Characteristics and enhanced quality control of drifting buoy observations of sea surface temperature from the International Comprehensive Ocean Atmosphere Data Set (ICOADS)

By Simone Morak-Bozzo

Over the last two decades drifting buoys have become the most prevalent in situ measurement method for sea surface temperature(SST). Drifting buoy data are particularly popular because of their high spatial and temporal coverage. Their freedom of movement allows them to provide data offside shipping routes and moorings and the high temporal resolution gives insight not only into seasonal but also diurnal variations of SSTs, as shown in Morak-Bozzo et al, 2015.

Drifting buoy observations need to be verified before they can be used for quantitative studies involving SST data and cleaned from artefacts linked to erroneous observations or faulty instruments. The main archive of drifting buoy observations is the International Comprehensive Ocean Atmosphere Data Set (ICOADS). In our current project we investigate the life story of approximately 26000 buoys from 1979 to 2014, from the time of their deployment to their “death”. We also observe and document how the network of drifters evolves over time and we assess the quality of the data.

Figure 1: For the subperiod 1982-2012 and a total of 22678 buoys, the position of deployment is shown on the left and the position of their last signal on the right, gridded on a 2-by-2 degree grid.

Figure 2: This figure is showing the distribution of the buoy lifetime for all the buoys in the ICOADS archive. We can see that less than 40% of the buoys make it past the first year after deployment and less than 20% survives more than two years.

Figure 3: The maps are showing the cumulative number of drifting buoy observations in each 2-by-2 degree box for seven 5-year periods (1980-1984, 1985-1989, etc.). In the top right corner of each subfigure we show the sum of all observations in that 5-year period. The majority of buoys reports at least every 3 hours, if not hourly.

An important step to determine the reliability of the data is the creation of an automated protocol for quality control (QC) to complement the limited QC product provided by ICOADS. The set of algorithms we developed can detect e.g. stranded and misplaced buoys or non-plausible SST reports.

Compared to an independent data set of SSTs, such as Reynolds AVHRR-only (Reynolds et al. 2007, Banzon et al. 2016) the ensemble of QC flags shows a clear improvement of the quality of the drifting buoy data set.

The resulting data and associated uncertainty will be contributing to a new SST reconstruction dating back to 1850, under the project HOSTACE (Historical Ocean Surface Temperatures: Accuracy, Characterisation and Evaluation).

References:

Morak-­Bozzo, S., Merchant, C. J., Kent, E. C., Berry, D. I. and Carella, G. ,2016:Climatological diurnal variability in sea surface temperature characterized from drifting buoy data. Geoscience Data Journal, 3 (1),  20­-28, doi: https://doi.org/10.1002/gdj3.35.

Banzon, V., Smith, T. M., Chin, T. M., Liu, C., and Hankins, W., 2016: A long-term record of blended satellite and in situ sea-surface temperature for climate monitoring, modeling and environmental studies. Earth Syst. Sci. Data8, 165–176, doi:10.5194/essd-8-165-2016.

Reynolds, R. W., T. M. Smith, C. Liu, D. B. Chelton, K. S. Casey, and M. G. Schlax, 2007: Daily high-resolution-blended analyses for sea surface temperature. Journal of Climate20, 5473–5496, doi:10.1175/JCLI-D-14-00293.1.

Posted in Climate, Climate change, Data collection, Data processing, earth observation, Measurements and instrumentation | Leave a comment

Smoke, science, and sharks

By Ross Herbert

In the August of 2017 the Cloud-Aerosol-Radiation Interactions and Forcing – Year 2017 (CLARIFY) measurement campaign took place on a tiny island in the middle of the southeast Atlantic Ocean where we were surrounded by whales, sharks, and most importantly, stratocumulus clouds. During August and September dense layers of strongly absorbing smoke from biomass burning in central Africa were transported over these semi-permanent clouds. These events provided us with a unique opportunity to understand the interactions between the cloud, smoke layers, and radiation; processes which remain key uncertainties within our understanding of global aerosol radiative forcing, cloud feedbacks, and, ultimately, climate change. In this blog I will discuss my experience of participating in the measurement campaign and also outline the high-resolution cloud modelling work that I am currently doing.

The island (with Bear Grylls)

The Ascension Island is situated at 8°S 14°W in the middle of the Atlantic Ocean, 1600km from the coast of Africa and 2250km from Brazil. The island, with a diameter ~ 10km, is a volcanic island natively populated by nesting turtles, colourful land crabs, and (once) huge bird colonies, and non-natively by rats (that appeared with the boats), cats (that were let loose on the island to control the rats but decided to eat birds instead and are now outlawed), and donkeys (that are tolerated). In the mid-nineteenth century, following encouragement from Charles Darwin, the royal navy began bold plans to increase precipitation on the island by ‘greening’ the upland reaches of the largest mountain with anything that would grow. The result: bamboo, bananas, wild ginger, and guava, and the birth of the aptly named ‘Green Mountain’; however, most of the island remains volcanic and devoid of vegetation.

