22nd June 2021
Climate Science: The Basics
Prof Keith P Shine, Regius Professor of Meteorology and Climate Science, University of Reading
The 2015 Paris Agreement to the United Nations Framework Convention on Climate Change aims to limit “the increase in global average temperature to well below 2°C above pre-industrial levels” and pursue “efforts to limit the temperature increase to 1.5°C”. This talk aims to provide the climate science basics needed to understand how the “net zero” concept emerges from this goal. The first need is to understand the relationship between atmospheric CO2 changes and global temperature change and establish that CO2 changes are the major driver of climate change. This defines the approximate limits in CO2 concentration change that are consistent with the Paris Agreement goals. The second need is to understand what CO2 emissions from human activity are consistent with this concentration change. Because of the way our CO2 emissions perturb the natural carbon cycle, the temperature goal requires bringing our (net) CO2 emissions to zero.
Net Zero: What, Why and When?
Prof Myles Allen, Professor of Geosystem Science, Environmental Change Institute and Dept of Physics, University of Oxford; Director, Oxford Net Zero
To read the full abstract, please click here
Science Needed to Achieve Net Zero: A Policy View
Dr Sarah Honour, Head of Climate Science, Department for Business, Energy & Industrial Strategy (BEIS)
In the lead up to COP26 this November, there has bever been a greater political focus on achieving net zero and the goals of the Paris Agreement. Science has been instrumental in setting this target and will be essential if we are to meet it. This talk will look at a few of the many policy questions related how we can achieve the goals of the Paris Agreement, both globally and as the UK, and the role that science can play in answering them. Questions include:
What emission reduction technologies and paths take us to net zero and how certain are we that they will get us there? What are the implications of different pathways?
How will we know we’re on track to meet Net Zero and how do we know when we’ve got there?
What impacts will we see in a less than 2°C world and how do we adapt to those impacts?
This talk won’t have all the answers – it won’t even have all the questions but it will provide a policy perspective on some of the cross-cutting science needed to deliver a net zero future.
Room One - Climate
The Impact of Atlantic Multidecadal Variability in a Coupled Climate Model Ensemble
Dr Dan Hodson, Research Scientist, NCAS, University of Reading
During the 20th Century, North Atlantic sea surface temperatures (SSTs) underwent pronounced multidecadal variability. We examine the impacts of this Atlantic Multidecadal Variability (AMV) on climate in an ensemble of five coupled climate models at both low and high spatial resolution. We use a heat-flux SST nudging scheme specified by CMIP6's Decadal Climate Prediction Project Component C (DCPP-C) to impose a persistent positive/negative phase of the AMV in otherwise freely running coupled simulations. The large-scale response to the AMV involves widespread warming over Eurasia and the Americas, with cooling over the Pacific Ocean and monsoon regions, and a northward displacement of the inter-tropical convergence zone (ITCZ). The accompanying changes in global atmospheric circulation lead to widespread changes in precipitation. We use Analysis of Variance (ANOVA) to examine how this response changes between models and resolutions.
Climate and Composition Implications of a Hydrogen Economy
Dr Nicola Warwick, NCAS Research Scientist, NCAS, University of Cambridge
Hydrogen is a key component of future energy pathways to the target of net zero greenhouse gas emissions by 2050 recommended by the UK’s Committee on Climate Change. Moving to hydrogen would lead to a significant reduction in carbon dioxide emissions, but any leakage of hydrogen could have negative environmental effects. A global atmospheric model is used to explore the impact of hydrogen leakage on climate and composition. We find that hydrogen increases lead to an increase in radiative forcing due to increases in the methane lifetime, tropospheric ozone and stratospheric water vapour. Including co-benefit emission reductions offsets these increases to varying extents depending on the hydrogen leakage rate. We provide an update to the single literature estimate of the H2 GWP.
The Challenge of Temporal Merging to Obtain Seamless Climate Information Beyond Decadal Time Scales
Dr Daniel Befort, Postdoctoral Research Assistant, Atmospheric, Oceanic and Planetary Physics, Department of Physics, University of Oxford
There is a high demand for skilful, reliable and seamless climate information for the next 1-40 years. Traditionally, information for different time scales is provided by different sources, e.g. decadal predictions for multi-annual time scales or projections providing long-term information. As predictions are initialized with the observed climate state at the start of the integration, they are often more skilful for lead times of a few years (depending on variable and region) compared to uninitialized climate projections, designed to provide information on long-term variability. Thus, most useful climate information for the next 1-40 years would likely need to draw upon information from both sources.
