging towards the idea of the station model (see Figure 7). There are still no fronts.

 

DWR-1941-04-01-p3m-2 DWR-1941-04-01-p2m-2

Figure 5. 1 April 1941. The style of the charts is similar to nine years ago, although there has been a change in the style of the wind arrows, and weather on the hemispheric chart is now shown using symbols rather than Beaufort letters. Wartime has brought a reduction in the availability of observations. There are none from most of the European continent. Some days there are observations from Russia, America, or the Western Atlantic, but not always (compare Figures 6 and 7). There are still no weather fronts.

 

DWR-1941-05-01-p2m-2

DWR-1941-05-01-p3m-2

Figure 6. One month later – 1 May 1941. Look carefully and you will see the first fronts to appear in this section of the Daily Weather Report. In the SE corner of the UK chart there are two occluded fronts: one heading SE, one heading NE. A key to the fronts has been added by hand under the Further Outlook. Fronts are not yet shown on the hemispheric chart.

 

DWR-1942-03-01-p2m-2 DWR-1942-03-01-p3m-2

Figure 7. Ten months later – 1 March 1942. The UK chart has been expanded to a full page. A full station model has been adopted for plotting the UK observations. This enables the reinstatement of the low cloud that was lost in the early 1930s, but the sea disturbance, which had been plotted since the first charts in 1872, has been dropped. Fronts now appear on the hemispheric chart, and there is a printed key to a rather elaborate system of fronts.

Examination of the copies of the Daily Weather Report available on the web therefore gives the impression that fronts did not appear until 1 May 1941, but that is not the full story. In 1919 the Daily Weather Report was split into three sections: a British Section, an International Section, and an Upper Air Section. We have been looking at the British Section, which is the only section currently available on the web. Fronts first appeared in the International Section on 1 March 1933. Nevertheless, 1 May 1941 (75 years ago this month) is an important date, for it represents the arrival of fronts at the heart of the Met Office’s activities.

 

Posted in Historical climatology, History of Science | Tagged , | Leave a comment

A PhD student’s overview of the European Geosciences Union (EGU) General Assembly 2016

By David Flack

Last week (18 – 22 April) 13,650 scientists from 109 countries descended upon Vienna for the European Geosciences Union (EGU) general assembly. This includes a range of different disciplines, not just those associated with meteorology and hydrology, and amongst these were a large number of scientists from the UK (around 1,300). As a member of the Flooding From Intense Rainfall (FFIR) project I was obviously very interested in a lot of the work associated with precipitation and flash flooding, hence the angle of this blog. In this blog I’m going to try to give a brief overview of what EGU is like (from a PhD student’s point of view) and highlight some of the interesting topics among the hydro-meteorology community at EGU.

Indeed for the FFIR team, EGU started first thing on Monday with Matt Perks (Newcastle University) being schdeuled as one of the first talks of the conference. His talk was a summary of his recent work looking at unmanned aerial vehicles (UAVs) and their use for taking observations whilst floods are occurring, and how this can be used in modelling the water flow in flash flooding situations. Other highlights from the morning session included a talk from the ECMWF (European Centre for Medium Range Weather Forecasting) on a global flash flooding forecast system that they are developing and the links with high-resolution weather forecasts that are able to improve representation of heavy rain

Then after a range of other talks was a poster session in the evening in which Adrian Champion (a member of the Meteorology department here at Reading) was presenting his work on atmospheric precursors to flash flooding amongst various other interesting posters.

One thing that I started to notice, as a PhD student at my first international conference, was the size. There are so many interesting posters and presentations that you can’t get to all the ones or find all the people that you wanted to speak to. However, that size isn’t necessarily a bad thing as it allows you to meet a range of people presenting.

Another good aspect about EGU is the location in Vienna, you are never too far from the city centre via the underground, so you were able to go out in the city centre in the evening and look at the wonderful architecture and experience the Viennese culture (see below).

2016 04 29 David Flack Fig 1 Vienna - IMG_2143 (972 x 648)

Throughout the week there were lots of talks and posters on precipitation including talking about how intense rainfall would vary with climate change in terms of frequency and intensity and hence the impact for flash flooding. Also interesting from my point of view were the various talks on modelling precipitation and the different ways of measuring them and the advances in satellite technology.

I had a couple of posters at EGU presenting my previous work looking at convective regimes in the UK and my current work on uncertainty in models with these regimes. Many people were interested which is always useful when presenting material at a conference. The poster sessions for me were the most useful as you were able to interact with many different people and make useful contacts and collaboration ideas for the future.

