Why has there been a rapid increase in heat-related extremes in Western Europe since the mid-1990s?

By Buwen Dong

In the last few decades, Europe has warmed not only faster than the global average, but also faster than expected from anthropogenic greenhouse gas increases (van Oldenborgh et al., 2009). With the warming, Europe experienced record-breaking heat waves and extreme temperatures, such as the 2003 European heatwave, 2010 Russian heatwave, and 2015 European heatwave, which imposed disastrous impacts on individuals and society.

Illustrated in Figure 1 are time series of the area averaged summer (June to August, JJA) surface air temperature (SAT), and summer or annual temperature extreme anomalies over Western Europe (35oN-70oN, 10oW-40oE, land only) relative to the climatology over the time series expanding period. One of most important features is the abrupt surface warming since the mid-1990s and rapid increases in temperature extremes. The changes in SAT and temperature extremes during the recent 16 years (1996-2011) relative to the early period 1964-1993 are more than 1.0 degrees Celsius (degC).

Figure 1. Time series of summer (JJA) or annual mean anomalies relative to the climatology (mean of the whole period) averaged over Western Europe (35oN-70oN, 10oW-40oE). (a) Surface Air Temperature (SAT, degrees Celsius),Tmax, Tmin, and Daily Temperature Range (DTR, degC), (b) annual hottest day temperature (TXx) and warmest night temperature (TNx) (degC), TXx and TNx based on two data sets of HadEX2 and E-OBS are shown. Black and red range bars indicate the earlier period of 1964–1993 and the recent period of 1996–2011.

What has caused the rapid summer warming and increases in high temperature extremes over Western Europe? Relative to an early period of 1964-93, sea surface temperatures (SSTs) have warmed, particularly in the North Atlantic and Indian Oceans, and sea ice extent (SIE) has decreased. Due to air quality legislation, anthropogenic aerosol (AAer) precursor emissions in Europe and North America have decreased since the 1980s and greenhouse gas (GHG) concentrations have increased. In order to understand the relative importance of these forcing factors on the rapid Western European summer warming and increases in hot temperature extremes, numerical experiments with the atmospheric component of a state of the art global climate model have been performed in a study by Dong et al (2016).

Some area averaged summer or annual changes in temperature extreme indices over Western Europe between two periods for both observations and model simulations are illustrated in Figure 2. There is good agreement between the model forced by changes in all forcings and observed changes in summer seasonal mean SAT, Tmax, and Tmin. In response to changes in all forcings, the model simulates an area-averaged summer mean SAT change of 1.16 ± 0.21 degC over Western Europe, which is very close to observed change of 0.93 degC. The changes in SST/SIE explain 62.2 ± 13.0% of the area-averaged SAT signal, with the 37.8 ± 13.6% explained by the direct impact of changes in GHGs and AAer. Both changes in SST/SIE and AAer lead to an increase in Tmax, while the increase in Tmin is predominantly due to the change in SST/SIE. The direct impact of AAer changes act to increase Daily Temperature Range (DTR), but change in DTR is countered by direct impact of GHG forcing. However, DTR change in response to all forcings is overestimated by the model. Results also suggest that the direct impact of AAer changes plays an important role in the increase in the annual hottest day temperature (TXx) (explaining 45.5 ± 17.6% of the signal in the response to changes in all forcings) while the increase in the annual warmest night temperature (TNx) is mainly mediated through the warming of the ocean.

Figure 2: Observed and model simulated summer seasonal mean (JJA) changes between two periods for SAT, Tmax, Tmin, and DTR, the annual changes in annual hottest day (TXx) and warmest night (TNx) temperature, averaged over Western Europe. SAT, Tmax, Tmin, DTR, TXx, and TNx are in degC. (a) Observed changes (based on CRUTS3.2 and HadEX2) data sets, and simulated responses to changes in SST/SIE, GHG concentrations, and AAer precursor emissions. The coloured bars indicated the central estimates and the whiskers show the 90% confidence intervals based on a two tailed Student t-test. (b) Model simulated changes in response to different forcings. SST & SIE is the response to changes in SST/SIE. GHG is the response to GHG concentrations, and Aerosols is the response to changes in AAer precursor emissions.

Whilst each forcing factor causes summer mean surface warming and associated temperature extreme changes over Western Europe, the physical processes are distinct in each case. For example, SST/SIE changes lead to more or less uniform summer mean warming at the surface. In contrast, changes in AAer lead to a band of surface warming and temperature extreme changes in latitude of 40oN-55oN. The results in this study illustrate the important role of the direct impact of changes in AAer not only on summer mean temperature but also on temperature extremes. Reduction of AAer precursor emissions not only induces increased downward solar radiation through aerosol-radiation and aerosol-cloud interactions, but also induces local positive feedbacks between surface warming and reduced cloud cover, reduced precipitation, soil moisture, and evaporation.