In this work, we discuss some of the challenges encountered when aiming to temporally combine decadal predictions and climate projections, e.g. discontinuities of the central estimates or other moments of the distributions at the respective transition point. A framework -aiming to overcome these issues- based on the idea to constrain uninitialised climate projections using decadal predictions is presented. The application to surface temperatures over the North Atlantic Subpolar Gyre region in CMIP5 model simulations suggests that such a method can provide more skilful seamless information beyond decadal time scales. Besides discussing advantages and limitations of this approach, other methods to temporally merge information from decadal predictions and climate projections will be discussed.
The Role of Earth System Interactions in Climate Forcing
Fiona O’Connor, Scientific Manager, Met Office Hadley Centre
Quantifying forcings from anthropogenic perturbations to the Earth system (ES) is important for understanding changes in climate since the pre-industrial (PI) period and into the future. Here, using a range of model experiments following the protocols defined by the Radiative Forcing Model Intercomparison Project (RFMIP) and the Aerosol and Chemistry Model Intercomparison Project (AerChemMIP), we quantify and analyse a wide range of present-day (PD) anthropogenic effective radiative forcings (ERFs). Quantifying ERFs that include rapid adjustments within a full Earth System Model (ESM) enables the role of various chemistry–aerosol–cloud interactions to be investigated.
The anthropogenic perturbations analysed include changes in greenhouse gas concentrations, aerosol and aerosol precursor emissions, ozone precursor emissions, and land use. For methane and non-methane ozone precursors, their ERFs are strongly influenced by cloud adjustments through chemistry-aerosol-cloud interactions. In the case of methane, these interactions increase the magnitude of the forcing but in the case of nitrogen oxide emissions, the cloud adjustment offsets the forcing from ozone. We will present a process-based understanding of these interactions and suggest that rapid adjustments included in ERF estimates need to include chemical as well as physical adjustments to fully account for complex ES interactions.
Room Two - Air Quality and Composition
Changes In The Level of NO2 and O3 in the UK During The COVID-19 Pandemic
James Lee, Research Professor, NCAS
The COVID-19 pandemic has caused worldwide lockdowns resulting in a reduction of road traffic as people are ordered to stay at home. As emissions from traffic accounts for 31% of total NOx in the UK this has led to reductions in NO2 during the pandemic, approximately by 33% for NO2 across 2020. We examined concentrations of NO2 and O3 from a total of 30 automatic monitoring sites from the UK’s Automatic Urban and Rural Network (AURN) between January 2020 and March 2021. Urban background and urban traffic sites were chosen for the analysis to represent the effect on city-wide and roadside air pollution respectively across the UK. Data was compared to the same period from the previous 5 years as well as a random forest model which predicted ambient concentrations of NO2 and O3 based on present day meteorological conditions. Agreement was found to be good for the two methods. We used Google Transit Mobility as a proxy for traffic, allowing us to assess the relationship between traffic and NO2 from early 2020 to spring 2021. During the spring 2020 UK lockdown, NO2 levels near the roadside decreased by 52 ± 3% on average. However, the most recent restrictions (January - March 2021),which were similar in stringency those in the spring of 2020, have resulted in only a 28 ± 3% reduction in NO2 on average. This is most likely due to a greater contribution to NO2 from domestic emissions in winter compared to spring.
O3 concentrations were elevated with an average 29% increase across the urban background sites. The total oxidant, Ox, (sum of NO2 and O3) experienced marginal change (< 1%) indicating the changes in NO2 and O3 were largely due to photochemical repartitioning. This has highlighted the importance of O3 in urban locations in a future low NOx environment in the UK when electric vehicle fleets are adopted.
How have Surface NO2 Concentrations Changed as a Result of the UK's COVID-19 Travel Restrictions?