Figure 1:On top of Sister’s peak looking towards Green Mountain

The measurement campaign

My role on the month-long campaign was a mission scientist. Along with several others we worked as a team alongside the pilots to plan the sorties and then join the crew onboard the FAAM BAe-146 research aircraft. For the sorties we would use forecasts and satellite observations to plan flights depending on what our objectives were; this might include incoming plumes of smoke, co-ordinated satellite overpasses, inter-comparison flights, and interesting cloud or convective features.

Figure 2: The FAAM BAe-146 (right) and NASA P3 (left) on Wideawake Airfield

The pilots required detailed flight plans of bearings, altitudes, distances, and times, so the planning would often take several hours. On flight days we would be up at 6am to check the most recent observations and make any last-minute changes to the sortie plan before heading to the airfield. During flights, the mission scientists used real-time data and the expertise of the instrument operators to fine-tune the sortie so that we were focusing on the correct feature. Smoke layers and stratocumulus-to-cumulus transitions are very poorly forecast so we had to make rapid decisions based on what we observed during flight. This information would be relayed to the mission scientist who sat in the cockpit and would ultimately make the final decision before informing the pilots.

This slideshow requires JavaScript.

The campaign was a huge success; we flew 28 sorties totalling 99 hours and collected data in every cloud-aerosol-radiation regime we hoped for and more. As well as the standard atmospheric state measurements (temperature, pressure, relative humidity etc..) we made detailed measurements of the clouds, the particles, and the gases within the atmosphere and smoke layers. This included number, size, and composition, and also radiative properties such as ability to scatter and absorb radiation. The measurements will help towards improving our understanding of how the smoke layers are formed, transported, evolve, and also how they affect radiative fluxes and interact with the stratocumulus clouds. We can also use the measurements to improve our representation of smoke layers and related processes in the computer models that help us predict the climate and forecast the weather. FYI the aircraft data will be available on the Centre for Environmental Data Analysis (CEDA)archive (ask if you want more specifics).

End of campaign. Time to go back to the office

So enough of the campaign, what am I actually doing on the project? I am working with Nicolas Bellouin and Ellie Highwood to investigate what happens to the cloud radiative properties (i.e., how much energy from the sun and the earth’s surface passes through and out of the cloud) when the layer of smoke is elevated above the cloud; this is a scenario that we encountered numerous times during the campaign and that we observe in satellite observations. Previous studies suggest that the smoke may act to thicken the clouds, which makes them brighter and more reflective to incoming solar radiation (sunlight), thus acting to cool the climate. The stratocumulus clouds and smoke layers cover vast areas of the ocean, therefore small changes to the radiative properties of the cloud may have important impacts. My work will help us understand this effect in more detail and understand how sensitive it is to the properties of the smoke layer and cloud.

Smoke above cloud + Sunlight = Heat = Cloud response = Semi-direct radiative effect = Cooling? or Warming?

In my work I am using the MetOffice Large Eddy Model (the LEM) to simulate the evolution of a stratocumulus cloud deck with an elevated layer of absorbing smoke. The LEM has very high spatial resolution (imagine dividing the atmosphere into boxes – high resolution means lots and lots of little boxes) that allows us to simulate the main turbulent motions that drive the movement of energy, winds, and moisture in the atmosphere (typical climate-scale models have spatial resolutions ~ 200 times greater). It is also coupled to a radiation scheme that allows us to represent upward and downward fluxes of radiation through the atmosphere, smoke layer, and cloud. From this we can represent the additional heating that is caused by the smoke and any subsequent changes to the cloud field and radiative properties; we can then finally determine whether the smoke is acting to cool or warm the climate. This specific radiative effect is commonly referred to as the semi-direct effect and in the context of the IPCC AR5 report can be seen as a rapid adjustment to the instantaneous radiative effect of aerosols.

These simulations are giving us very interesting results that are currently being written up for publication (sneak preview: it’s not as simple as we previously thought). Watch this space!

A longer version of this blog can be found at https://rossherbert.org

Posted in Africa, Atlantic, Atmospheric chemistry, Atmospheric circulation, Atmospheric optics, Climate, Climate change, Climate modelling, Clouds, Data collection, earth observation, Energy budget, Environmental hazards, Greenhouse gases, Measurements and instrumentation, Microphysics, Numerical modelling, Solar radiation, Weather forecasting, Wind | Leave a comment

Determining the Earth’s energy and water cycles

By Christopher Thomas

The Earth’s energy and water cycles govern the distribution and movement of energy and water in the atmosphere, oceans and land. Both energy and water are constantly being transported between different regions of the globe, and the two cycles are closely linked (‘coupled’) together. It is very important to determine the size and variability of these transports as well as the long-term average distributions of each quantity. This will enable us to detect trends due to climate change, as well as to provide insight into short term climatic variability such as the El Niño Southern Oscillation.

The Sun provides energy to the Earth via solar radiation; simultaneously, some energy is emitted back into space. It is therefore quite common to consider the energy imbalance at the top-of-atmosphere (TOA). There is currently a small downward imbalance which means the Earth is gaining heat. The incoming radiation is distributed between the atmosphere, land and oceans both vertically and horizontally. Water is transported around the globe by (e.g.) ocean currents, rivers, evaporation, and precipitation. The last two processes provide the crucial link between the energy and water cycles. When water changes state from vapour to liquid or vice versa it either releases or absorbs energy in quantities which are significant enough that they must be included in the energy cycle.