Looking forward in the next few decades, greenhouse gas concentrations will continue to rise and anthropogenic aerosol precursor emissions over Europe and North America will continue to decline. Our results suggest that the changes in seasonal mean SAT and temperature extremes over Western Europe since the mid-1990s are most likely to be sustained or amplified in the near term, unless other factors intervene.

References

Dong, B.-W., R. T. Sutton, and L. Shaffrey, 2016: Understanding the rapid summer warming and changes in temperature extremes since the mid-1990s over Western Europe. Clim. Dyn. doi:10.1007/s00382-016-3158-8

van Oldenborgh GJ, Drijfhout S, van Ulden A, Haarsma R, Sterl A, Severijns C, Hazeleger W, Dijkstra H, 2009: Western Europe is warming much faster than expected. Clim Past 5(1):1–12

Posted in Aerosols, Atmospheric chemistry, Climate, Climate change, Climate modelling, Environmental hazards, Numerical modelling | Leave a comment

The physics behind a physics scheme

By Alan Grant

When I joined the Met Office (or, as it was then, The Meteorological Office), I was posted to the boundary layer group. I spent a number of years investigating the atmospheric boundary layer, using data from aircraft and tethered balloons. The justification for the work was to increase our understanding of the boundary layer, which would hopefully lead to improvements in the parametrization of the boundary layer in forecast models. Fast forwarding to the present, I now work on the boundary layer that forms below the surface of the ocean, using high resolution large eddy models, instead of autonomous underwater vehicles (AUVs) and gliders.  The aim of the work remains the same, to develop better parametrizations.

Figure 1. The sea surface in a North Pacific Storm. Photo credit – NOAA

A simple approach to parametrizing the ocean boundary layer is to use parametrizations developed for the cloud-free, atmosphere boundary layer, but upside down (making appropriate changes to account for different densities and heat capacities of air and water). This is a reasonable strategy, but it turns out that there is more to the ocean boundary layer than this, and unsurprisingly the source of the difference between the oceanic and atmospheric boundary layers is the boundary condition.

The possible effects of the surface waves, one of the more striking features of the ocean surface, is an obvious difference between the oceanic and atmospheric boundary layers. Breaking waves, and the interaction between turbulent currents and the Stokes drift of the surface waves (a Lagrangian drift which arises from the non-linearity of the Navier-Stokes equations) has dramatic effect on the properties of the turbulence in the boundary layer.

A more fundamental difference between the oceanic and atmospheric boundary layers is the effect that the surface stress has on the boundary layer flow. In the atmospheric boundary layer momentum is transferred to the surface, so that the surface exerts a drag on the atmosphere. Along with the transport of momentum to the surface, there is also a transport of the mean kinetic energy of the flow (not to be confused with turbulent kinetic energy) from the outer region of the boundary layer towards the surface. This flux of mean kinetic energy maintains the flow near the surface, and supplies the energy needed for the large dissipation rates that occur in the surface layer. The ultimate source of this kinetic energy flux is the work done by the pressure gradient.

In the oceanic boundary layer, the momentum transferred from the atmosphere to the surface acts to generate the mean current. Along with the transfer of momentum into the ocean there is, again, a transfer of mean kinetic energy, but now it is directed away from the surface into the ocean.  This energy flux supports the generation of turbulence at the base of the well-mixed portion of the boundary layer, and turbulent mixing in the stratified layer below. The turbulent mixing in the stratified layer is an important feature of the upper ocean, but is poorly represented in current parametrizations.

To improve parametrizations of the mixing in the stratified layer we need to understand in more detail the process outlined above, and large-eddy simulation can be used to make detailed studies of this and other processes. By understanding the fundamental physics that lies behind the physics scheme we can hopefully improve the parametrization of the surface boundary layer in ocean models.

Posted in Boundary layer, Environmental physics, Numerical modelling, Oceans, Waves | Leave a comment

Changing wet and dry seasons

By Richard Allan

The fickle nature of weather patterns is ultimately responsible for the where and when of tropical rainfall extremes which wreak damage on agriculture, infrastructure and people. Tropical cyclones, such as Enawo which battered Madagascar in March, can severely impact low-lying, highly populated regions through intense rainfall combined with strong winds and storm surges. Explosive thunderstorms operating at smaller spatial scales can generate flash flooding and may lead to devastating landslides in mountainous terrain. A sustained dearth of rainfall or multiple failed seasonal rains, as implicated in drought currently impacting Somalia, Kenya and Ethiopia, are also inextricably linked with evolving weather patterns, often driven by the slower heart-beat of the oceans as they pace out the internal rhythm of El Niño Southern Oscillation and its decadal physiognomies.

