Climate change in the Mediterranean Sea

Posted on 14 May, 2018 by Stephen Burt

By Fanny Adloff

The Mediterranean is the largest semi-enclosed sea on our planet. Acting as a miniature ocean, this basin is appropriate to study climate change impact on the ocean. The residence time of the Mediterranean waters – of about a century – is smaller than in the world ocean, so that a quicker response to climate change is expected in this vulnerable basin. Hot-spot of marine biodiversity, the Mediterranean hosts 10,000 to 12,000 marine species, one fourth of them being endemic. The Mediterranean suffers from high anthropic pressure at the shore and at the open sea, which substantially increases its ecosystem vulnerability.

IPCC global climate models do not have enough resolution accurately to represent the complex circulation and the water masses of the Mediterranean basin. High-resolution regional ocean models have been thus developed to study the impact of climate change on this specific region.

A six-member ensemble regional climate simulation covers the period 1961-2099 to study the oceanic response to climate change and its uncertainty (Adloff et al., 2015). The ocean response can vary depending on (i) the socio-economic scenario, (ii) the Atlantic boundary conditions, (iii) the fresh water fluxes from river input and Black Sea inflow, or (iv) the atmospheric surface fluxes. To assess the impact of climate change, the 30 year reference period 1961-1990 is compared to the ‘end of 21st century’ period 2070-2099.

The spatially averaged sea surface temperature increase is predicted to be 1.7°C under low-emission scenario, and up to 3°C under the larger emission scenario (Figure 1). The warming could reach 4°C in some specific regions like the Balearic Islands. The surface warning spreads non-homogenously toward depth, and the deep-water formation dynamics are also affected.

Figure 1: Composite of sea surface temperature anomalies maxima (top) and minima (bottom) for the 2070–2099 period w.r.t. 1961–1990 (°C). The largest (maxima) or smaller (minima) anomaly out of the six scenario simulations is represented at each grid point

Sea surface salinity could increase by 1 g of salt per kilogram of water by the end of the century. These hydrographic changes will lead to substantial modification of the water masses properties of the Mediterranean and of its thermohaline circulation (THC). This three-dimensional circulation is mainly driven by deep water formation processes taking place in winter when surface water becomes denser and sinks towards the bottom of the ocean, bringing oxygen to the deepest layers.

In the present climate, this phenomenon mainly occurs in the western Mediterranean (in the Gulf of Lion and in the Adriatic Sea). In future climates, the simulations show considerable changes of the Mediterranean THC, with a large increase of deep water formation in the Eastern Mediterranean (in the Levantine basin).

Significant modifications of surface currents and of mean sea level are also simulated, the latter being very sensitive to the chosen Atlantic boundary conditions (Figure 2, Slangen et al. 2017).

Figure 2: Cumulative thermosteric sea-level change w.r.t. 1961–1990 (cm), averaged over the Mediterranean Sea from the six-member ensemble scenario simulations from Adloff et al. (2015). In blue, the uncertainties linked to the choice of the prescribed hydrographic conditions of Atlantic waters west of Gibraltar, and in red, the uncertainties linked to the choice of the socio-economic scenario

These changes to the physical characteristics of the Mediterranean Sea could severally affect its marine ecosystems. Because of the warmer waters, many species migrate northward and get trapped by a ‘cul-de-sac’ effect at the north coasts of the Gulf of Lion, the Adriatic Sea and the Aegean Sea. Also, massive extinction of non-migratory species such as Gorgons or Posidonias could take place under climate change.

References

Adloff F. , S. Somot, F. Sevault, G. Jordà, R. Aznar, M. Déqué, M. Herrmann, M. Marcos, C. Dubois, E. Padorno, E. Alvarez-Fanjul, D. Gomis, 2015. Mediterranean Sea response to climate change in an ensemble of 21st century scenarios. Climate Dynamics, doi:10.1007/s00382-015-2507-3.

Benedetti F., F. Guilhaumon, F. Adloff, S.D. Ayata, 2017. Investigating uncertainties in zooplankton composition shifts under climate change scenarios in the Mediterranean Sea. Ecography, doi:10.1111/ecog.02434.

Slangen A.B.A., F. Adloff, S. Jevrejeva, P. W. Leclercq, B. Marzeion, Y. Wada, R. Winkelmann, 2016. A review of recent updates of sea-level projections at global and regional scales. Surveys in Geophysics, doi:10.1007/s10712-016-9374-2

Posted in Climate change, Climate modelling, Environmental hazards, Oceans | Leave a comment

The “size” of the NWP/DA problem

Posted on 8 May, 2018 by Stephen Burt

By Javier Amezcua

There is a professor in the University of Reading that likes to say that the Data Assimilation (DA) problem in Numerical Weather Prediction (NWP) is larger than the size of the universe (estimated to be around 1080 atoms, give or take). I thought it would be interesting to explain what he means by that and how one arrives at such an assertion. While doing this, I will also discuss some basic mathematical concepts that are involved in the combined NWP/DA problem.