For my first time at an international conference I thought the range of disciplines and size of the conference would put me off, but having attended I found the complete reverse happening and would definitely go back again.

Posted in Conferences, Environmental hazards, Hydrology, Students | Tagged | Leave a comment

Polar Prediction School

By Jonny Day

During the last 2 weeks Dr Jonny Day spent two weeks lecturing and coordinating a Polar Prediction School for graduate students and early career researchers. The school is a joint initiative from the World Weather Research Programme (WWRP) – Polar Prediction Project, World Climate Research Program (WCRP) – Polar Climate Predictability Initiative and Bolin Centre for Climate Research.

The school was based at the Abisko Scientific Research Station in northern Sweden – an appropriately Arctic environment (Figure 1). It brought in 28 PhD students and early career researchers from all over the world and a wide range of disciplines for nine days of lectures and practical exercises on the theme of polar prediction. Organised by Jonny Day (Reading University) and Gunilla Svensson (Stockholm University), the invited lecturers included Ian Brooks (University of Leeds), James Screen (Exeter University), Helge Gossling and Thomas Jung (AWI), Cecilia Bitz (University of Washington), Don Perovich (CRREL), Erik Kolstad (University Bergen), Jen Kay (Colorado University), and Matthew Chevallier (Meteo France).

2016 04 21 Jonny Day - Fig 2 - IMG_20160413_141615364

Figure 1. Fieldwork at the Abisko Scientific Research Station in northern Sweden

Polar regions are experiencing rapid changes to their climate; this is opening up new possibilities for businesses such as tourism, shipping, and oil and gas extraction. At the same time it brings new risks to these delicate environments. Effective weather and climate prediction is essential to managing these risks. The complexity of the polar environmental systems, and very limited measurements in these remote regions, make them very challenging environments to provide accurate forecasts for any time scale from days to decades.

 

2016 04 06 1700z - 037

Figure 2. Making measurements of near-surface wind and temperature profiles and the surface energy budget using a micro-meteorology mast erected on the frozen surface of Lake Torneträsk, northern Sweden

As well as an intensive program of lectures and modelling exercises, students conducted practical work based around measurements of near-surface wind and temperature profiles and the surface energy budget made from a micro-meteorology mast erected on the frozen surface of Lake Torneträsk (Figure 2). Radiosondes were released each day, with one day of intensive measurements where radiosondes were released every 3 hours for 24 hours to study the diurnal cycle of boundary layer structure (Figure 3). All the observations were drawn together on the final day to study the full range of processes governing the surface energy balance over the previous week. Other lectures and exercises covered chaotic systems and predictability, operational ocean prediction, modelling polar boundary layer processes, ensemble climate prediction, sea ice processes, and polar lows.

Links: Storify by Denis Sergev (UEA)

2016 04 21 Jonny Day - Fig 1 - IMG_20160412_133320299

Figure 3. Radiosondes were released each day, with one day of intensive measurements where radiosondes were released every 3 hours for 24 hours to study the diurnal cycle of boundary layer structure

 

Posted in Climate, Climate change, Cryosphere, Measurements and instrumentation, Polar, University of Reading | Tagged , | Leave a comment

Energy flows, rainfall patterns and climate

By Richard Allan

The subtle differences in the way heat is distributed between the northern and southern hemispheres, either side of the equator, are important in determining global rainfall patterns and climate. Over many tens of thousands of years, periodic cycles in Earth’s orbit around the sun initiate massive glacial to interglacial swings in climate and the changes in heating between hemispheres cause profound past shifts in the tropical rainy belt with huge impacts on ecosystems. Recent work to observe and simulate the current hemispheric heating imbalance also highlight its great importance for present human societies.

Land generally reflects more sunlight than dark ocean surfaces. Yet although most (around two thirds) of the global land surface area is contained in the northern hemisphere, observations show a remarkable symmetry in hemispheric albedo (both hemispheres reflect just as much sunlight over the course of the year) and this symmetry is additionally explained by the greater cloudiness in the south. However, thermal infra-red energy emitted to space also affects heating differences between each hemisphere and satellite observations show that the southern hemisphere emits less infrared radiation to space and therefore absorbs more energy overall than the northern hemisphere, as depicted in the diagram below.