Despite the dominant role of chaotic atmospheric and oceanic fluctuations in meting out extremes of weather, there are several controlling factors at the largest terrestrial scales. Barely perceptibly, they nudge the distribution of extremes away from present day patterns through global climatic changes caused by the influence of human activities on the planet’s energy budget. Rising concentrations of greenhouse gases in the atmosphere are slowly but inexorably perturbing the climate system, modifying the ultimate fuel for rainfall extremes: water and energy. While detailed computer simulations capture the fluid, thermodynamic essence of the atmosphere, the processes building up to rainfall extremes, the fine-scale detail of convective storms, must be approximated using a simplified approach but very much rooted in physics. The reliability of future projections out across progressively more distant decades relies on fundamental linkages between the skill in depicting essential physics at the tens to hundreds of kilometre scales and the unrepresented detail which determines impacts. Nevertheless, there is a physical basis for anticipating substantial yet contrasting changes across wet and dry meteorological regimes as the climate warms.

Figure 1: Monthly anomalies in observed tropical total column integrated moisture and surface temperature. Using updated data from Allan et al. (2014) Surv. Geophys.

Figure 1: Seasonal anomalies in observed tropical total column integrated moisture and surface temperature 1988-2016 relative to 1995-2000 climatology using updated data from Allan et al. (2014) Surv. Geophys.

As global temperatures increase, the invisible, gaseous water vapour in the air becomes more abundant, to the tune of 7% for each degree Celsius of tropical warming. This is incontrovertible, observable (see Figure 1) and has huge implications for climate. Water vapour amplifies climate change through the most well-known vicious cycle (warming leads to more atmospheric moisture and a more potent greenhouse effect which feeds back on the warming, amplifying the magnitude of the climate’s response to rising greenhouse gases). This increase in moisture crucially also fuels a greater intensity of rainfall.

When considering the vast areas experiencing the seasonal progression of the tropical rainy belt, the simplest result of a warming planet is the intensification of the wet seasons. The chaotic nature of weather patterns dictates that seasonal rains may fail one year while be supercharged in another depending on the precise position and mood of this rainy belt as it straddles the hemispheres. Yet when conditions are right for a seasonal deluge, it will be more intense in a warmer world due to the more copious quantities of available moisture blown in by the winds.

Things get more complicated when considering the dry zones and dry seasons. Averaged over many years, more atmospheric moisture arrives from over the ocean than is exported from land. If this wasn’t the case, then rivers would have to run in reverse! But for the dry season, a warmer, thirstier atmosphere can more effectively sap the ground of its moisture and temporarily export moisture out of the region. All this points toward an intensification of the wet and dry seasons as the planet warms but this simplistic message is complicated by several factors which include:

  1. The tropical rainy belt alters in precise location from year to year and its position also responds to climate change;
  2. Tropical cloud and winds rapidly adjust to modified heating patterns from greenhouse gases and aerosol particle pollution that ultimately are driving global climate change over longer time-scales, the slow pace of which is determined by the capacity of our vast oceans to take up excess heat;
  3. The effect of these radiative forcings of climate and chaotic weather fluctuations confound detection of wetter wet seasons and drier dry seasons that relate to increasing moisture;
  4. Possible long-term changes in moisture within the soil can feed back on the atmospheric circulation and rainfall characteristics, complicating the picture locally; 
  5. Finally, how arid a location is depends on much more than merely precipitation.

While “dry-gets-drier” does not apply in a simplistic way over land, intriguing signals of amplification in the seasonality of rainfall are apparent over tropical land with chaotic fluctuations superimposed upon longer-term trends (Figure 2). More intense dry seasons have potentially serious consequences for ecosystem productivity. Yet these signals are confounded by dominating variations in weather patterns and by movement in the climatological locations of wet and dry regimes over time. Better defining of the wet seasons is an important step in detecting and monitoring changes important for impacts on societies. The intensification of wet and dry seasons appears a likely but not guaranteed response in a warming climate yet substantial changes in rainfall, increases or decreases, are projected for considerable proportions of tropical land over the 21st century.

Figure 2: Changes in rainfall over tropical land for the wet season (top) and the dry season (bottom) in gauge-based observations (blue, red), an atmosphere-only simulation with prescribed observed sea surface temperature (brown) and fully-coupled climate models (black with shading showing ±1 standard deviation across 15 models). Updated from Liu and Allan (2013) ERL.

Figure 2: Changes in rainfall over tropical land for the wet season (top) and the dry season (bottom) in gauge-based observations (blue, red), an atmosphere-only simulation with prescribed observed sea surface temperature (brown) and fully-coupled climate models (black with shading showing ±1 standard deviation across 15 models). Updated from Liu and Allan (2013) ERL.

The where, when and scale of drought and deluge will continue to centre on the complex evolution of the atmosphere and ocean circulation but weather patterns are already being nudged away from their climatological normal. Limiting impacts requires substantial and sustained cuts in greenhouse gas emissions to tackle the root causes of warming but also crucial is better preparedness through supporting the development of climate resilient societies.

References

Allan et al. (2014) Physically consistent responses of the global atmospheric hydrological cycle in models and observations, Surv. Geophys., doi:10.1007/s10712-012-9213-z.