Helen Dacre, Professor of Meteorology, University of Reading
Restrictions as a result of the COVID-19 pandemic have led to fewer vehicles on UK roads. Since fuel combustion is responsible for a large fraction of UK emissions it is expected that surface NO2 concentrations would reduce as a result. However, over parts of the UK, surface NO2 concentrations have increased following the implementation of travel restrictions. NO2 measurements from 142 Automatic Urban and Rural Network sites are combined with meteorological data from the Met Office high-resolution weather prediction model to build site specific models. These models predict NO2 concentrations given no change in emissions. It is found that both meteorological and emission changes contribute to the observed changes in NO2 concentrations. Given no change in emissions, changes in meteorology between pre- and post-lockdown periods would have led to a mean increase in NO2 concentrations of +6%. Conversely, changes in emissions would have led to a mean reduction in NO2 concentrations of -18%, resulting in the observed total change in NO2 concentrations of −12%. However at some sites the reduction in emissions is smaller than the increase in NO2 concentrations due to meteorology. The largest increases associated with changes in the meteorology are seen at rural sites (+20%) where NO2 measurements are representative of large areas and thus dominated by the regional advection of secondary NO2 from Europe. Conversely, the largest decreases associated with reduced emissions are found at urban traffic and urban background sites (−27% and −14% respectively) where NO2 concentrations are representative of local areas and thus dominated by local reduction in emissions from vehicles. As lockdown measures are relaxed, NO2 concentrations are likely to return to pre-COVID levels, but these results demonstrate that changes in our behaviour can result in positive impacts on air quality and illustrate the effectiveness of travel-reducing strategies in urban areas.
Intensive Observations at the NERC Air Quality Supersites
James Allan, NCAS Scientist, NCAS, University of Manchester
The NERC air quality supersites are facilities in London, Birmingham and Manchester that house a large suite of state-of-the art composition and meteorological measurements, for the purposes of studying the evolving properties, sources and processes of air pollution in the UK. As part of the OSCA project and the wider Clean Air Strategic Priorities Fund (SPF), intensive measurements are taking place at the sites in the summer of 2021 and the winter of 2022. These involve a number of collaborators from a number of UK Universities, NCAS, CEH and the Met Office, bringing together a suite of online and offline measurements that will deliver a deeper understanding the processes that govern air quality in the UK and help us to meet the air quality challenges of the future, as we strive to continuously improve air quality while transitioning to a low-carbon economy and adapt to changes in climate. Here we will present an overview of the project and present preliminary results from the intensives.
Future Raman Lidar Plans at Chilbolton Observatory
Judith Jeffery, Lidar and Meteorological Instruments Scientist, STFC/NCAS
Raman lidar measurements of water vapour and aerosol extinction have been made on a case-study basis at the STFC Chilbolton Observatory since 2001, funded by NERC as a part of the Atmospheric Measurement and Observation Facility (AMOF) of the National Centre for Atmospheric Science (NCAS). Now there are plans to purchase a new Raman lidar system for the site to take advantage of the many advances in the technology in the previous 20 years including continuous, remotely-controlled operation. The system will meet the minimum requirements of ACTRIS (Aerosol, Clouds and Trace Gases Research Infrastructure) with scope for future upgrades. We will share plans for the new lidar and the measurements that will be possible.
Room Three - Weather
The Role of Atmospheric Science in Tomorrow's Severe Weather Warning
Prof Brian Golding, Fellow in Weather Impacts, Met Office
There is growing recognition that weather-related hazard warnings should take more account of user needs. Research in the WMO HIWeather project has shown these needs to include information on hazard impact and response, as well as on the hazard itself, and for information to be delivered through multiple channels and formats. At the same time communication technology and artificial intelligence are offering the delivery of messages automatically tailored to user preferences. For the past 60 years, meteorology has been driven primarily by advances in the technology for running NWP, with the aim of producing forecasts accurate enough to save lives and reduce damage. In the future, we may expect to see meteorology increasingly influenced by how those forecasts are used, and in particular by the technology for communicating forecasts and warnings. Taking this as a starting point, my talk will consider how this rebalancing of the warning process may change the priorities for atmospheric science. For instance, while user input will continue to drive an emphasis on the precision and accuracy of very short range severe weather forecasts, the shift from weather to its impact will reinforce the trend towards coupled hazard prediction and a greater use of statistical modelling.
Application of a Holistic Approach to Understanding Heavy Precipitation Response to Climate Change
Dr Abdullah Kahraman, Research Associate, Newcastle University
Increases in the frequency and intensity of short-duration heavy precipitation events and the impact on flooding is an important hazard of anthropogenic climate change. Prior studies have mostly focused on the driving thermodynamic factors, mainly stemming from an increase of water vapour in the atmosphere with warming. Here, we apply a holistic approach to understanding heavy precipitation response to climate change, combining high moisture, fast condensation (via updraft speed), and a potential long-duration factor, the speed of the heavily-precipitating systems. We also discuss changes in the remaining ingredient for high precipitation rates, namely the precipitation efficiency. Pan-European convection-permitting climate simulations are used to analyse the future changes in the co-existence of favourable conditions across Europe. We confirm that moisture increases are the most important for heavy precipitation changes, but increases in instability during summer months are also found to be a contributor. A dramatic frequency increase in slow-moving intense rainstorms is found, which is 45 % higher than increases in all storms. The storm speeds appear to be getting slower especially during the autumn, which is the peak season for intense rainstorms, resulting in enhanced flood risk. This is likely linked to weakening jets due to Arctic Amplification in the late warm season.