Our aim is to combine a variety of Earth observation (EO) data sets in order to determine the energy and water cycles with as much precision as possible. This work was pioneered by the NASA NEWS team (L’Ecuyer et al. 2015, Rodell et al. 2015) and as a first step we have reproduced their work. The Earth is divided into 16 regions (seven land and nine ocean) as shown in Figure 1. In each region we consider the energy and water content both in the atmosphere and at the surface, as well as the TOA radiation imbalance (no water is lost to space).

Figure 1: The regions used in this study.

Several complementary EO measurements of energy and water transport are combined in each of these regions. These measurements are expressed as fluxes (the rate of flow of energy through an area). The larger the flux in a particular region, the more energy is being added to (or removed from) that region. Both vertical and horizontal fluxes are exploited in this method. Figure 2 shows the initial net vertical energy flux at the Earth’s surface in each of the 16 regions. The initial net flux is obtained by simply adding the downward fluxes together and subtracting the upward fluxes from the total. The values shown in the figure are the average of about a decade’s worth of observations.

Figure 2: Initial values of net surface flux in each region. Positive values indicate heat gain from the atmosphere and negative values indicate heat loss to the atmosphere.

Now, we know that there are various physical balances that should be respected (such as conservation of energy). The measured fluxes don’t satisfy these balances so we have to modify each flux by adding or subtracting a particular amount to it. When doing this it is particularly important to consider the uncertainty on each flux, which quantifies how well we (think we) have measured it. Observations that are poorly measured have large uncertainties and will be allowed to move more than those that are well known. For example, the TOA radiation has been extremely precisely determined, so has a small uncertainty, but the horizontal movement of water in the atmosphere is poorly understood and has a large uncertainty. The latter can therefore be moved relatively further from its initial value when trying to ensure all the physical balances are respected.

The net surface fluxes obtained in this procedure are reproduced in Figure 3. Some features of the solution indicate that it may be possible to make improvements. For example, the North Atlantic (region 13 in Figure 1) is losing less heat than may be expected. A large amount of heat is transported northwards from the South Atlantic by a process known as the Atlantic Meridional Overturning Circulation (see e.g. Buckley and Marshall 2016). This sort of discrepancy has motivated us to look into ways to incorporate extra information into the solution and hopefully mitigate the problems. One method which seems promising is to account for large-scale correlations in satellite measurements of heat loss from the ocean to the atmosphere (technically known as latent and sensible heat fluxes). If the flux measurements over the oceans are positively correlated then they are more likely to all be measured too high or low at once. This could arise due to the data processing methods used to determine the fluxes from satellite data, for example.

Figure 3: Output values of net surface flux in each region. Positive values indicate heat gain from the atmosphere and negative values indicate heat loss to the atmosphere.

We have estimated the heat flux correlations over the oceans and if we include them in the fit we find that more heat is transported from the South to the North Atlantic, which consequently loses more energy to the atmosphere. This is a promising result and we are currently preparing a paper to describe it more fully. We also intend to explore additional possibilities such as using additional data sets, including reanalysis constraints, and enhancing the spatial resolution of the solution by dividing the globe into more regions. In the long term we have ambitions to include the carbon cycle alongside the energy and water cycles.

References:

L’Ecuyer et al., 2015: The Observed State of the Energy Budget in the Early Twenty-First Century. J. Clim. 28(21), 8319–8346, 10.1175/JCLI-D-14-00556.1

Rodell et al., 2015:The Observed State of the Water Cycle in the Early Twenty-First Century. J. Clim. 28(21), 8289–8318,  10.1175/JCLI-D-14-00555.1

Buckley and Marshall, 2016: Observations, inferences, and mechanisms of the Atlantic Meridional Overturning Circulation: A review. Rev. Geophys. 54(1), 5–63, 10.1002/2015RG000493

Posted in Climate, Climate change, Data processing, earth observation, Energy budget, ENSO, Hydrology, Solar radiation, Water cycle | Leave a comment

Storylines of regional climate change

By Giuseppe Zappa 

An outstanding question for climate science is quantifying how global warming will regionally affect the aspects of climate that are most directly relevant to society, such as precipitation, windiness and extremes. But achieving this task is proving not to be simple. The main available tool consists in computer simulations performed using ensembles of climate models. These models are used to run scenarios in which greenhouse gas concentrations increase with time, so that the climate response to warming can be evaluated. But, for the aspects of climate that are controlled by the atmospheric circulation, there still remains substantial spread across the model projections thus leading to uncertainties in how regional climate will respond to global warming.

Figure 1:Decadal evolution of cold season (November to April) Mediterranean precipitation as a function of global warming in the simulations from 36 climate models from the 5th phase of the Coupled Model Inter-comparison Project (CMIP5) for the RCP8.5 emissions scenario. Precipitation and temperature changes are evaluated relative to the 1960 to 1990 mean. The thick line shows the multi-model. For presentation purposes, the horizontal axis starts at 0.2 K. 