Berg et al.(2016) Land-atmosphere feedbacks amplify aridity increase over land under global warming, Nature Climate Change, doi:10.1038/nclimate3029

Bony et al. (2013) Robust direct effect of carbon dioxide on tropical circulation and regional precipitation, Nature Geoscience doi:10.1038/ngeo1799

Byrne & O’Gorman (2015) The response of precipitation minus evapotranspiration to climate warming: Why the “wet-get-wetter, dry-get-drier” scaling does not hold over land, J. Climate, doi:10.1175/JCLI-D-15-0369.1

Chadwick et al.(2015) Large rainfall changes consistently projected over substantial areas of tropical land. Nature Climate Change, doi:10.1038/nclimate2805

Chou et al. (2013) increase in the range between wet and dry season precipitation, Nature Geoscience, doi:10.1038/ngeo1744

Dunning et al. (2016) The onset and cessation of seasonal rainfall over Africa, J Geophys. Res., doi:10.1002/2016JD025428

Greve et al. (2014) Global assessment of trends in wetting and drying over land, Nature Geoscience, doi:10.1038/ngeo2247

Hegerl et al. (2015) Challenges in quantifying changes in the global water cycle. Bull. Am. Meteorol. Soc., doi: 10.1175/BAMS-D-13-00212.1

Held & Soden (2006) Robust responses of the hydrological cycle to global warming. J Clim doi: 10.1175/JCLI3990.1

Liu & Allan (2013) Observed and simulated precipitation responses in wet and dry regions 1850-2100, Environ. Res. Lett., doi:10.1088/1748-9326/8/3/034002

Murray-Tortarolo et al. (2016), The dry season intensity as a key driver of NPP trends, Geophys. Res. Lett., doi:10.1002/2016GL068240

Trenberth et al. (2014) Global warming and changes in drought. Nature Climate Change, doi:10.1038/nclimate2067

Posted in Climate, Climate change, Climate modelling, drought, Numerical modelling | Leave a comment

Potential links between Arctic sea ice loss and mid-latitude weather: revisiting an influential earlier study

by Len Shaffrey

The Arctic is changing rapidly due to human emissions of greenhouse gases. Arctic sea ice extent has been declining by 12% per decade since reliable satellite estimates began in 1979. By summer 2012, Arctic sea ice extent had been reduced by nearly half of its value in 1979. The reduction of sea ice has been accompanied by rapid warming, with Arctic surface temperatures increasing at nearly twice the pace of global temperature rise.

The dramatic changes in the Arctic have raised questions about whether the loss of sea ice is having an impact on mid-latitude atmospheric circulation and on the frequency of extreme weather events. One of the first papers to address this question was Francis and Vavrus (2012) (hereafter FV12). FV12 presented observational evidence that Arctic sea ice loss (and in particular the rapid loss of Arctic sea ice loss after Summer 2007) was associated with a weakening of the mid-latitude jet stream over North America and the North Atlantic. The apparent weakening of the jet stream from 1980 to 2010 can be seen in the time series of October-November-December (OND) 500 hPa zonal winds shown in Figure 1 (which is based on Figure 3b in FV12). FV12 went on to argue that the weaker jet stream was associated with large, persistent meanders that resulted in extreme weather events such droughts and severe cold spells.

Figure 1. Time series of OND 500 hPa zonal winds (averaged over 140°W to 0°E and 40°N to 60°N) from 1979 to 2010 (black) and with a 5-year running mean applied (blue). Note that for the running-mean time series the 2 years at the beginning and end of the time series are determined by averaging 3 or 4 years years rather than 5. This figure is based on Figure 3b of Francis and Vavrus (2012). Data from NCEP reanalyses.

FV12 received substantial attention from the scientific community. Subsequent a large number of papers were published that also argued for a link between Arctic sea ice loss and changes in mid-latitude circulation (e.g. Lui et al. 2012; Tang et al. 2013, etc.) or argued that the links weren’t statistically robust (Barnes, 2013; Screen and Simmons, 2013; Barnes et al. 2014, etc.). For more details of the debate, see the review papers of Cohen et al. (2014), Barnes and Screen (2015) and Shepherd (2016).

As five years have passed since the publication of FV12, I thought it might be worth revisiting the paper to see if a few extra years of data might provide some additional insight. Figure 2 extends the time series of 500 hPa seen in Figure 1 to OND 2016. It also extends the time series backwards in time to 1948 to provide some additional context. Figure 2 shows no evidence for a weakening of the OND 500 hPa zonal winds over North America and the North Atlantic in the extended time series between 1948 and 2016. It also suggests that the apparent weakening of the OND 500 hPa zonal winds seen in FV12 was due to the choice of the end date of OND 2010. OND 2010 was characterised by the extremely negative North Atlantic Oscillation (NAO) pattern over the North Atlantic, which resulted in a weakened jet stream. Within a few months, the NAO and the strength of the jet stream would return to normal. In the following year, OND 2011, the NAO was strongly positive and the jet stream was stronger than usual. It seems that the apparent link between Arctic sea ice loss and the mid-latitude jet stream is very sensitive to adding a few extra years of data. It also suggests that the arguments and conclusions of FV12 haven’t stood the test of time.