First we have to understand NWP. In 1904 Vilhem Bjerknes – one of the fathers of meteorology – stated in a seminal paper that predicting the weather is “simply” a physics problem that can be solved mathematically. After all, the atmosphere is a fluid (made of nitrogen, oxygen, argon, water vapour, etc) that sits on top of a rotating sphere (more or less) in which we happen to live. Motion in this fluid is driven by solar radiation, constrained by the rotation of the Earth (effects such as the Coriolis ‘force’ and Taylor columns come to mind), and subject to boundary conditions (the ocean and land). If we consider the variables: velocity (a 3-dimensional variable), air density, humidity, temperature and pressure, then the evolution of the atmosphere is governed by the equations in Figure 1. Although they seem complicated, these expressions come from simple basic principles: conservation of mass, energy and momentum, as well as an equation of state (perhaps this is the least intuitive one).

Figure 1: Equations governing the evolution of the variables in the atmosphere.

The problem does not seem too big so far, right? Seven equations for seven variables (remember that velocity has three components). These variables, however, depend on one time component and three space components (latitude, longitude and height), i.e. these are spatial fields. Six of them are partial differential equations, i.e. they contain derivatives both in space and time.

“Solving” the equations in Figure 1 means performing the time-integration of the seven spatial fields. This is not an easy task. Simplifications can be done (e.g. by considering limits, equilibrium cases, scale analysis) and one can reach some analytical solutions for the simplified problems. However, in the general case (which is the one used in NWP) the equations have to be solved numerically. This requires both spatial and temporal discretisation. For the spatial part, imagine the atmosphere as a hollow shell, and imagine that we divide this shell in small boxes (recall this is a three-dimensional problem). The smaller the boxes the more precise our solution is.

In a simple latitude-longitude grid, the number of boxes (called gridpoints) is

Nboxes = Nlatitudes x Nlongitudes x Nlevels.

And since in each of the boxes we have seven variables we would have the following number of effective variables:

Nvariables = Nboxes x 7.

Let’s say we discretise every degree in both longitude and latitude, and we assign 20 vertical levels. This already results in Nvariables ~ 107  (in operational centres this number is closer to 107). Suddenly we have gone from 7 fields to 107 grid variables! On top of that, these variables evolve in time.

Now, let us think of the kind of system that we are dealing with. As I have discussed in an older post in this same blog, the atmosphere belongs to a class of dynamical systems known as chaotic. In these systems, a single evolution does not really mean much after a given lead-time, since tiny perturbations in the initial conditions can render completely different trajectories after a time. The dream would be to evolve probability density functions and update them with new observations when these are available. This is the ultimate (yet perhaps unattainable) objective of DA.

As a mental exercise, let us think of constructing the pdf’s empirically (histograms) for the variables of the atmosphere at a given time. For simplicity let’s think we divide the range of each variable in 10 bins (not very high resolution, but this will do). For each bin there is a frequency count. For two variables we would have 102 bins, and it would look something like Figure 2. Can you imagine a histogram in three or more dimensions? (I cannot).

Figure 2: A 2-dimensional (bi-variate) histogram (centre) with 1-dimensional histograms in the marginal axes. (Source: Python documentation)

Now, let us go back to our problem. How many frequency counts do we need to store in order to have an empirical probabilistic view of the atmosphere? Nbins = 10Nvariables. If we go back to our calculations we would get: Nbins = 1010000000 … which is colossal! If we stored a number in each and every atom of the universe, we would still have a huge quantity of numbers without a place to be stored. This is what the original assertion refers to. A more precise way of saying it would be: “an empirical probabilistic representation of the discretised state of the atmosphere is bigger than the size of the universe”.

You may be thinking: can the problem be simplified? Perhaps. For instance, if the distribution of the variables is a joint multi-variate Gaussian (a common assumption), then this distribution is totally determined by a mean vector and a covariance matrix. They are still massive entities: the vector has  109 elements and the covariance matrix has 1018 elements. Imagine storing these elements and doing computations with them … well, this is what we do in supercomputers. But this approximation (Gaussianity) is sometimes out of its depth, so we are always looking for ways to improve our probabilistic representation.

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

Hidden in the clouds

Posted on 30 April, 2018 by Stephen Burt

By Nicolas Bellouin

Our atmosphere contains varying amounts of tiny liquid or solid particles called aerosols. Some aerosols have a natural origin, like the mineral dust particles that form sandstorms, or the sea spray emitted by breaking waves. Other aerosols are due to human activities, from the combustion of fossil fuels to generate electricity or power our transport systems, to agricultural and forest clearing fires.

Aerosols play an important role in the formation of liquid clouds. To form a liquid cloud droplet in normal atmospheric conditions, water vapour needs to condense on to aerosols. Human activities or sporadic natural sources like wild fires and volcanoes increase the number of aerosols in the atmosphere. Cloud water ends up being distributed over more aerosols, leading to the formation of more, smaller liquid cloud droplets.

Distinguishing polluted from unpolluted clouds is generally not possible with the naked eye. But using satellite instruments that measure in the near-infrared part of the electromagnetic spectrum reveals the impact human activities and natural sources have on clouds. In the near-infrared, clouds made up of smaller droplets appear brighter than clouds made up of larger droplets.

The following image is a true colour image taken by a NASA satellite instrument called the Moderate Resolution Imaging Spectroradiometer (MODIS). We are over Vanavara in Central Russia, on 8 October 2016.

The clouds on the western side of the picture are clearly brighter than the clouds on the eastern side, but why? Are they simply thicker? Or have their droplets been made smaller by aerosol emissions? Looking at the same region, but this time using a near-infrared wavelength, brings the answer.