Let’s consider Mediterranean precipitation as an example. Figure 1 shows that while the average of 36 climate model projections, as well as most of the individual models, indicate a future decline in winter Mediterranean precipitation, the magnitude of the precipitation reduction, even for a given warming of the planet, remains highly uncertain. Notably, the projected drying at 2 degrees warming in some models can be larger than the drying at 4 degrees warming in other models. Taking the multi-model mean provides a simple, and often adopted, approach to summarise the ensemble and communicate the regional projections to stakeholders and decision makers. But substantial information on the uncertainty is lost by simply averaging the model responses. So is this fully justified?

In a recent paper, Zappa and Shepherd propose to use an alternative storyline approach to characterise the uncertainty in regional climate projections from ensembles of climate models. To think in terms of storylines, it is necessary to realise that regional atmospheric circulation can be driven by remote aspects of climate. This is true on seasonal timescales, for example in response to the development of El Nino or La Nina events in the tropical Pacific, but it is equally true for the long timescales associated with climate change. In particular, Zappa and Shepherd identify two remote drivers of atmospheric circulation whose response to climate change is both uncertain and capable of influencing the European and Mediterranean climate: the magnitude of the upper tropospheric warming in the tropics and the strength of the Northern Hemisphere stratospheric vortex.

Figure 2:Four different plausible storylines of cold season Mediterranean precipitation change per degree of global warming. The different  storylines depend on the magnitude of the tropical amplification of global warming and on the strength of the stratospheric vortex response as indicated above the figures. See Zappa and Shepherd 2017 for more details. 

By applying a statistical framework to the climate models output, four different plausible storylines of Mediterranean precipitation change are identified for different combinations in the two remote drivers responses (see Figure 2). The patterns of regional precipitation change per degree of global warming within each storyline are found to be rather diverse: depending on the storyline, the Mediterranean precipitation response can be larger (Figure 2b) or weaker (Figure 2c), or it can be more focused on the eastern (Figure 2d) or on the western (Figure 2a) Mediterranean. A worst case storyline of Mediterranean climate change is identified for a large tropical amplification of global warming and a strengthening of the stratospheric vortex (Figure 2b), in which case the Mediterranean drying per degree of global warming is expected to be particularly enhanced. 

Until there is sufficient physical understanding or observational evidence to discard one of the above combination of driver responses, these four storylines should be considered equally plausible future realisations of Mediterranean regional climate change. It is worth to highlight that, until discarded, the worst case storyline could still be realised. This should be kept in mind when evaluating the risks of climate change and developing local adaptation plans. 

Reference:

Zappa, G. and T.G. Shepherd, 2017: Storylines of atmospheric circulation change for European regional climate impact assessment. Journal of Climate,30, 6561-6577, https://doi.org/10.1175/JCLI-D-16-0807.1

Posted in Atmospheric circulation, Climate, Climate change, Climate modelling, Greenhouse gases, Numerical modelling, Stratosphere, Troposphere | Leave a comment

How the Hadley Cells work

By Gui-Ying Yang

The Hadley Cell, named after British meteorologist George Hadley who discovered this tropical atmospheric overturning circulation, is one of the basic concepts in weather and climate. Figure 1 shows the zonal mean overturning circulation in a latitude height plane for Boreal summer June-July-August (JJA), based on 30 years (1981-2010) of ECMWF data. It is seen that the JJA Hadley Cell is dominated by its ascent near 10°N and descent near 20°S, with motion towards the summer hemisphere near the surface and a return flow towards the winter hemisphere in the upper troposphere. This classic picture of the zonally averaged Hadley Cell gives a smooth impression of the cross-equatorial flow moving from one hemisphere to the other. The basic theories of the Hadley Cell are based on angular momentum conservation with the additional consideration of some mixing and friction near the surface. However, angular momentum conservation from zero velocity at the equator moving to another latitude, φ, gives a zonal wind u=aΩsin2 φ/cos φ (134 m s-1 at 30o latitude and 700 m s-1 at 60o latitude) that is many times larger than that observed as seen in Figure 1. It is clear from Figure 1 that the angular momentum is far from uniform and the motion crosses angular momentum contours. Consistent with this, the actual subtropical jet maximum of about 40 m s-1 at 30oS is very much smaller than the value suggested by the theory based on angular momentum conservation in the upper branch of the Hadley Cell. This implies that eddy angular momentum mixing processes are actually of order one importance. In this study, we will reveal some interesting features associated with the Hadley Cell.

Figure 1:  The JJA zonal mean overturning circulation in a latitude-height cross section, based on 30 years of ERA-Interim data.  Colours indicate absolute angular momentum and blue contours indicate zonal winds.

Firstly, to investigate the nature of the JJA momentum flux, Figure 2 shows the steady and transient momentum flux in JJA 2009. It is seen that in addition to the expected strong transient momentum flux in the midlatitude, whose importance for the Hadley Cell has often been stressed, both steady and transient fluxes show a maximum in the tropical upper troposphere, extending from the region of tropical convection in the summer hemisphere into the sub-tropics of the winter hemisphere. The dividing line between tropical positive values and sub-tropical negative values looks almost identical in the steady and the transient, suggestive that there is a motion with SW-NE tilts north of about 15oS and NW-SE south of that with an amplitude that fluctuates in time. The latitude-longitude pictures at 150 hPa show that this signature comes predominantly from the Indian Ocean (not shown). This can be seen in the case studies below (Figure 5).