Figure 2. Time series of OND 500 hPa zonal winds (averaged over 140°W to 0°E and 40°N to 60°N) from 1948 to 2016 (black) and with a 5-year running mean applied (blue). Data from NCEP reanalyses.

References
Barnes, E. A., 2013. Revisiting the evidence linking Arctic amplification to extreme weather in midlatitudes. Geophys. Res. Lett. 40, 1–6.

Barnes, E. A., Dunn-Sigouin, E., Masato, G. & Woollings, T., 2014. Exploring recent trends in Northern Hemisphere blocking. Geophys. Res. Lett. 41, 638–644.

Barnes, Elizabeth A. and James Screen, 2015. The impact of Arctic warming on the midlatitude jetstream: Can it? Has it? Will it?. WIREs Climate Change, 6, doi: 10.1002/wcc.337.

Francis, J. A., and S. J. Vavrus, 2012. Evidence linking Arctic amplification to extreme weather in mid-latitudes, Geophys. Res. Lett., 39, L06801, doi:10.1029/2012GL051000.

Liu, J., Curry, J. A., Wang, H., Song, M. & Horton, R., 2012. Impact of declining Arctic sea ice on winter snow. Proc. Natl Acad. Sci. USA 109, 4074–4079.

Screen, J. A. & Simmonds, I., 2013. Exploring links between Arctic amplification and mid-latitude weather. Geophys. Res. Lett. 40, 959–964.

Shepherd, T.G., 2016. Effects of a warming Arctic. Science, 353 (6303). pp. 989-990. ISSN 1095-9203 doi: 10.1126/science.aag2349

Tang, Q., Zhang, X., Yang, X. & Francis, J. A., 2013. Cold winter extremes in northern continents linked to Arctic sea ice loss. Environ. Res. Lett. 8, 014036.

 

 

Posted in Climate, Climate change, Cryosphere, Polar | Tagged | Leave a comment

Water vapour transport by tropical cyclones over East Asia

By Liang Guo

When talking about tropical cyclones (TCs), people tend to think about gusty winds and heavy rain. These weather phenomena impress us due to the immense impacts on our surroundings. However, these weather phenomena are short-lived. Most TCs reduce to much weaker weather systems in several days after making landfall.

Therefore, when discussing water vapour transport (WVT) in a larger and longer context, the contribution of TCs is often neglected. Especially over East Asia (EA), where the summer monsoon (EASM) is strong and dominates the WVT. However, according to observations (i.e., International Best Track Archive for Climate Stewardship, IBTrACs), EA and the adjacent north-western Pacific is also the most active TC basin. Therefore, quantitatively measuring WVT by the EASM and TCs will help us understand TCs’ role in long-term large-scale WVT.

Figure 1 shows the mean seasonal cycle (averaged between 1979 and 2012) of WVT by the EASM and TCs via two boundaries which are defined within the figure. The first obvious difference is the peaking period. The WVT peak of TCs is in late summer and early autumn, which is later than that of the EASM. This is because TC genesis is sensitive to the sea surface temperature (SST), and SST is higher in late summer and early autumn. As shown in Figure 1, in most of the time, the WVT by EASM is about one order of magnitude larger than that of TCs. However, with the EASM retreating from EA landmass in August, the contribution from TCs is as important as the EASM, and is even larger in September and October.

Figure 1. Seasonal cycle of monthly mean vertically integrated moisture flux passing through the southern (blue) and eastern (red) boundaries. The mean-flow moisture fluxes are shown as solid lines (using the left-hand vertical axis) and TC eddy moisture fluxes as dashed lines (using the right-hand vertical axis). The inner panel shows the definition of the southern and eastern boundaries. Positive values indicate moisture is transported into the EA landmass, and negative values indicate moisture is transported out of the EA landmass. Note that the scale of the right-hand vertical axis is an order of magnitude smaller than the left-hand vertical axis.  Units: kg/s.

The second difference is the direction of WVT. The WVT by the EASM is imported via the southern boundary and is exported to the Japan and the Korean peninsula via the eastern boundary, while the WVT by TCs is in the reverse direction. The WVT import by TCs via the eastern boundary is mainly due to the westward-moving TCs, which bring warm and moist air from the Pacific. The WVT export by TCs via the southern boundary, on the other hand, is smaller and is linked to the northward moving TCs. The small magnitude is due to the air flow coming from land and is weaker and drier after the northward moving TCs making landfall.

An interesting detail of the WVT via the southern boundary is its sudden sign change in September, due to the reverse of the meridional gradient of the mean specific humidity over the EA landmass. During the EASM (June-August), the maximum of the mean specific humidity is located over south China (about 25°N, therefore, a southward gradient). However, with the retreat of the EASM in late August, the maximum of the mean specific humidity also rapidly retreats to near the equator (becomes a northward gradient).