2016 04 11 Richard Allan Fig 1 DEEPC_Loeb_energyflow_200006-201505

Above: New observations of energy flows in the climate system in petawatts (PW: millions of billions of watts of energy flux) and the location of the Inter Tropical Convergence Zone (ITCZ)

The diagram above (updated from Loeb et al. (2015) using data from Liu et al. (2015) for June 2000 to May 2015 and other sources included in references and notes below) shows that 0.35 petawatts (PW: millions of billions of watts of energy flux) is building up in the southern hemisphere while the northern hemisphere is actually losing more energy to space than it is gaining, despite the fact that the surface is warming up! How can this be?

Here’s how things seem to stack up:

  • (1) Earth is heating up: there is more energy arriving than leaving our planet. Averaging both hemispheres, this amounts to 0.35 – 0.04 = 0.31 petawatts (or 0.6 watts per square metre if you divide by Earth’s surface area of 510 million million square metres). Almost all of this excess energy accumulates in, and warms, the oceans; the rest warms the air (calculated as +0.01 petawatts) and also heats up the solid ground and expends energy melting ice.
  • (2) Much of this ocean heating is in the southern hemisphere (+0.27 petawatts, see notes below)
  • (3) The oceans (mainly the Atlantic) are transporting huge quantities of energy across the equator from the south to the north (0.30 petawatts in the diagram)
  • (4) Much of this transported energy is released into the atmosphere north of the equator (calculated as 0.27 petawatts)
  • (5) This extra heating of the northern hemisphere atmosphere affects the global atmospheric circulation and explains why the tropical rainy belt (or Inter Tropical Convergence Zone, ITCZ) is more in the northern hemisphere than the south
  • (6) The position of the tropical rainy belt helps move a large fraction of the heat transported by the ocean back south of the equator by the atmospheric winds (around 0.23 petawatts). A startling further result from our analysis is that many sophisticated computer simulations, that are used to make projections of future climate change, are unable to simulate the correct energy flow in the atmosphere between hemispheres. Whereas the observations show a southward movement of energy, many simulations produce the reverse.
    2016 04 11 Richard Allan Fig 2 Loeb_rain

    Above: A schematic showing simulations (open symbols) that produce a northward flow of energy across the equator produce less rainfall north of the equator than south of the equator which disagrees with observations (filled red circle) [source: Loeb et al. (2015), Figure 7d]

    We also find that the simulations with the wrong direction of atmospheric energy flow between hemispheres also simulate more rainfall in the southern hemisphere than the northern hemisphere, again the reverse of what occurs in the real world (see schematic above). This is important because it offers a new perspective on how to improve our climate simulations.

    Global rainfall patterns, including monsoon systems that are part of the tropical rainy belt, are of utmost importance to human societies. So this is an important and exciting area of research and we hope to make further progress in observing, simulating and understanding how energy flows and rainfall patterns are changing as the climate continues to be affected by human-caused emission of greenhouse gases and other pollutants.

  • [Research is ongoing under the DEEP-C and SMURPHS projects funded by the Natural Environment Research Council].

    References

Broecker & Putnam (2013) Hydrologic impacts of past shifts of Earth’s thermal equator offer insight into those to be produced by fossil fuel CO2, PNAS, doi: 10.1073/pnas.1301855110

Dong and Sutton (2015) Dominant role of greenhouse-gas forcing in the recovery of Sahel rainfall, Nature Clim. Ch., doi: 10.1038/nclimate2664

Frierson et al. (2013) Contribution of ocean overturning circulation to tropical rainfall peak in the Northern Hemisphere, Nature Geoscience, doi: 10.1038/ngeo1987

Haywood et al. (2016) The impact of equilibrating hemispheric albedos on tropical performance in the HadGEM2-ES coupled climate model, Geophys. Res. Lett., doi:10.1002/2015GL066903

Liu et al. (2015) Combining satellite observations and reanalysis energy transports to estimate global net surface energy fluxes 1985-2012, J. Geophysical Research, doi:10.1002/2015JD023264

Loeb et al. (2015) Observational Constraints on Atmospheric and Oceanic Cross-Equatorial Heat Transports: Revisiting the Precipitation Asymmetry Problem in Climate Models, Climate Dynamics, doi:10.1007/s00382-015-2766-z

Maidment et al. (2015) Recent observed and simulated changes in precipitation over Africa,Geophys. Res. Lett., doi: 10.1002/2015GL065765

Roemmich et al. (2015) Unabated planetary warming and its ocean structure since 2006, Nature Climate Change, doi: 10.1038/nclimate2513