The Boreal winter picture indicates a similar tropical upper-tropospheric flux maximum but with the sign reversed as expected (not shown).

This indicates that angular momentum flux in the tropical upper troposphere, which has been neglected, is very important for the existence of the Hadley Cell.

Figure 2: Northward momentum flux of westerly wind in JJA 2009.

Then to examine the zonal and temporal variation of winds and convection in the Hadley Cell region, Figure 3 shows the JJA mean motion at (a) 200 hPa and (b) 950 hPa and Outgoing Longwave Radiation (OLR).  It is clear that the zonal average motion described by the Hadley Cell occurs in longitudinally confined regions that can be associated with the tropical convective heating regions (Low OLR). The lower tropospheric meridional motion contains the flow from the S Indian Ocean into the S Asian and the W Pacific regions of convection. It also shows the flow from the S Hemisphere into the E Pacific and Atlantic Inter Tropical Convergence Zone (ITCZ) heating regions. In the upper troposphere there is a return flow in each of these regions.  

Figure 3: Climatology JJA motion.

To illustrate the transient behaviour, Hovmöller of 5°N-10°S v at 200hPa (used to show the upper-tropospheric cross-equatorial motion) and  0-20°N OLR for 2009 JJA season are presented in Figure 4. The motion is seen to be localised in longitude and time.  The cross equatorial motion in the upper troposphere is strongest in transient waves associated with convective events over the Indian and W Pacific region.

 Figure 4: Transient motion in JJA 2009.

Finally, individual synoptic events in different longitudinal sectors are analysed. As a case study, Figure 5 (a),( b) show the upper and lower tropospheric horizontal winds for early July 2009 with contours of Potential Vorticity (PV) and OLR, respectively. On 6 July when convection is predominately in Indian sector, the lower tropospheric inflow is seen to have its origin from 30-40oS near 60oE. Two days later (8 July), in the upper troposphere, return flow reaches 30oS where it interacts with the S Hemisphere winter subtropical jet and the eastward moving synoptic waves on it, with a horizontal tilt consistent with that suggested by the momentum flux shown in Figure 2.  A filament of N Hemisphere positive PV moves towards the anti-cyclonic side of the S Hemisphere jet.

On 11 July, when convection is predominately in the Philippine sector, similar features are seen in a region shifted to the east.

Figure 5: (a) 370K PV and 150-hPa winds and (b) OLR and 950-hPa winds in early July 2009, 5 days apart with OLR and lower level winds leading the upper level features by two days.

In summary, this study gives evidence that:

(1) The existence of the Hadley cell involves not only the expected strong transient momentum flux in the midlatitude, but also the strong momentum flux in the tropical upper troposphere.

(2) The upper branch of the Hadley Cell is concentrated in certain longitudinal sectors and intensified cross-equatorial flow is associated with flaring in organised convection in those regions.  The tropical upper tropospheric motions associated with convection are crucial to the existence of the Hadley Cell.

(3) Filaments of the upper tropospheric air move from the summer to the winter hemispheres or are mixed in; they can reach the anti-cyclonic side of the winter subtropical jet and interact with the weather systems on it.

These observed features/processes have important implications on weather, climate and climate change, therefore it is important to know how well they are represented in weather and climate models. Also knowing that the cross-equatorial transfer of trace chemicals in the atmosphere occurs in filaments may have significant implications for atmospheric chemistry models, with almost undiluted summer hemisphere air moving deep into the winter hemisphere.

Posted in Climate, Climate change, Climate modelling, earth observation, Equatorial waves, extratropical cyclones, Tropical convection, Troposphere, Waves, Weather, Wind | Leave a comment

Mechanisms of Climate Change in the Indian Summer Monsoon

By Jon Shonk

Over one billion people are reliant on the rainfall of the Indian Summer Monsoon. During the wet season, which usually spans June to September, some parts of India receive over 90% of their total annual rainfall. Deficits or excesses of rainfall can have devastating effects, such as drought, inundation, crop failure and health issues. Bouts of extreme weather, such as short periods of very intense rain, can also have detrimental effects via flash flooding and the triggering of landslides.

It is therefore important to get an idea of how monsoon rainfall might change in a warmer future climate. Climate prediction uses numerical models to advance an initial global “snapshot” of the atmosphere and ocean forward in time using a supercomputer, and then examines the statistics of the weather over some period in the future. Nowadays, many institutions around the world run their own climate models. While these are all based around the same physical principles, the formulation and structure of the models can be very different. This means that the behaviour of models, even if initialised from the same global snapshot, can be quite different after 100 years of simulation.

The five maps in Figure 1 show the projected change in rainfall (averaged from June to August) over India in a future world that is 1.5 °C warmer than pre-industrial conditions (about 0.8 °C warmer than today), for five different climate models. There is a clear disagreement in the pattern of change, with no obvious consensus on which parts of India are likely to become wetter or drier.

Figure 1. Projected changes in rainfall as a result of a 1.5 °C warming, according to five climate models. Rainfall is averaged over June, July and August. Data from the HAPPI project (see Mitchell et al, 2017 for details).

So can we infer anything about the future Indian Summer Monsoon from these models? An advantage of using multiple models is that we can build an “ensemble” of predictions – that is, a number of plausible future climate projections. But the challenge is then how to statistically combine the projected changes to produce a single, clear, robust message.