As we can see, due to the difference in the season cycle compared to EASM, TCs’ role on WVT is non-negligible over EA.

Reference

Guo, L., Klingaman, N. P., Vidale, P., Turner, A. G., Demory, M.-E., & Cobb, A., 2017. Contribution of Tropical Cyclones to Atmospheric Moisture Transport and Rainfall over East Asia. Journal of Climate, DOI: 10.1175/JCLI-D-16-0308.1, in press.

 

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

Nice weather, atmospheric blocking and forecasts

By Oscar Martinez-Alvarado

With the beginning of spring (either the ‘meteorological’ spring on 1 March or the ‘astronomical’ spring on 20 March, as the Met Office explains here), the UK and indeed the whole Northern Hemisphere start experiencing warmer weather. The next few days are forecast to see settled weather over the UK with warm temperatures during the day, chilly overnight. This will be in part due to the prevalent large-scale atmospheric conditions, featuring what is called an episode of atmospheric blocking. An atmospheric block is a long-lived, usually more than five days, high-pressure system that remains almost stationary over a particular region. While the high-pressure system is there, other sources of weather variability such as mid-latitude cyclones and storms are deflected either to the north or to the south, so that the block acts as a shield for the blocked region against wet and windy weather. This might sound good, but if the blocking conditions remain for too long, this might lead to heat waves and drought in summer or very cold temperatures in winter.

Even though these blocks depend on the large-scale atmospheric conditions, which are generally easier to forecast than conditions at a particular location, forecasting the time when a blocking episode will start or finish is notoriously difficult. Figure 1 is an example of a blocking episode over Scandinavia, observed during the field campaign of the North Atlantic Waveguide and Downstream Impacts Experiment (NAWDEX, see also the blog entry on 31 October 2016). The blocking episode properly began on 4 October 2016 and remained there at least until 17 October. Two blobs representing atmospheric blocking on 4 October 2016, one over Greenland and a second one over Scandinavia, can be seen in Figure 1.Figure 1. Atmospheric blocking over the North Atlantic and Europe on 4 October 2016.

Figure 2 represents the same day, but as it appeared in the forecast six days earlier (i.e. the forecast issued on 28 September 2016). The colour shading represents the probability of having a block. White represents zero probability of a block; light yellow represents a very low probability of a block; dark red represents a very high probability of a having a block. The shading over Scandinavia indicated that a blocking over that region was almost equally likely to happen or not happen. Furthermore, the forecast is not at all confident about the occurrence of blocking over Greenland. In addition to these two locations, there was a very small but non-zero chance of blocking over Spain and a few other spots over Canada and the North Atlantic.

Figure 2. Atmospheric blocking over the North Atlantic and Europe on 4 October 2016 as forecast six days before on 28 September 2016.

Researchers at the Department of Meteorology and as part of the NAWDEX project are investigating the causes for the low blocking forecast skill in forecasts beyond five days and whether it is possible to improve it. It very well may be that this aspect of the atmosphere is actually unpredictable at those time scales, or it might be that we can improve our numerical models better to represent this phenomenon. Being able to predict blocking with confidence more days in advance would bring benefits such as planning ahead for the extreme weather conditions (e.g. heat waves, cold spells and drought) that might arise as a result. This makes the endeavour of improving blocking forecasts worth pursuing.

At the time of writing this blog entry, the 5-day forecast indicated that the current block over the UK should end on Monday 10 April, while a new strong block was forecast to be located over the North Atlantic. I am curious to know whether the forecast was correct or not.

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

Can the use of CCTV images improve urban flooding forecasts?

By Sanita Vetra-Carvalho

Urban flooding can result from intense rainfall, flash floods, coastal floods or river floods, the same as in rural areas. However, in cities, unlike in rural areas, there is very little open soil available for water storage and most floodwater needs to be to be transported to surface water or the sewage system. It is possible for rainwater entering the sewer system in one part of the city to exit through a manhole, flooding a different part of the city. Having early and accurate warning of potential flooding allows cities to prepare the drainage systems by ensuring there is an adequate water drainage capacity through sewage system and draining canals.

Due to the inherent complexity of cities characterized by a dense network of buildings with basements, roads, public transport, and a large number of people and businesses operating in close proximity, flooding in urban areas can be extremely costly and disruptive. The good news is that there are growing amounts of data available about our cities, such as CCTV images and citizen-sourced smartphone images, as well as scientific river gauge data and satellite images. Urban areas are rich with observation networks such as CCTV cameras looking at buildings, streets, parking lots, and rivers with images being available on a minute timescale. Many cities have a dense network of such cameras, and it is often open access, for example, London Traffic Cameras (JamCams) have around 800 CCTV traffic cameras distributed around London freely open to the community. Other organisations such as Highway Traffic Cameras have many cameras across England monitoring motorways and pass through/around cities. River cameras (e.g. Farson Digital Watercams) are another source of free open data (Figure 1).