 

Clouds appearing white on the picture are brighter because they are made of smaller droplets than the clouds that appear orange. The linear structure of the brighter clouds betrays discrete sources of aerosols: in this case, wild fires raging in the Siberian forest.

Still over Russia, but on 29 June 2010 this time. The true colour image appears unexciting.

But the same scene observed in the near-infrared reveals a long, narrow strip of clouds that are brighter, hence made of smaller droplets.

What is the aerosol source in this case? A refinery of the Vankor Oil field in Eastern Siberia.

Moving to warmer climes and a different type of clouds. We are now in the Pacific Ocean, on 6 July 2016, looking at the islands forming Vanuatu, and at cumulus clouds. The true colour image appears again rather unremarkable.

But looking in the near-infrared reveals polluted clouds, which are more difficult to identify in such broken cloud conditions. For that reason, they have been marked by the yellow arrows.

Human activities are not the cause of the smaller droplets in this case. The perturbations are caused by two volcanoes, Ambrym and Yasur, degassing sulphur-rich aerosols into the atmosphere.

Closer to home, now. The following image, taken on 1 March 2013, shows cold, pristine air blowing from Scandinavia into northern England.

But look in the near-infrared and suddenly the locations of power stations, which inject aerosols emitted from the burning of oil and coal, appear clearly.

Those images are not just beautiful to look at. Because a brighter cloud reflects more sunlight back to space, man-made aerosols indirectly modify the energy budget of the Earth through their modification of cloud droplet sizes. By identifying and studying hundreds of those “cloud tracks”, my colleague Dr Velle Toll and I have extracted quantitative information about the response of clouds to aerosol emissions by human activities. With our dataset, we can determine if cloud responses simulated by climate models are realistic. Our research will hopefully lead to improved representations of aerosol-cloud interactions in climate models, which are used to predict future climate changes.

Reference

Toll, V., M. Christensen, S. Gassó, and N. Bellouin. Volcano and Ship Tracks Indicate Excessive Aerosol‐Induced Cloud Water Increases in a Climate Model. Geophysical Research Letters, doi:10.1002/2017GL075280, 2018. Available at https://agupubs.onlinelibrary.wiley.com/doi/full/10.1002/2017GL075280

Images from NASA Worldview https://worldview.earthdata.nasa.gov/

Posted in Aerosols, Atmospheric chemistry, earth observation, Remote sensing | Leave a comment

U.K. Spring Weather and the Natural World

Posted on 23 April, 2018 by Stephen Burt

By Pete Inness

We are now just over half way through April, so about half way through meteorological Spring which is defined as March, April and May. Despite the warm weather of the last few days it’s been a fairly cold Spring so far, with the mean temperature across most of the UK being 2 degrees C below normal during March. Easterly winds in late February and March, dubbed the “Beast from the East” in the press, brought cold air across the UK from continental Europe, and there were several spells of snow which caused significant disruption to transport.

To many people the arrival of Spring is not so much set by a date in the calendar but by the occurrence of particular aspects of the natural world. People keep records of events such as the first flowering of particular plants or the emergence of particular insects from hibernation. The Nature’s Calendar website, run by the Woodland Trust, allows people to upload their recordings and produces maps which show how the seasons are progressing in the natural world.

For me personally, the appearance of blackthorn flowers in the hedgerows is a sign that Spring has really started. Blackthorn is a ubiquitous hedging shrub around much of the UK and its white flowers, blossoming before the leaves appear, transform the appearance of the countryside between early March and mid April. In Autumn, blackthorn is also the source of the dark blue sloes needed for the manufacture of sloe gin. With the cold weather in February and March this year it’s not surprising that blackthorn has flowered later than usual. By 1 April last year, with March temperatures a degree or so above normal, there had been many recorded sightings of blackthorn in flower across the whole of the UK, including all of Wales, Northern Ireland and even Northern Scotland (see Figure 1, Top). This year, sightings of blackthorn reported to Nature’s Calendar up to 1 April were restricted to England and South Wales, with no reports at all in Northern Ireland or Scotland and very few reports north of the English Midlands (Fig. 1, Bottom). The very cold Spring of 2013, when March was actually colder than February, led to a very similar pattern of delayed flowering in blackthorn.  In 2016, a year with Spring temperatures very close to the long term average, there were many more reports of blackthorn in flower by 1 April, including in Northern Ireland and the central belt of Scotland.

 

Figure 1. Sightings of blackthorn in flower on or before 1 April, reported to the Nature’s Calendar website. Each dot represents an individual report. (Top) 2017. (Bottom) 2018. Figures downloaded from naturescalendar.woodlandtrust.org.uk.

Year-to-year variations in the dates of particular natural events are not unexpected as the weather has a first order impact on the behaviour or plants, animals, birds and insects. However, climate change has also started to have a discernible impact on long term trends in the natural indicators of the arrival of Spring. A study by Amano et al, published in 2010, used flowering records going back 250 years to create an index of flowering dates for the UK. They found that in the most recent 25 year period in their study, plants flowered between 2.2 and 12.7 days earlier than any other consecutive 25 year period in the study, with flowering occurring, on average, 5 days earlier for each degree of warming.

In some ways this might be good news as a longer growing season for crops in the UK might mean that agricultural productivity will increase. However, there may be downsides in the natural world. Some species may adjust to temperature changes at different rates. To give a simple example, the arrival and nesting of a particular migratory bird species in the UK may be timed to coincide with an abundance of a particular insect larva on which they feed their chicks. However, if that insect starts producing larvae very much earlier than the arrival of the birds then the birds themselves may struggle to feed their young.