The simplest option is to take an average across the projections from the five models (Figure 2a). The result is a weak pattern of slightly wetter conditions over eastern India and Bangladesh. However, the averaging process leads to areas where an increase of rainfall in one model cancels out a decrease in another, and understanding the reasons why models project such differences could provide extra clues as to how the monsoon might change.

Figure 2. Projected changes in rainfall, shown as the average across the same five models used in Figure 1. The changes for a warming of (a) 1.5 °C and (b) 2.0 °C are shown. Data from the HAPPI project.

By examining the behaviour of the models individually we can build an idea of the mechanisms by which the rainfall distribution changes in a warmer climate. This has been the focus of my recent work. I have also been looking at the differences in rainfall change between a world that is 1.5 °C warmer and one that is 2 °C warmer (Figure 2b). A paper on this should be ready soon…

 

Posted in Climate, Climate change, Climate modelling, drought, Environmental hazards, Flooding, Monsoons, Numerical modelling, Oceans | Leave a comment

Image conscious atmospheric science

By Giles Harrison

A frequently-heard mantra in physics is “Like charges repel and unlike charges attract”. At face value this paraphrase of Coulomb’s Law seems useful for clouds too, as, quite apart from the obvious example of thunderclouds, water drops in clouds are almost always charged to some extent. However, as it turns out, there are further subtleties to explore in the case of cloud droplets. The simple summary of the 1785 experimental findings of the French engineer and physicist, Jean Auguste Coulomb, made using a sensitive torsion balance, only applies to point charges, which, small though they are, cloud droplets are not. In fact they are sufficiently large for the charge within them to move around, i.e. to use a technical description, water droplets are polarisable. This means that, should there be another charge nearby, the charges within a water drop will re-arrange themselves in response. If this second charge is carried by another droplet, the charge in one will be re-arranged in response to the charge in the other. This is electrostatic induction: overall, the total charge on each droplet does not change, but its distribution within the droplet alters.

This concept is visualised below in figure 1. In the left-hand picture, there are two drops, both carrying negative charges. If they were solely point charges, they would repel each other in accordance with Coulomb’s Law. In the right-hand picture, in which a smaller droplet has been moved closer to the larger drop, a positive charge – known as an image charge – is induced on the droplet’s side of the drop by the negative charge. If the drops were brought closer still, the induced image charge in one would induce a stronger opposite charge in the other, which, perhaps counter-intuitively for two negatively charged objects, leads to attraction. Consequently when charged drops are driven together by turbulent motions and collide, the strong electrostatic attraction which always occurs between the image charges is likely to make them coalesce, and discourage them bouncing off each other. Collision and coalescence occurs continuously in clouds, and allows drops to grow sufficiently that they can ultimately fall as rain. Our initial work indicates that this process is accelerated by droplet charging.

Figure 1. Electrical forces between a small water droplet and a larger water drop, each carrying an overall negative charge (left). As the droplet approaches the drop, a positive image charge is induced in the drop (right), leading to an attraction.

These and related matters were discussed at a recent workshop at Reading on Microphysics of electrified clouds. In work funded at Reading by the United Arab Emirates Rain Enhancement Programme, a team of scientists and engineers is investigating how droplet charging affects droplet collisions and the formation of rain, and whether this can be used practically to influence clouds. Our project is pursuing these questions using a combination of numerical modelling and experimental work. A novel aspect of the numerical work is inclusion of a full description of the turbulent flow usually present in clouds (figure 2).

 

Figure 2. A system of droplets subjected to a turbulent flow field. An animation of the simulation is available here.

A second strand of work concerns the electrical environment of clouds in the UAE. This has been little explored, so, to obtain new information, we have established an automatic measurement site that provides a combination of cloud and atmospheric electricity data (figure 3).  

Figure 3. Measuring equipment being installed in the UAE to provide continuous data on atmospheric properties. Data is obtained by remote interrogation from Reading.

Finally, we need an inexpensive and flexible means to actually get into clouds, to make further measurements and undertake experiments on the effects of introducing charge. As well as the established Reading techniques exploiting modified meteorological balloons, we are using Unmanned Aerial Vehicles (UAVs) for this, designed specially by our collaborators at the Engineering Department at the University of Bath (figure 4).

Figure 4. UAVs developed by the Engineering Department at the University of Bath. (a) launch system and (b) the airframe planned to carry the meteorological instrumentation for the field experiments. Test flights can be viewed here.

This combination of new technologies, surface monitoring equipment and numerical modelling is allowing direct exploration of charge effects in non-thunderstorm clouds. In this, we are conscious that the often neglected electrostatic image force between water droplets seems likely to play a central role.

Follow UK atmospheric electricity activities at ctrwiae.org and on twitter: @atmos_elect

Posted in Clouds, earth observation, Measurements and instrumentation, Microphysics, Numerical modelling, University of Reading, Weather | Leave a comment

Wind generation in the UK during the summer of 2018

By Daniel Drew

The record breaking summer of 2018 has featured in a number of recent blog posts (link1 and link 2), but one area not discussed is the impact of the prevailing hot, sunny and calm conditions on the electricity system in the UK- particularly the level of wind power generation. I was able to experience this first hand as in the spring of 2018 I started a 1-year placement in the energy forecasting team at National Grid (as part of the UKRI Industrial Innovation Fellowships ).