Detecting Robust Non-Stationary Signals from the Noise in Ensemble Data Using Wintertime Euro-Atlantic Circulation Regimes
Swinda Falkena, PhD Candidate, University of Reading
We use a regularised k-means clustering algorithm to identify a robust signal of the non-stationary dynamics for six wintertime circulation regimes over the Euro-Atlantic sector. On inter-annual timescales we find a predictable signal in the occurrence rate for two zonal flow regimes, the positive phase of the North Atlantic Oscillation (NAO) and a negative blocking over Scandinavia, with the signal strength being similar between observations and model. In contrast the model signal strength for an NAO-index is weaker by a factor of two compared to observations. The regime analysis suggests that the model underprediction of the NAO-index signal is mainly associated with the negative NAO phase, as we find poor predictability for the corresponding NAO- regime.
Weather Regimes in Southeast Asia: Sub-Seasonal Predictability of the Regimes and the Associated High Impact Weather
Paula L M Gonzalez, Research Scientist, NCAS, Adjunct Research Scientist , International Research Institute for Climate and Society (IRI) , The Earth Institute
This work considers the sub-seasonal predictability of two sets of weather regimes for South East Asia: a two-tiered assignment, that first considers large-scale patterns and then assigns synoptic-scale regimes, and a flat classification, which only considers the synoptic scale. In the two-tiered approach, the tier 1 large-scale regimes, which capture ENSO and seasonal variations, are each partitioned into South East Asia regional clusters that capture synoptic variability, including cold surges, the MJO, the Borneo Vortex and tropical cyclones. Both sets of regimes explain around 10% of seasonally anomalous precipitation variance over land.
The sub-seasonal predictability of both the standard and tiered regimes is assessed using UKMO GloSea5 hindcasts and forecasts for lead times of up to 5 weeks. We find that the GloSea5 system presents an accurate representation of the regimes’ climatology and a good level of skill for their assignment. Nonetheless, the predictability depends on the specific regimes and some significant forecast drifts are also identified. Additionally, the predictive skill of high impact precipitation events obtained statistically from the prediction of the regimes is assessed and compared with the probabilistic precipitation forecasts of the GloSea5 ensemble. We find that these regime-conditioned predictions skilfully predict heavy rainfall for longer leads than simulated precipitation (up to around day 20) and we explore further improvements through spatial and temporal aggregation of the predictions.
Mid-latitude Dynamics 1
Evidence of Reduced Wintertime Nordic Seas Storminess
Ben Harvey, Research Scientist, NCAS, University of Reading
Extratropical cyclones (ETCs) present one of the major weather risks in the mid and high latitudes. Understanding how and why ETCs will respond to climate change is an essential component of assessing future weather risks and informing climate change adaptation strategies. Climate model projections reveal a complex picture for North Atlantic ETCs, with increased storminess expected over NW Europe and reduced storminess expected further north, over the Nordic Seas. Motivated by recent evidence suggesting the Nordic Seas reduction is particularly strong in those models with the most realistic representation of the present-day storm track, this study investigates the extent to which a trend in Nordic Seas storminess may already be discerned from observations.
Reanalysis products exhibit a marked reduction in Nordic Seas storminess since the mid-2000s, coinciding with a period of rapid sea ice loss. This trend is present in both bandpass-filtered sea level pressure variance and Lagrangian cyclone-tracking diagnostics. However, the NE Atlantic is known to exhibit substantial decadal variability in storminess and the accuracy of the reanalysis products before the satellite era is questionable. To place the recent trend in a longer-term context we compute storm track diagnostics directly from long-term station observations in the region. There are only two such stations with continuous records extending back 100 years and these both show that the recent reductions are unprecedented during that period. Therefore, with the caveat that the available observational data in this region is severely limited, we conclude that the expected future reduction of storminess in the Nordic Seas region may already be present in observations.