In farming too, changes to Spring temperatures may also cause problems. Although Spring is getting warmer on average, there are still examples such as 2013 and 2018 when there can be very cold spells. If blossom on fruit trees has already appeared prior to a cold spell then yields of that particular variety of fruit can be badly hit.

The specific atmospheric conditions that led to the low Spring temperatures this year and in 2013 are now reasonably well understood and there is no indication that these conditions will become less likely to occur in a warmer world. Hence we can continue to expect considerable variations in Spring temperatures from year to year even as the mean climate warms over the next 50 years or so.

Reference
Tatsuya Amano, Richard J. Smithers, Tim H. Sparks, William J. Sutherland., 2010. A 250-year index of first flowering dates and its response to temperature changes. Proceedings of the Royal Society (B).

 

Posted in Climate, Climate change, Phenology | Leave a comment

Skirting the Issue

Posted on 16 April, 2018 by Stephen Burt

By Geoff Wadge

During a major explosive volcanic eruption a set of three main processes transfers mass and heat from the solid earth to the atmosphere. These three processes are: a gas thrust (jet) extending up from the volcanic vent, a middle tower of vigorously convecting gas and ash with air entrainment and topped by a widening region of diffusion and gravity flow near the level of neutral buoyancy. Each eruption tends to produce different plumes, reflecting the variable behaviour of the magma flux at the vent and the ambient atmosphere. Whilst Pliny the Younger provided the original description of a major volcanic explosion plume, the first physical model was produced by Morton et al (1956). Here we focus on a variant of the classic eruption plume – the skirt cloud.

Skirt Clouds
On 17 April 1979 a former colleague of mine took off in a small aircraft to observe up close the eruption of Soufrière volcano, St Vincent. One of the photographs he took (shown below) made the cover of Science. The convection tower, of about 2 km in diameter and seen here from a distance of about 5 km, is partly encased in a “skirt cloud” of unusual complexity. Such clouds had been observed during the era of subaerial nuclear weapons testing, but not from such a close vantage point!

Photograph – K C Rowley

So what is the mechanism leading to the formation of a skirt cloud, in particular this one? There is no definitive analysis that I’m aware of and not all plumes have skirts. Any explanation needs to take account of its location in the convection tower and the high angle layering created. Barr (1982) argued that this example was initiated from thin sub-horizontal layers with high moisture interleaved with drier layers like pileus clouds, the layering being formed by horizontal shearing of cumulus cloud. These sub-horizontal layers are then deformed by the development of the high velocity (> 50 m s-1) eruption column through the layer(s). The heat flux profile at the base of the convective plume should be annular and variations in vent geometry or pressure could help create multiple skirts, as seen here.

The 1979 eruption at Soufrière, St Vincent yielded some excellent science, though my bold photographer colleague was lost to science and he went on to become the current Prime Minister of Trinidad and Tobago.

REFERENCES

Barr, S., 1982. Skirt clouds associated with the Soufrière eruption of 17 April 1979. Science, 216, 1111-1112.

Morton B.R., Taylor, G., Turner, J.S., 1956. Turbulent gravitational convection from maintained and instantaneous sources. Proc. R. Soc. Series A, Mathematical and Physical Sciences, 234(1196), 1-23.

Posted in Aerosols, Environmental hazards, Environmental physics | Tagged , | Leave a comment

Playful floods

Posted on 9 April, 2018 by Stephen Burt

By Sanita Vetra-Carvalho

Flooding is no fun for those who have been affected by it. However, being able to ‘experience’ flooding in some sense is one of the best ways to communicate flood risks as well as the potential solutions to audiences ranging from school children to decision makers within industry and government sectors. Equally important is to be able to build physical models based on simplified reality which can be used in research to learn about less known aspects of flooding such as extreme events with long return periods, study behaviour of the water, or to develop and/or improve numerical computer models used in producing flood forecasts. As such, flood demonstrators have an important role to play in people learning to live in the world where floods are predicted to increase in both frequency and intensity in the near future due to climate change (Sayers et al, 2015).

Below I have summarised portable flood demonstrators known to me developed and available within the UK. These range from completely physical to virtual reality flood demonstrators, and will be presented in this order for no particular reason.

The first one that comes to my mind is Wetropolis developed by Professor Onno Bokhove and his team at University of Leeds. Wetropolis (Figure 1) is an interactive toy model of an extreme rainfall event demonstrating how such an event can lead to flooding in a city. It is aimed at both the general public as well as scientific community effectively and realistically demonstrating city flooding. From the scientific perspective Wetropolis is a very interesting and useful tool because the design of the river channel and rainfall model is based on mathematical techniques and the occurrence of flooding is determined by rainfall of the current and previous days, as in the real world.

Figure 1. The schematic of the Wetropolis flood demonstrator. Image taken from http://www1.maths.leeds.ac.uk/mathsforesees/edin/MFEdinb2016s.pdf

To study and demonstrate the key principles of floods and coastal risk management JBA Trust has created a range of physical models to demonstrate. Their models range from a portable wave tank to a trailer-mounted hydraulic flume demonstrator, and are very effective in visualising and understanding water behaviour in simple river channels and coastal regions.