Stephens, G. L., D. O’Brien, P. J. Webster, P. Pilewski, S. Kato, and J. Li (2015), The albedo of Earth. Rev. Geophys., 53, 141-163. doi:10.1002/2014RG000449

Notes

The energy flow diagram is calculated as follows:
(1) Top of atmosphere/surface fluxes are from the Liu et al. (2015) Version 2 geodetic weighting product. The +0.01 petawatt term represents heating of the northern and southern hemisphere atmosphere; the atmospheric heat transport is calculated using the method in Loeb et al. (2015) as 0.5x((0.35 – –0.04) + (0.57 – 0.27)) = 0.23 PW.
(2) Upper ocean (0-2000m depth) heating is 0.228 PW in the south and 0.025 PW in the north based on results for 2006-2013.
(3) Consistent with previous work, I assume an additional 0.035 PW of heating of the deep ocean (primarily in the south) while heating of the ground (including ice) and melting of ice is estimated as another 0.01 PW (70% of which is assumed to affect the northern hemisphere due to the greater land mass)
(4) From this I estimate that 0.27 PW (0.228 + 0.035 + 0.003 = 0.266 PW) is being stored in the southern hemisphere ocean and 0.03 PW (0.025 + 0.007 = 0.032 PW) is being stored in the northern hemisphere ocean but including within both estimates the small heating of land and heating/melting of ice (0.003 + 0.007 PW).
(5) The ocean heat transport is calculated using equation (4) from Loeb et al. (2015): 0.5x((0.57 – –0.27) – (0.27 – 0.03)) = 0.30 PW.

Posted in Climate change, Monsoons, Numerical modelling | Tagged | Leave a comment

Can specific extreme weather or climate events be attributed to climate change?

By Ted Shepherd

Whenever an extreme weather or climate event occurs, scientists are invariably asked whether it can be blamed on anthropogenic climate change. The usual response from climate scientists has been that it is not possible to attribute the cause of specific events, because of the chaotic nature of weather and climate, but that the observed event may be an example of something that is expected to become more (or perhaps less) common because of climate change. In recent years, methods have been developed to try to provide a more concrete answer to this question, a subject known as “extreme event attribution”. However there is controversy about the methodologies, and published studies can draw apparently opposite conclusions about the same event – leading to confusion and the impression we scientists don’t know what we’re doing. It’s fair to say that extreme events have been exploited by both sides of the climate debate.

In response to this situation, the US National Academy of Sciences was asked to undertake a study on the science of extreme event attribution — the first of its kind. The report was developed over the last six months and released on 11 March. (Electronic copies are available for free download.)

(Full disclosure: I was on the author team for the report.)

One important set of conclusions and recommendations concerned the framing of event attribution studies. The question of whether climate change caused an extreme event is generally ill-posed, if cause is understood in the usual deterministic sense, because natural variability almost always plays a role — so there is a random element. An alternative question of whether climate change influenced an extreme event is also ill-posed; given that climate has changed, it is almost tautological that all events have been influenced to some degree by climate change. Yet that is often the question that studies address, with a yes or no answer provided; and it is certainly the question that the media normally ask. Instead, the relevant questions are: what was the influence, and how large was it? Even then, exactly how the question is formulated affects the answer. Examples of such questions are:

  • Are events of this severity becoming more or less likely because of climate change?
  • To what extent was the storm more or less intense because of climate change?

The take-home message from the report is that it is now possible to estimate the influence of climate change on some types of specific extreme events. Confidence is highest for extreme heat and cold events, followed by hydrological drought and heavy precipitation.  There is little or no confidence in the attribution of severe convective storms and extratropical storms.

The overall summary, by extreme event type, is shown in Figure 1 in graphical form, with the basis for the assessment summarized in tabular form in Figure 2.

If you would like further background on extreme event attribution, at a less technical level, I can also recommend my recent review paper on the subject.

2016 03 17 Ted Shepherd Figure 1

Figure 1. The committee’s assessment of overall confidence in capabilities for attribution of specific events, by event type, versus the understanding of the effect of climate change on that event type, in general. As overall understanding improves (horizontal axis), the potential for specific-event attribution increases (vertical axis). A position below the 1:1 line indicates the potential for improvement by technical means alone, but this potential is limited by the level of physical understanding. Source: National Academy of Sciences Attribution of Extreme Weather Events in the Context of Climate Change, March 2016.