Figure 1. Map of London Traffic Cameras, with a camera looking at the Thames

Currently this source of information is not used in producing flood forecasts, however, CCTV images have the potential to be very valuable in producing more accurate urban flood forecasts. The way to make most of such information is to use a technique called data assimilation (DA) which combines a model forecast with observations such as river levels and water extent in streets to produce a more accurate flood forecast. Using information from CCTV images and assimilating them directly into a flood forecasting model is one of the novel ideas behind the Data Assimilation for the REsilient City (DARE) project with an aim to assess the impact and benefit such novel observations can offer for urban area flood forecasting and improve the accuracy of urban flood forecasts.

Posted in data assimilation, Environmental hazards, Hydrology, Urban meteorology | Leave a comment

Melt ponds over Arctic sea ice

By Daniela Flocco

Melt ponds develop over Arctic sea ice during the melting season from the accumulation of melt water from ice and snow. These have become increasingly important over the last few decades because they have been more prevalent and absorb much more solar energy due to their dark colour compared to the highly reflective white sea ice (Perovich et al., 2002). Where ponds form, the ice beneath becomes thinner due to increased melting. Towards the end of the summer, the air temperature drops and a thin layer of ice forms over melt ponds. The ponds’ melt water trapped in the ice acts as a heat store and does not allow the underlying ice to start thickening until all the pond’s water is frozen. Ponds are up to 1.5 m deep and it can take over two months to freeze their volume of water. Considering that ponds cover up to 50% of the sea ice extent their impact cannot be neglected (Flocco et al., 2015).

Credit: Donald Perovich

A strong negative correlation exists between the change in successive mean winter ice thicknesses and the length of the intervening melt season, suggesting that summer melt processes play a dominant role in determining mean Arctic sea ice thickness for the following winter (Laxon et al., 2003). Another indication of the importance of melt ponds in explaining thinning of sea ice is that melt ponds are present in the Arctic more than in the Antarctic, where the sea ice thinning is less striking.

Ponds are rather irregular in shape but occur at a higher percentage over thin young ice: since the area of young ice is increasing (relatively to the total amount of ice which is instead decreasing), the impact of melt ponds will also become increasingly important. This will lead to a positive feedback effect in which thin ice will start thickening later in winter and will possibly be a preferential area for the formation of melt ponds in the following spring. Furthermore, corresponding to where melt ponds form, specular lenses of fresh water form under the sea ice cover, impacting the freezing point of water at the ice–ocean interface. At the beginning of the season sea ice is impermeable, so once ponds form they can be above sea level. When they start melting the ice, it becomes more permeable and when the ponds are fully developed they are in hydrostatic balance with the ocean so they drain to sea level.

Schemes handling melt ponds have only recently been included in global circulation models and are rather crude: the melt water was assumed to be flushed into the ocean without dwelling on the sea ice. Recent studies have shown that the lack of a melt pond parameterization can give an overestimation of sea ice thickness of up to 40% during summer (Flocco et al 2010, 2012). Model results have shown a good ability to forecast the minimum September ice extent, relating it to the melt pond area calculated by the model in May (Schröder et al 2014). This is one demonstration of how we have used the principles of physics to understand the changes we have observed in the cryosphere.

References

Flocco, D., D. L. Feltham, and A. K. Turner, 2010. Incorporation of a physically based melt pond scheme into the sea ice component of a climate model. J. Geophys. Res., 115, C08012, doi:10.1029/2009JC005568.

Flocco, D., D. Schröder, D. L. Feltham, and E. C. Hunke, 2012. Impact of melt ponds on Arctic sea ice simulations from 1990 to 2007. J. Geophys. Res., doi:10.1029/2012JC008195.

Flocco, D., D. L. Feltham, E. Bailey, and D. Schröder, 2015. The refreezing of melt ponds on Arctic sea ice. J. Geophys. Res. Oceans, 120, 647–659

Laxon, S., N. Peacock and D. Smith, 2003. High interannual variability of sea-ice thickness in the Arctic region. Nature, (425) October 30, 947-950.

Perovich, D.K., W.B. Tucker III, and K.A. Ligett, 2002. Aerial observations of the evolution of ice surface conditions during summer, J. Geophys. Res., 107 (C10), 8048, doi:10.1029/2000JC000449.

Schröder D., D. L. Feltham, D. Flocco, M. Tsamados, 2014. September Arctic sea-ice minimum predicted by spring melt-pond fraction. Nature Clim. Change, DOI: 10.1038/NCLIMATE2203.

Posted in Climate, Climate modelling, Cryosphere, Polar | Tagged | Leave a comment

Observation uncertainty in data assimilation

By Sarah Dance

Approximately 4 million properties in the UK are at risk from surface-water flooding which occurs when heavy rainfall overwhelms the drainage capacity of the local area. Several national weather centres have been developing new numerical forecasting systems to improve prediction of such events.  Weather forecasts are based on the output from a numerical computer model. Such models are built from a mathematical description of physical laws that govern the behaviour of the atmosphere and evolve an estimate of the current state of the system forward in time.  The estimate of the current state of the system may be obtained by a sophisticated mathematical blending of information from previous forecasts with recent observations in a process known as data assimilation.