The JBA Trust wave tank effectively demonstrates the performance of various coastal defences under different wave regimes including beach during a storm surge. It is a very simple, effective and portable wave tank demonstrator.

JBA Trust hydraulic flume demonstrators range from mini to trailer-mounted versions. All of which simulate simple channel flow driven by a system of re-circulating pumps, and feature scale models of typical engineered structures such as bridges, weirs, debris screens etc.

Figure 2. The JBA Trust free standing hydraulic flume. Picture taken from http://www.jbatrust.org/how-we-help/physical-models/hydraulic-flume-free-standing/

Another visually effective way to physically demonstrate flooding, its causes, effects and to test various flood defence mechanisms is using a sandbox and water. While the basic sandbox is cheap to make it can be messy and tedious to work using physical water. This issue is alleviated in the recently developed augmented reality sandbox (AR SandBox). Originally developed by Oliver Kreylos at UC Davis it combines the physical sand in a box with 3D computer visualization tools which simulates the rain and water flow.

Figure 3. An AR SandBox at the GameScienceCenter in Berlin, Germany. Photo by Sjors Houkes.

The AR sandbox is a very interactive way to understand how topography affects water moving through a catchment. Here participants can assemble real or hypothetical topographical environments which are then directly overlaid with an elevation map and contour lines. The rain in the AR sandbox is simulated by holding a hand over the sand and in real time observing how the virtual water moves in the catchment. The effect on water due to changes in the land can be explored by moving the physical sand in the box which will automatically adjust the augmented reality. Currently there are more than 150 AR SandBoxes worldwide adapted for wide range of audiences including JBA Trust, River Way Trust, University of Southampton and many others (explore more here). You can also build your own AR Sandbox following instructions here.

Stepping completely into the virtual reality there is the Flash Flood game which demonstrates the effects of flash flooding due to extreme rainfall events. The game was commissioned by NERC Flooding from Intense Rainfall (FFIR) programme to help share their research and produced by SeriousGeoGames. Flash Flood is based on a real intense rainfall event from 2007 in Northumberland, vividly highlighting the dangers of flash flood, how rapidly it forms, and how damaging it can be. It challenges participants to survive a flash flood. This tool is open access and can be downloaded from the mentioned site for free.

Figure 4. Flash Flood being tried out by Prof Brian Golding at the Flooding From Intense Rainfall annual conference in 2017. Image taken from http://blogs.reading.ac.uk/flooding/2017/12/08/the-ffir-annual-conference/

The same SeriousGeoGames have also developed TideBox (previously known as a Humber in a Box), which merges a research hydraulic model with a gaming engine. Here a hydraulic model is used to simulate long-term development of the Humber estuary and players can watch the tides and flow of the water in 3D. A player can increase or decrease the sea level to see how flood risk might change in the future.

This is by no means an exhaustive list of portable flood demonstrators available in the UK and if you know of or are in a possession of a demonstrator not mentioned here I would love to hear about it, you can email me.

References

Sayers, P.B, Horritt, M, Penning-Rowsell, E, and McKenzie, A., 2015. Climate Change Risk Assessment 2017: Projections of future flood risk in the UK. Research undertaken by Sayers and Partners on behalf of the Committee on Climate Change. Published by Committee on Climate Change, London.

Posted in Environmental hazards, Flooding, Hydrology | Tagged | Leave a comment

Estimating the risks of climate change: what are the effects of climate policy?

Posted on 26 March, 2018 by Stephen Burt

By Nigel Arnell

I am writing this from Beijing, where the 13th National People’s Congress has just reaffirmed the Chinese commitment to control future emissions of greenhouse gases and meet the aspirations of the Paris Agreement on Climate Change. This agreement, struck in 2015, commits countries to reduce emissions so that the increase in global mean temperature is limited to ‘well below’ 2 oC above pre-industrial levels, and to aim to limit the increase in temperature to 1.5 oC. I am here to attend the final workshop of a joint UK-China project which is providing policy makers in China and elsewhere with information on climate risks at the global and Chinese scales. We are looking at risks under ‘high’ emissions pathways, and are comparing these with the risks that arise under pathways that are consistent with the Paris Agreement. Policymakers want to know this partly so they can understand and communicate the risks of not reducing emissions, and partly in order to prepare for risks which cannot be avoided.

The analysis builds on several years of research funded by NERC and the Department for Business, Energy and Industrial Strategy (and its predecessors), most recently through the AVOID2 programme. The headline results from the AVOID2 programme were recently published by Lowe et al. (2017), showing that the impacts under high emissions were much larger than under lower emissions – particularly for impacts related to heat extremes – and also that in order to achieve the Paris targets it is necessary to reduce emissions rather sooner than implied in the commitments made by countries in Paris.

Figure 1 below (simplified from Arnell et al., 2018) shows the proportion of impacts that are avoided with a specific policy target (along the x axis), relative to the impacts that would occur at higher increases in temperature, for four different indicators of potential impact: the population exposed to drought, the area of cropland exposed to drought, the population exposed to flooding and the population exposed to heatwaves. The orange line shows the proportion of impacts that would be avoided with a particular policy target compared to impacts at an increase of 4 oC in 2100. Meeting the 1.5 oC target would reduce the population exposed to drought by around 80%, compared with a 4 oC world, and meeting a 2 oC target would reduce impacts by around 65%. The proportional effects are greater for the heatwave indicator and lower for the cropland drought indicator – suggesting that achieving climate policy targets would have the greatest effect on impacts generated by high temperatures.