2016 03 17 Ted Shepherd Figure 2

Figure 2. The committee’s assessment of confidence in the ‘three legs of the stool’ needed for attribution of specific events: models, observations, and physical understanding. Source: National Academy of Sciences Attribution of Extreme Weather Events in the Context of Climate Change, March 2016.

Posted in Climate, Climate change, Environmental hazards | Tagged | Leave a comment

How do emissions in Abidjan affect the price of chocolate in the UK?

By Peter Hill

The majority of the cocoa beans required to supply the world’s ever increasing demand for chocolate come from southern West Africa. Unfortunately, the volume produced, and consequently the cost of cocoa beans, is heavily dependent on the regional rainfall, which is highly variable.

West African precipitation depends on the West African monsoon (WAM); south-westerly winds during the northern hemisphere summer bring moist air from the Gulf of Guinea, which results in the wet season when most of the annual rainfall occurs (Figure 1). The timing and magnitude of monsoon rainfall is very variable and difficult to predict. Many weather and climate models have problems related to representation of the WAM and climate models show little consensus on the monsoon response to CO2 increases.

2016 03 10 Peter Hill Fig 1

Figure 1: Zonal (10°W – 10°E) mean precipitation for 2000-2014. Note how north of 7°N the precipitation increases from May to August as the monsoon winds  bring increasing moisture further north. Precipitation rates are taken from the Global Precipitation Climatology Centre (GPCC) analysis that is based on rain gauges. The thin light red lines show July mean precipitation from individual years and demonstrate the highly variable nature of the monsoon precipitation.

So how does this relate to emissions in Abidjan? Population growth, economic growth and urbanisation are leading to large increases in anthropogenic emissions in Abidjan and the other cities of southern West Africa. Evidence suggests that anthropogenic aerosol emissions in other monsoon regions affect the strength of the monsoon. For example, reflection of sunlight by sulphate aerosols in India has been linked to a decrease in Indian monsoon rainfall (as explained by Andrew Turner). However, the effect of aerosols on the WAM is unknown, not least because aerosol emissions and concentrations in southern West Africa are not currently well measured or understood.

The EU-funded DACCIWA project (Dynamics-Aerosols-Chemistry-Cloud Interaction in West Africa) brings together scientists from across Europe (including the University of Reading) and southern West Africa to study the effects of anthropogenic emissions in southern West Africa on human health and the regional climate. Measurements of emissions are underway in Abidjan, Ivory Coast and Cotonou, Benin (Figure 2). An extensive field campaign using aircraft and three extensively instrumented ‘supersites’ will be conducted in June-July 2016 in order to improve our understanding of the clouds, precipitation and radiation in this region and how they interact with aerosols. Numerical modelling studies will be used to link these findings to the monsoon dynamics.  Ultimately we aim to understand how emissions affect the West African monsoon (and the production of cocoa!)

2016 03 10 Peter Hill Fig 2

Figure 2: Location of the DACCIWA field sites. The area highlighted by the dotted box shows the broader DACCIWA domain that will be used in modelling studies.

Posted in Aerosols, Climate change, Monsoons | Tagged | Leave a comment

Predicting the airborne spread of hazardous releases in urban areas

By Omduth Coceal

The threat of terrorist attacks, like the risk of accidents, is an unfortunate probability that we need to take seriously and be prepared for. A particularly challenging problem is to be able to predict the spread of potentially toxic material released (accidentally or deliberately) in populated areas. Answers must be provided quickly – ideally in a matter of minutes – in order to guide first responders in making critical evacuation and rescue decisions. At the same time, predictions have to be robust enough while taking into account complex factors such as the local urban environment and incomplete information about the source of the release. These constraints pose a particular set of challenges for formulating and solving the underlying scientific problems and for designing practical models.

2016 03 03 Omduth Coceal Fig 1 - viz

Figure 1. Visualisation of complex turbulent flow over model buildings in a wind-tunnel experiment (Credit: Paul Hayden, EnFlo, Univ. of Surrey)

A whole spectrum of approaches for modelling dispersion exists. At one end of the scale computational fluid dynamics (CFD) models can provide a detailed representation of the urban geometry and reproduce details of the flow field necessary to reproduce the dispersion of pollutants accurately. However, these models are too computationally expensive to run in an emergency response scenario. At the other end of the scale simple Gaussian plume models (GPM) are able to provide very fast predictions but are lacking in accuracy because they fail to take into account the presence of buildings and how these modify the airflow. It is therefore necessary to design models that have some degree of building-awareness but without the need to resolve detailed flow patterns, and that can therefore run fast enough.