A Climate Model Evaluation of 3 Hourly Compound Precipitation and Wind Extremes Over Europe
Laura Owen, PhD Student, University of Exeter
Extreme precipitation and winds can have a severe impact on society, particularly when they occur at the same place and time. Studies have investigated the frequency of co-occurring extreme precipitation and wind using observational data. Due to the rarity of such events, these results are limited when looking at the risk of very extreme events, since a large number of samples is needed to get robust estimates. Additionally, it is very difficult for estimates based on observations alone to help us understand the risk of future rare or unprecedented events. Using the UNSEEN method (UNprecedented Simulated Extremes using ENsembles) this risk can be estimated from large ensembles of climate simulations. The Met Office's Global Seasonal forecast system version 5 (GloSea5) model ensembles are evaluated against ERA5 reanalysis data to find out how well they represent extreme precipitation, extreme wind and extreme co-occurring events over Europe. This model has not been evaluated in such a way before and this is needed before the model can be used to estimate the likelihood of unprecedented events using the UNSEEN method. We find that although the intensity of precipitation and wind extremes differ between the model and observations (by up to 12 mm and 9 m/s), the frequency of co-occurring events is well represented. However, significant differences in frequency are found around and over some areas of high topography. The model’s co-occurring events at individual locations investigated occur with very similar synoptic patterns to ERA5, indicating that the compound extremes are produced for the correct reasons. The model ensembles can then be used to assess the present day likelihood of unprecedented 3 hourly compound precipitation and wind extremes for winter over Europe, and to find out how the North Atlantic Oscillation influences the frequency of co-occurring events over Europe.
Attributing the Role of Climate Change in the 2014 Wintertime Heavy Accumulated Precipitation Event in Southern England and the European Heatwave 2003, using Global Spectrally Nudged Storylines
Linda van Garderen, PhD Student, Helmholtz Zentrum Hereon
Extreme weather events are generally associated with unusual dynamical conditions, yet the signal-to-noise ratio of the dynamical aspects of climate change that are relevant to extremes appears to be small, and the nature of the change can be highly uncertain. On the other hand, the thermodynamic aspects of climate change are already largely apparent from observations and are far more certain since they are anchored in agreed-upon physical understanding. The storyline method of extreme-event attribution, which has been gaining traction in recent years, quantitatively estimates the magnitude of thermodynamic aspects of climate change, given the dynamical conditions. There are different ways of imposing the dynamical conditions. We present a method where the dynamical conditions are enforced through global spectral nudging towards reanalysis data of the large-scale vorticity and divergence in the free atmosphere, leaving the lower atmosphere free to respond. We simulate the historical extreme weather event twice: first in the world as we know it, with the events occurring on a background of a changing climate, and second in a “counterfactual” world, where the background is held fixed over the past century. The European 2003 heatwave shows a high signal-to-noise ratio on daily timescales and at local spatial scales. For the 2014 heavy accumulated precipitation event in southern England we see that the response of precipitation to changes in climate forcing differs strongly on the local scale.
Forecast-Based Attribution of a Winter Heatwave within the Limit of Predictability
Nicholas J. Leach, PhD Student, University of Oxford
Attribution of extreme weather events has expanded rapidly as a field over the past decade. However, deficiencies in climate model representation of key dynamical drivers of extreme events have led to some concerns over the robustness of climate model-based attribution studies. In addition to the concerns over model fidelity, it has been suggested that the conventional risk-based approach to event attribution, which does not impose significant conditioning on the event, may result in false negative results due to dynamical noise overwhelming any climate change signal. The Storyline attribution framework, in which the impact of climate change on each of the individual drivers of an extreme event is examined, aims to mitigate against these concerns. Here we propose a novel methodology for attribution of extreme weather events, in which a model that successfully predicted the event is used. The use of a successful forecast ensures both that the model is able to accurately represent the event in question; but also that the analysis is unequivocally an attribution of the specific event in question, rather than a mixture of multiple different events that share some characteristic. Since this attribution methodology is conditioned on the component of the event that was predictable at forecast initialisation, we show how adjusting the lead time of the forecast can flexibly set the level of conditioning desired. This flexible adjustment of the conditioning allows us to effectively synthesize between a Storyline (highly-conditioned) and a risk-based (relatively unconditioned) approach. We demonstrate this forecast-based methodology through a partial attribution of the direct CO$_2$ effect on the exceptional European winter heatwave of February 2019.
Tropical Meteorology 1
The Influence of the QBO on the Tropical Circulation in the Met Office Unified Model
Jorge L García Franco, DPhil Student, University of Oxford
Previous observational studies have suggested that the stratospheric quasi-biennial oscillation is linked to the Walker circulation but a short record and the influence of El Niño-Southern Oscillation makes the evaluation of the QBO influence difficult.