Figure 1: The daily mean wind power output for Great Britain during the summer of 2018. Based on data from  www.gridwatch.templar.co.uk

The proportion of the UK’s electricity provided by wind power has been growing rapidly over the last 10 years from only 1.5% in 2008 to 17% in 2017 (more than double the amount provided by coal). Wind generation is typically lower in the summer months, in 2017 wind provided 12.9% of the UK’s electricity needs from June to August. Several media reports have speculated this figure will be a lot lower for 2018, however we have to wait for the official energy figures to be published by the Department for Business, Energy and Industrial Strategy to confirm this.

Initial data provided by GridwatchUK, suggests that between June and August this year the level of wind generation was generally low. Expressed in terms of capacity factor (energy production as a proportion of the theoretical maximum), the level of wind generation in the UK was approximately 19.0%. Additionally, there were several periods where generation was persistently low for a number of days. For example, from the 11th – 14th July the capacity factor of wind generation was below 10% for 74 consecutive hours.

Given the relatively short period of time the wind farms have been operating, it is difficult to place events like this into context using measured power output data. Fortunately, work carried out by University of Reading and National Grid  developed a method for creating a synthetic long term hourly time series of wind generation (1980-present) based on the current distribution of wind capacity. Using the data produced in this study, shows that while the summer 2018 wind generation is lower than 38-year summer mean of 21.0%, it is far from the lowest value in the dataset (16.6% based on the meteorological conditions experienced in 1983). Additionally, the period of persistently low generation experienced in July (74 hours below 10%) occurs on average twice per year across the 38-year period. In summary, based on the figures currently available, the level of wind generation during the summer of 2018 was lower than average but within the limits of a 38-year climatology.

Posted in Climate, Climate change, Historical climatology, Renewable energy, University of Reading | Leave a comment

Clouds, climate and the Roaring 40s

By Richard Allan

In our new research we have traced large and long-standing biases in computer simulations of climate, affecting the tempestuous Southern Ocean, to errors in cloud that emerge rapidly within the atmospheric models. Biases evolve over time through knock on effects that shift the location of the battering winds known as the Roaring 40s. Our new method combines detailed computer simulations with observations of energy exchanged between the oceans and atmosphere that allowed us to better understand how deficiencies in climate models emerge in this key region for climate. This offers a route to improve the complex simulations necessary to make reliable climate change projections of the future.

The Southern Ocean is a pivotal component of the global climate system yet it is poorly represented in climate simulations, with significant biases in upper-ocean temperatures, clouds and winds. It plays an important role in the uptake of excess heat and carbon dioxide generated through human activities. However most coupled atmosphere-ocean climate models have substantial warm biases in Southern Ocean Sea Surface Temperature (SST) (see Figure 1) that have been linked to a lack of reflective super-cooled liquid water clouds in simulations. Our work has helped to elucidate the link back from the SST biases to cloud-related errors in absorbed sunlight and we identified a slower response of the region of intense winds affecting the Southern Ocean that further modify the biases.

 

Figure 1: Warm biases in simulated sea surface temperatures cover the Southern Ocean (large orange region near the bottom of the map) (IPCC AR5 Chapter 9, Figure 9.2(b)).

In our study we find that coupled climate simulations with warm biases in the Southern Ocean also receive too much heat flux at the surface in simulations using just the atmospheric part of the model (Figure 1). This suggests deficiencies that develop rapidly in the atmosphere are strongly linked with the long-term climatological bias in the simulations. Further analysis identified that too much sunlight due to unrealistic cloud is primarily to blame, consistent with previous research.

Figure 2: Link between sea surface temperature (SST) biases in coupled CMIP5 simulations and surface energy flux bias in the atmospheric component of the simulations (AMIP5) from Hyder et al. (2018)

To interpret the results a detailed framework was developed that resulted in a candidate for longest methods section of the year award! We attempted to summarise the main points in a schematic where we assume SST biases are linked to the energy budget of the upper mixed layer of the ocean as similarly applied in studies understanding ocean temperature variability. A further finding is that although initial deficiencies in cloud develop rapidly in simulations, the overall biases also relate to a response in the location of the “Roaring 40s” or more specifically the latitude of maximum westerly wind. As can be seen in Figure 3 below, positional errors in this “zonal wind maxiumum latitude” (ZWML) are also correlated with errors in the surface energy fluxes in the atmospheric simulations.

Figure 3: Errors in coupled model “zonal wind maxiumum latitude” (ZWML) correlate with errors in the surface energy fluxes in atmospheric simulations from Hyder et al. (2018)

Importantly, further detailed analysis demonstrates that our interpretive framework can be applied in targeting improvements to climate simulations that avoid “pasting over cracks” where one bias compensates for another. This offers a route to further improve the climate model simulations that are vital in providing realistic projections of how climate will change over the coming decades. The work was led by colleagues at the Met Office, involved a collaboration of many scientists and was conduced as part of the NERC DEEP-C and SMURPHS projects. The detailed research is available as an open access research paper in Nature Communications.