DIPLOS (Dispersion of Localised Releases in a Street Network) is a collaborative project involving several institutions in the UK and France and aims to undertake underpinning scientific studies to improve the modelling of dispersion in cities. Recognising that the development and improvement of reliable fast modelling tools require good quality datasets and the development and testing of appropriate models, DIPLOS addresses these needs by focusing on the following core objectives:

(1) To perform detailed wind tunnel experiments and high-resolution numerical simulations of dispersion in arrays of buildings;
(2) To quantify and model the main exchange processes in streets and intersections and represent the effect of different flow processes;
(3) To develop empirical and theoretical methods to estimate concentration levels and fluctuations quickly close to the source, where the danger is greatest; and
(4) To implement and test the resulting methods in an operational fast dispersion model, and to evaluate their performance in simulating realistic case studies in central London.

The project is now well under way and updates will be reported on the project website. See also: Fast network-based modelling of dispersion in city centres (NCAS Atmospheric Physics science highlight).

Posted in Environmental hazards, Numerical modelling, Urban meteorology | Tagged | Leave a comment

Ah, the sweet smell of rain …

By Ellie Highwood

Despite all the rain of the past winter, there is something about rain that I have missed – its perfume. As we head towards spring, with daffodils all around us already, I am looking forward to the first time I get enveloped in the unmistakeable strong sweet fragrance that says ”it’s just rained here”. This smell has a name – “petrichor” from the Greek “petros” meaning  stone  and “ichor” meaning  the fluid that flows in the veins of the gods in Greek mythology. It turns out to be inextricably linked to my research area.

In my research concerning atmospheric particles (aerosols) I care about rainfall primarily because it washes aerosol out of the atmosphere, for example occasionally leaving red Saharan dust films on cars in the UK. This “wet deposition” is a challenge to represent in climate models because you need to get the rainfall pattern right as well as the details of the aerosol since different particles respond differently to passing through rain. When this isn’t right in climate models it can mean that aerosol either stays in the air too long and is transported too far away from source, or doesn’t get transported far enough. In the case of petrichor however, the rainfall is responsible for the emission of aerosol upwards into the atmosphere, allowing it to become airborne and therefore reach our noses – essentially carrying the smell of rain. Petrichor was first described in a Nature paper in 1964 by Bear and Thomas as being airborne molecules from decomposing plant or animal matter that during dry weather become attached to rocks or soil and then are disturbed when the raindrops hit the ground. In fact there are lots of smelly molecules on the Earth’s surface that can be disturbed by raindrops and carried into the air as aerosols. Volatile oils from plants and trees that have collected on rocks can also be vaporised by rainfall – some of these are the inspiration for room fragrances, washing powders and perfumes. The slightly musty, earthy tones of petrichor are provided by soil based bacteria called actinomycetes. These common filamentary bacteria spread their tentacles through damp warm soil pretty much the world over. When the soil and its resident bacteria dry out, spores and a compound called geosmin are produced. The moisture and simple force of rainfall releases spores into the air so that they can drift towards our noses. This is why the “it’s just rained” smell is often strongest after the first rain following a long dry spell, and also more noticeable in spring and summer. Incidentally, the geosmin smells pleasant in the context of earth and gardening, but the same substance when present in drinking water or indeed wine is not so desirable, and is an acquired taste when present in beetroot!

2016 02 25 Ellie Highwood Fig 1 petrichor

Petrichor is the name for the smell after rain has fallen. However, there is also sometimes a different but equally distinctive sharp smell that precedes rainfall. This is not produced by aerosol as such, but by gas phase atmospheric chemistry, specifically ozone production. It’s particularly noticeable ahead of thunderstorms in which electrical charges in the atmosphere split nitrogen and oxygen molecules into individual atoms. Atmospheric chemistry manages to combine some of the oxygen molecules into trios – the ozone molecule. The strong downdrafts associated ahead of thunderstorms can transport this ozone down to nose-level even when it is generated up in the clouds.

2016 02 25 1459z - 033 (912 x 684)

The ‘wilderness path’ between two meteorology buildings on the University of Reading campus – drying nicely in the early spring sunshine, but always ready to resume its normal existence as a sea of sticky mud whenever it rains!

One of the fascinating things about weather and climate is the universality of some of the weather related experiences. The “smell of rain” occurs in poetry and song from across the world, thanks to the global presence of soil bacteria and their role in generating smelly aerosols.

Posted in Aerosols, Atmospheric chemistry, Weather | Tagged , | Leave a comment