Figure 1. The impacts avoided with different climate policy targets, relative to impacts at 2, 3 and 4 oC above pre-industrial levels. The plots show four impact indicators. The solid lines show the median estimate of the proportion of impact that is avoided, and the shaded area shows the uncertainty range.

The green line shows the proportion of impacts avoided relative to a temperature increase of 3 oC, which is close to the estimated increase that would occur if countries met the commitments they have already made to reduce emissions. Meeting the 1.5 oC target would reduce the population exposed to drought by approximately 65% and again the proportions vary between the indicators: heatwave impacts are reduced by around 90%. This suggests that additional efforts to reduce emissions beyond those already committed by countries would reduce future impacts substantially.

The blue line compares impacts at 1.5 oC and 2 oC, illustrating the benefits of moving to the more stringent Paris Agreement target. Drought impacts would be approximately 40% lower at 1.5 oC than at 2 oC, and heatwave impacts would be around 60% lower.

There are, of course, many caveats with this assessment. There is uncertainty in the projected benefits of climate policy, due to uncertainty in future regional changes in temperature and, particularly precipitation. This is shown by the shaded areas in the figure. There are also other potential indicators for the four impact areas shown here, and these may result in different apparent benefits of reducing climate change. The proportion of impacts avoided at lower temperatures varies between regions. However, together the results show that reducing emissions can reduce impacts substantially, and that impacts at 1.5 oC can be a lot lower than impacts at increases of 4, 3 or even 2 oC.

REFERENCES

Arnell, N.W., Lowe, J.A., Lloyd-Hughes, B. & Osborn, T.J., 2018. The impacts avoided with a 1.5 oC climate target: a global and regional assessment. Climatic Change 147: 61-76. Doi. 10.1007/s10584-017-2115-9. https://link.springer.com/article/10.1007%2Fs10584-017-2115-9

Lowe, J.A., Arnell, N.W., Warren, R., Gambhir, A., Bernie, D. & Thompson, E., 2017. Avoiding dangerous climate: results from the AVOID2 programme. Weather 72: 340-345. Doi 10.1002/wea.3176  https://rmets.onlinelibrary.wiley.com/doi/abs/10.1002/wea.3176

Posted in Climate, Climate change, Climate modelling, Conferences, Environmental hazards, Greenhouse gases | Tagged | Leave a comment

Is the Montreal Protocol really working?

Posted on 19 March, 2018 by Met Department

By Michaela Hegglin

The Montreal Protocol, which celebrated its 30th birthday last year, is an international treaty established in 1987 to protect the ozone layer from human-made ozone depleting substances. The Montreal Protocol has been hailed as the most effective international environmental agreement to date, and addressed one of the most pressing environmental issues of the 20th century. But … is the Montreal Protocol really working?

Montreal Protocol history in a nutshell
It was the English scientist James Lovelock who was the first to measure the abundance of chlorofluorocarbons (CFCs) with a homemade gas-chromatograph and to realise that these human-made substances were found ubiquitously in both the northern and southern hemispheres (Lovelock et al., 1973). The finding triggered Mario Molina and Sherwood Rowland’s hypothesis in the early 1970s (Molina and Rowland, 1974) that CFCs could only be destroyed in the stratosphere where they release chlorine atoms, which then would be able to destroy ozone catalytically and pose a threat to the ozone layer. The discovery of the Antarctic ozone hole in 1985 by Joe Farman and colleagues at the British Antarctic Survey (Farman et al., 1985) proved their hypothesis to be not only correct, but also far more threatening than had been imagined even by them. It spurred research activities to understand why such severe ozone depletion was found over Antarctica alone, and led to political action under the Montreal Protocol in 1987. The realization that more severe ozone depletion would spread further across the globe if we were to continue releasing CFCs into the atmosphere, along with technological advancements that made replacement of CFCs possible, helped governments to tighten the regulations on CFCs through several Amendments to the Montreal Protocol. 

Ozone layer research today
Almost 50 years after Molina and Rowland’s hypothesis, research on the stratospheric ozone layer is still ongoing, but now focuses on the question of whether the Montreal Protocol and associated Amendments is working and whether the ozone layer is beginning to recover. In particular, researchers now know about the confounding effects that climate change and tropospheric pollution can have on attempts to detect ozone recovery. In the WMO Scientific Ozone Assessment Report 2014, the key statements in the Summary for Policy Makers on this topic point out that indications of ozone recovery since 2000 are found in global total column observations (although not yet attributable to the decline in ozone depleting substances, ODS), and that ozone increases have been found in the upper stratosphere, half of which were attributable to ODS decline (with the other half attributed to climate-change and its effects on stratospheric temperatures) (WMO, 2014). Since the last assessment, studies by Shepherd et al. (2014) on the total column ozone evolution at mid-latitudes and Solomon et al. (2016) over Antarctica attributed ozone recovery to declining ODS concentrations with the help of complex model simulations that help distinguish ODS-related changes from those induced by climate parameters and other natural factors such as volcanic aerosol.   