Figure 1. Forecast-Assimilation Cycle

Remote sensing and data assimilation
In data assimilation we compare model forecast predictions and observations and adjust the model state so that it is closer to the observations, bearing in mind the uncertainty in the observations. However, the quantities predicted by the model are not usually the same as those being observed by the operational observation network. For example, in weather forecasting the model may predict wind, temperature, pressure and humidity. A weather radar on the other hand sends out pulses of electromagnetic waves and measures the intensity of the returned signal as the waves bounce off raindrops in the atmosphere. Thus we require a mathematical model that describes the physical relationship between the predicted quantities and the observations.  In data assimilation, this mathematical model is often termed the observation operator.

Observation uncertainty
When we compare the model predictions to the observations using the observation operator we are typically left with a residual known as the observation uncertainty. This image shows a measure of observation uncertainty for observations from the SEVIRI instrument used in numerical weather prediction over the UK. SEVIRI is a satellite-borne instrument measuring quantities sensitive to surface temperature (channels 7, 9 and 10) and upper level water vapour (channels 5 and 6) – see Figure 2. Channels 7, 9 and 10 are not used over land (the white areas in the picture) and the high uncertainty values shown around the coastline are due to a problem in the operational system quality control. Properly accounting for observation uncertainty in data assimilation is an important research topic. Our research has already led to improvements in operational weather forecast skill for global forecasting. Experiments are currently underway to find out if further improvements can be made for local forecasting of intense rainfall events.

Figure 2. Output from SEVIRI by channel

Reference

Waller, J. A., Ballard, S. P., Dance, S. L., Kelly, G., Nichols, N. K. and Simonin, D., 2016. Diagnosing horizontal and inter-channel observation error correlations for SEVIRI observations using observation-minus-background and observation-minus-analysis statistics. Remote Sensing, 8 (7). 581. ISSN 2072-4292
doi: 10.3390/rs8070581 http://www.mdpi.com/2072-4292/8/7/581 

Posted in data assimilation, earth observation, Numerical modelling, Remote sensing, Weather, Weather forecasting | Leave a comment

Confessions of an Admissions Tutor

By Hilary Weller

I am a postgraduate admissions tutor, so I see a lot of applications for PhD positions and I do a lot of interviewing. I would like to share some tips for applicants for PhD and post-doc positions and also some tips for interviewers. I have also done lots of interviewing for post-doc positions.

Your CV
If you are applying for a PhD position in a similar topic to your Bachelor’s or Master’s degree, then I really want to know how well you did in your degree – all the details – all the courses you did and the marks that you got in them. What did you do for your project? What mark did you get for your project? Did you do any relevant summer work? Include any other summer jobs or part time jobs – they tell me if you are hard working.

A CV for any position should document all periods of employment, unemployment and career breaks – even if you don’t want to tell me what the career break was for! I have two periods of maternity leave on my CV 🙂 which prompted positive comments from reviewers for my application for a fellowship. If you try to hide a career break I would be suspicious, but if you said “Jan 2014-Jul 2014: career break”, I would certainly be curious but I would be wary that I might be treading in a sensitive area. If you have had periods of unemployment, think of something useful that you did while unemployed. For example, did you read any relevant books, do any computer programming, spend time on any relevant hobbies or volunteering?

If you are asked for a personal statement I have some advice – tailor it to the job or PhD position that you are applying for, considering your aspirations as well as your experience, stick to the length limit, proof read it, then get someone else to proof read it.

Some very obvious interview questions to prepare for

  • Why are you interested in doing _THIS_ job/PhD?
  • What makes you well suited for _THIS_ job/PhD?
  • Tell us about a project that you have done?
  • Do you have any questions for us?

Some less obvious interview questions and tips for interviewers
For post-doc and PhD positions, it is important for interviewer and interviewee to find common ground. The obvious question “tell us about a project you have done” can leave the interviewer feeling uncritically impressed by the candidate. Conversely, questions along the lines of “what do you know about this aspect of my specific research area” can be unfair – a weak applicant with experience in the area could shine brighter than a strong applicant. So it is important to find areas of common interest. This requires preparation by the interviewer – you have their application and their academic transcripts – pick on a topic that you know about. Also, pick up on topics that the interviewee brings up and ask follow-up questions that you know the answer to. And interviewees – expect to be able to talk about anything mentioned in your application or transcripts! If you think that you did a relevant degree, you should be able to remember lots of what you learned.

An essential question for interviewing for a post-doc position – describe a paper that you have read recently.

Finally, remember that interviews are about candidates assessing the positions as well as vice-versa. So interviewers should be friendly and encouraging and offer plenty of information. Don’t spend too long on questions that the applicant is struggling with, move on with a smile rather than a shrug.

Posted in Academia, Teaching & Learning, University of Reading | Leave a comment