 

Posted in Climate, Climate change, Climate modelling, Clouds, earth observation, Energy budget, Numerical modelling, Oceans, Solar radiation | Leave a comment

Why was the sky Orange?

By William Davies

I was sitting in my house one morning in October 2017, engrossed in what I was doing. Gradually I noticed that an eerie darkness was smothering the natural light in the room. I stopped and looked outside. The sky was a dark orange! What was going on and where could I go for answers?

Earth observation satellites! These are good for this sort of event. From my experience using remote sensing instruments to study the atmosphere I know the value of this resource. This link: https://earthdata.nasa.gov/earth-observation-data/imagery provides access to a range of data useful for seeing what is going on with our planet. To see what ‘Worldview’ could tell us about 16th October 2017 look at Figure 1.

Figure 1. Worldview satellite image from 16th October 2017

The first thing Figure 1 tells us is that ex-hurricane Ophelia was playing a part – that’s the swirl of cloud to the west of the UK. There is a saying – ‘Red sky in the morning, shepherd’s warning’. That’s because our weather systems usually come from the west and the low morning sun to the east colours them red, as explained further below. But there is more – compare the dirty air over the UK with the whiteness of the storm’s centre. This air was coming from the south west – from the dusty Sahara and from Portugal and Spain where there were reports of wildfires.

These dusty, smoky particles are referred to by scientists as ‘aerosols’ along with other particles such as salt from the sea and nitrates and sulphates from pollution. Aerosols have a direct and indirect effect on the Earth’s climate. Their indirect effect comes about by playing the part of condensation nuclei that cause clouds to form (Davies et al., 2010). An increase in condensation nuclei means an increase in cloud formation with a reduction in water droplet size. This leads to an increase in cloud reflectance of sunlight – a cooling effect. The increase in reflectance happens because the total surface area of the water is greater when spread over more droplets (Twomey, 1974). The direct effect is that sunlight is absorbed and scattered by these aerosols (Davies and North, 2015). When aerosols absorb sunlight, this increases the atmospheric temperature – a warming effect. Some of the scattered sunlight is reflected back into space and this will also have a cooling effect. Understanding ‘aerosol – cloud’ interactions and the way that aerosols absorb and scatter sunlight is crucial in our understanding of the climate.

The red sky is caused by the way that light is scattered. Blue light has a shorter wavelength than red light and is scattered more easily by the molecules in the air (this is why a clear sky appears blue). Red light has a longer wavelength than blue light and is not scattered as easily. The orange sky was caused by the contribution the aerosols were making to the way the light was being scattered.

Here at University of Reading I am working on two projects that study the effect of aerosols. The CLoud-Aerosol-Radiation Interactions and Forcing Year 2017 (CLARIFY) field campaign flew from Ascension Island in the south eastern Atlantic. This has delivered airborne, surface-based and satellite measurements which will improve representation of aerosols and clouds and reduce uncertainty in their radiative effects in climate models. The Copernicus Atmosphere Monitoring Service (CAMS) provides analyses and forecasts that address environmental concerns relating to the composition of the atmosphere. At University of Reading we are producing climate forcing estimates for CAMS but there are many other CAMS teams across Europe covering a range of service themes.

Figure 2.  A recent CAMS dust aerosol forecast

Figure 2 displays a recent forecast for dust aerosol where it can be seen that dust off the north west coast of Africa was being blown over Spain and France towards the UK. By clicking on the icon at the top right of the global map one can choose a different type of aerosol to view.

Figure 3. A recent CAMS biomass burning aerosol forecast

In Figure 3 we can see the smoky aerosols that are the focus of the CLARIFY project. These biomass burning aerosols from Africa are emitted in August and September by agricultural waste burning and forest clearing. We can also see smoke from the wild fires in Northern California.

When I looked at these aerosol forecasts on 16th October 2017 the presence of dust and biomass burning aerosol over the UK was confirmed.

This orange sky was due to a combination of Ophelia, Saharan dust and wildfires over Portugal and Spain and was an unusual event which generated a lot of interest in the media (see https://www.bbc.co.uk/news/uk-england-41635906 ). It serves to remind us of the importance of aerosol research and the effect that varying aerosol optical properties can have on sunlight and on our climate.

References

Davies, W. H., North, P. R. J., Grey, W. M. F., and Barnsley, M. J.,2010. Improvements in aerosol optical depth estimation using multiangle CHRIS/PROBA images. IEEE T. Geosci. Remote, 48, 18–24. https://doi.org/10.1109/TGRS.2009.2027024

Davies, W. H., and North, P. R. J., 2015. Synergistic angular and spectral estimation of aerosol properties using CHRIS/PROBA-1 and simulated Sentinel-3 data. Atmos. Meas. Tech., 8, 1719–1731. https://doi.org/10.5194/amt-8-1719-2015

TWOMEY, S. A. ,1974. Pollution and the Planetary Albedo. Atmospheric Environment, 8, 1251–56. https://doi.org/10.1016/0004-6981(74)90004-3

Posted in Aerosols, Atmospheric chemistry, Atmospheric optics, Climate, Climate modelling, earth observation, Environmental hazards, Numerical modelling, Remote sensing, University of Reading | Leave a comment