A disconcerting finding …
More recently, however, a study published by Ball et al. (2018) found on the basis of observations alone that ozone in the lower stratosphere is in fact not recovering but in continuous decline (see Figure 1).  The study applied a more refined statistical method than usually used in the research field of stratospheric ozone, with which the authors were better able to take into account natural variations in ozone. The paper received much publicity, since its findings imply that the Montreal Protocol is not working as expected. On the other hand, some colleagues in the field were quick to denounce the paper and its conclusions. Is it time to worry?

Figure 1: Ozone changes as derived from different stratospheric ozone data records (taken from Ball et al., 2018).

While these findings are indeed disconcerting, the changes in the lower stratosphere do not seem to have had a discernible effect on total column ozone (at least not yet) (Weber et al., 2018). The changes are also (at least partially) compensated by increases in tropospheric ozone (Ball et al., 2018; Shepherd et al., 2014). Were the decline in lower stratospheric ozone to continue, however, the consequences could become more serious. In fact, scientists were expecting ozone decline in the lower stratosphere as a consequence of climate change due to a strengthening of the stratospheric circulation (Hegglin and Shepherd, 2009). These changes would lead to a substantial increase in harmful radiation reaching Earth’s surface in the tropics, where UV levels are low to begin with and where most people live.

The way ahead
What is ultimately needed to answer the question of whether the Montreal Protocol is working is to attribute the causes of the observed changes. Are they indeed the result of non-compliance with the Montreal Protocol’s regulations, or just the result of natural variability? Or are they instead due to climate change? If the latter were the case, the Montreal Protocol would be working but concern for the protection of the ozone layer would have to shift towards regulating climate change. More evaluations of the currently available ozone data record are needed, to confirm (or refute) the results of Ball et al. These evaluations should in particular take into account an aging fleet of ozone instruments flying in space, since they may well show signs of degradation potentially affecting the measurements. The finding also highlights that a renewed commitment to measure vertically resolved ozone in the stratosphere and ozone depleting substances in the troposphere is required to be able to satisfy future needs of monitoring the ozone layer. Both these are essential to see whether trends are to continue and to help attribute the changes to either increasing greenhouse gases or ozone depleting substances.

References

Ball, W. T., Alsing, J., Mortlock, D. J., Staehelin, J., Haigh, J. D., Peter, T., Tummon, F., Stübi, R., Stenke, A., Anderson, J., Bourassa, A., Davis, S. M., Degenstein, D., Frith, S., Froidevaux, L., Roth, C., Sofieva, V., Wang, R., Wild, J., Yu, P., Ziemke, J. R., and Rozanov, E. V., 2018. Evidence for a continuous decline in lower stratospheric ozone offsetting ozone layer recovery. Atmos. Chem. Phys., 18, 1379-1394, https://doi.org/10.5194/acp-18-1379-2018

Farman, Joseph C., Brian G. Gardiner, and Jonathan D. Shanklin., 1985. Large losses of total ozone in Antarctica reveal seasonal ClOx/NOx interaction. Nature 315, no. 6016: 207.

Hegglin, M. I., and T. G. Shepherd, 2009. Large climate-induced changes in UV index and stratosphere-to-troposphere ozone flux. Nature Geoscience 2, 687-691.

Lovelock, J. E.; Maggs, R. J.; Wade, R. J., 1973. Halogenated Hydrocarbons in and over the Atlantic. Nature 241, no. 5386: 194. doi:10.1038/241194a0.

Molina, Mario J., and F. Sherwood Rowland, 1974. Stratospheric sink for chlorofluoromethanes: chlorine atom-catalysed destruction of ozone. Nature 249.5460: 810.

Shepherd, T. G., D. Plummer, J. Scinocca, M. I. Hegglin, C. Reader, V. Fioletov, E. Remsberg, T. von Clarmann, H. J. Wang, 2014. Reconciliation of halogen-induced ozone loss with the total-column ozone record. Nature Geoscience, 7 (6), 443–449, doi:10.1038/NGEO2155

Solomon, Susan, Diane J. Ivy, Doug Kinnison, Michael J. Mills, Ryan R. Neely, and Anja Schmidt, 2016. Emergence of healing in the Antarctic ozone layer. Science: aae0061.

Weber, M., Coldewey-Egbers, M., Fioletov, V. E., Frith, S. M., Wild, J. D., Burrows, J. P., Long, C. S., and Loyola, D., 2018. Total ozone trends from 1979 to 2016 derived from five merged observational datasets – the emergence into ozone recovery. Atmos. Chem. Phys., 18, 2097-2117, https://doi.org/10.5194/acp-18-2097-2018

Posted in Atmospheric chemistry, Climate, Environmental hazards, Stratosphere | Tagged | Leave a comment

Improving estimates of soil moisture over Ghana

Posted on 12 March, 2018 by Stephen Burt

By Ewan Pinnington

This work aims to improve estimates of soil moisture over Ghana as part of the ERADACS project. In regions where the population relies on subsistence farming it is soil moisture, rather than precipitation per se, that is the critical factor in growing crops. The production of improved soil moisture forecasts should therefore enhance the drought resilience of these regions through improved capacity for early warning of agricultural drought. The seasonal cycles of precipitation and soil moisture over Ghana are shown in a video here, in this video we can see how the response of soil moisture to increased rainfall is lagged as it takes time for the rain to infiltrate into the soil.

Mathematical models of the land-surface are useful tools to inform soil moisture forecasts, but model errors are problematic. In order to improve soil moisture estimates from the Joint UK Land Environment Simulator (JULES) land surface model over Ghana we have combined satellite observations of precipitation and soil moisture with model predictions using the technique of data assimilation.

We have built a four-dimensional variational data assimilation system that ingests soil moisture observations from the European Space Agency (ESA) Climate Change Initiative to update the soil model parameters of JULES. In our experiments we drive the JULES model with precipitation observations from TAMSAT. Figure 1 shows a data assimilation experiment for one grid box. In this experiment we assimilated one year of soil moisture observations (2009) and then ran a five year hindcast (2010-2014) to judge the model performance against independent data. We can see that the JULES model being run with updated parameters after data assimilation (dark grey line) fits the ESA observations of soil moisture better than our prior model run. We can also see the reduction in bias for the hindcast period over the whole of Ghana in Figure 2, we see that before data assimilation the model is too wet over much of the country and that this bias is reduced after data assimilation.

Figure 1. Soil moisture data assimilation results for a north Ghana grid. Light grey line: prior JULES trajectory. Dark grey line: posterior JULES trajectory. Black dots: ESA CCI soil moisture observations. Faint grey vertical lines: error bars for observations. Vertical dashed line represents the end of the assimilation window.

Figure 2. JULES modelled soil moisture bias over Ghana for period 2010-2014. Left: before data assimilation. Right: after data assimilation.

Overall we find a 44% reduction in root-mean-squared error for our 5-year hindcast after assimilating a single year of soil moisture observations to update model parameters. The initial results of using this system are encouraging, but more work is needed to judge our results against “ground-truth” observations of soil moisture. From this work we also conclude that rainfall data has the greatest impact on model estimates during the seasonal wetting-up of soil, with the assimilation of remotely sensed soil moisture having greatest impact during drying down. For more information on this work please see our Hydrological and Earth System Sciences Discussions paper (Pinnington et al., 2018).

Reference
Pinnington, E., Quaife, T., and Black, E., 2017 (in review). Using satellite observations of precipitation and soil moisture to constrain the water budget of a land surface model. Hydrol. Earth Syst. Sci. Discuss., https://doi.org/10.5194/hess-2017-705

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

A simple way to find out where the moisture for regional rainfall comes from

Posted on 5 March, 2018 by Stephen Burt

by Liang Guo

Moisture tracing is an interesting scientific topic that has fascinated meteorologists and hydrologists for decades. Methods for tracing moisture are numerous, from observations to numerical modelling, from water isotopes to remote sensing, from online tracking to off-line tracking, and both Eulerian and Lagrangian methods are used.

A simple method involves a two-dimensional box model. To build a simple model, assumptions are needed. There are three assumptions:

  1. Vapour in the box remains constant at monthly time scales or longer;
  2. No matter from where the moisture comes, it is well mixed within the box;
  3. Evaporation and precipitation are constant within the box. Then, you can derive a simple relationship from the atmospheric water vapour conservation equation:

ρ = E / E+2Fin

This relationship is developed by Brubaker et al. (1993); ρ is the precipitation recycling ratio, which is the fraction of precipitation within the box that originates as the evaporation from the same box: E is the evaporation with the box and Fin is the horizontal moisture flux into the region, which is vertically integrated through the height of the box.

If the moisture does not come from the evaporation with the box, then it must come from outside in form of the moisture advection. Therefore,

α = 1-ρ = (2Fin)/(E+2Fin)

Where α is the ratio of precipitation arising from advected moisture to the total precipitation within the box.

If we further divide the Fin according to the directions, then we can calculate the contribution of advected moisture from different directions. Together with the contribution from the local evaporation, we can figure out from where the moisture to the precipitation within the box comes.

ρ + αW + αE + αN + αS =1

Where, W, E, N and S represent directions.

Take the central-eastern China for example (Figure 1, left). Applying the aforementioned equations to this region shows the seasonal cycle of the moisture contributions from all directions in Figure 1 (middle). It is clear that the summer monsoon (via the southern boundary) makes a significant contribution during the June-July-August, especially in July (40%). However, the contribution via the western boundary is equivalent or larger. In the winter, the moisture predominantly comes via the western boundary, although the mean precipitation is small (Figure 1, right).

Applying the simple model to a realistic case requires caution. However, similar results have been found from other studies using more sophisticated methods. Besides, a statistical test done by Guo et al. (2018) shows that about 70% of the precipitation interannual variation can be explained by the moisture flux via all these boundaries.

Figure 1 (Left) The study region. The boundary is divided into west (green), east (black), north (red) and south (blue). (Middle) Percentage contributions to precipitation from the moisture influxes from different directions, as well as from the local evaporation in grey. (Right) The mean seasonal cycles of precipitation calculated from the ERA-Interim re-analysis during 1979-2012, units mm/month. The precipitation is separated into colours according to the moisture contributions from each section of the boundary.

Reference

Brubaker, Kaye L., Dara Entekhabi, and P. S. Eagleson, 1993. Estimation of Continental Precipitation Recycling. Journal of Climate.

Guo, Liang, Nicholas P. Klingaman, Marie-Estelle Demory, Pier Luigi Vidale, Andrew G. Tuner, and Claudia C. Stephan, 2018. The contributions of local and remote atmospheric moisture fluxes to East Asian precipitation and its variability. Climate Dynamics, on-line DOI: 10.1007/s00382-017-4064-4.

 

Posted in China, Climate, Climate modelling, Hydrology | Leave a comment