Simulations with a nudged-stratosphere in the UK Met Office HadGEM3 climate model are presented where the zonal winds in the equatorial stratosphere are relaxed to a reanalysis in atmosphere-only and coupled ocean-atmosphere simulations.
The impact of nudging over the tropopause region and the effect of the QBO on the tropical circulation, monsoons and convection are presented and discussed.
The Four Regional Varieties of South Asian Monsoon Low-Pressure Systems
Akshay Deoras, PhD Student, University of Reading
We explore four regional varieties of South Asian monsoon low-pressure systems (LPSs) that formed during June–September 1979–2018. These four regional varieties include LPSs over the Arabian Sea, Sri Lanka, and short-lived as well as long-lived LPSs over the Bay of Bengal. We use the ERA-Interim reanalysis and TRMM precipitation datasets to examine the track statistics, thermal and moisture structure, and precipitation contribution of each LPS variety. Furthermore, we explore the modulation of LPS activity by the tropical intraseasonal variability, which is commonly monitored by the Boreal Summer Intraseasonal Oscillation, Monsoon Intraseasonal Oscillation and Madden-Julian Oscillation indices.
The results show that short-lived LPSs that form over the Bay of Bengal are the most frequent, whereas those over the Arabian Sea are the least frequent. All four LPS varieties feature a warm-over-cold core thermal structure, and LPSs over Sri Lanka feature the weakest thermal as well as moisture structure. Short-lived BoB LPSs produce the most precipitation over eastern India, and the most LPS-related precipitation over interior parts of India occurs due to long-lived BoB LPSs. The tropical intraseasonal variability strongly modulates LPS genesis — the anomalous occurrence of Arabian and Sri Lankan LPSs peaks when the convectively active phase of the intraseasonal oscillation (ISO) is over these regions, whereas that of BoB LPSs peaks several phases later with the northward/northeastward propagation of the ISO.
Exploring the Relationship Between the QBO and the MJO in the Hadley Centre Climate Model
Dr Martin B. Andrews, Senior Scientist, Met Office Hadley Centre
Observations suggest that the stratospheric equatorial QBO (Quasi-Biennial Oscillation) influences the strength of the tropospheric equatorial MJO (Madden-Julian Oscillation) in boreal winter (December-January-February). The MJO is stronger during QBO-easterly compared to QBO-westerly conditions. This relationship is explored in a 500-year CMIP6 pre-industrial control simulation using the Hadley Centre climate model. We find no statistically significant relationship between the phase of the QBO and the strength of the MJO in the free-running simulation. We compare the simulated zonal mean wind, temperature, and residual circulations with ERA5 reanalysis to highlight regions where model dynamics diverge from observations
Simulated Modulation of Convective Heating Profiles over the Maritime Continent by the MJO in Convection Permitting and Parametrized Convection Models
Emma Howard, Research Scientist, NCAS
Atmospheric convection interacts with large-scale circulation primarily through the heating and drying associated with latent heat release as water vapour condenses. The vertical distributions of these heating and drying tendencies after modification by eddy fluxes (Q1 and Q2 of Yanai (1973)) are key to the large-scale atmospheric response to convection. However, these distributions are often poorly represented by atmospheric models that parameterise convection. An under-representation of the variability of the vertical profiles of convective tendency may be one reason why such models struggle to simulate the eastward propagation of the Madden Julian Oscillation (MJO) - the leading mode of sub-seasonal rainfall variability in the tropics (Holloway et al 2013).
This study investigates convective heating and drying tendency profiles in a 3-month long simulation of the Maritime Continent, generated using the regional UM. Two model resolutions are considered: one convection-permitting with a 2km grid-spacing, and the other with parameterised convection using a N1280 cut-out domain. We study the dependence of tendency profile shapes on the phase of the MJO, and the extent to which this dependence can be attributed to variations in the proportion of shallow, congestus, deep and stratiform clouds under each phase. Consistent with observational studies, we find that in the convection-permitting model the heating and drying profiles are more top-heavy when the MJO is active than when it is inactive. The convection-parameterised model simulates considerably less variability in tendency profile shapes, particularly above the freezing level. The contribution of diurnally propagating convection to the total convective tendencies is also considered.
Orographic Rainfall Processes in India - Results of the IMPROVE Project
Dr Andy Turner, Associate Professor of Monsoon Systems, NCAS, University of Reading
Much of the focus on Indian weather and climate research is on the Indian monsoon, which supplies 80% of annual rainfall. However, less attention is paid to the spatial variations in rainfall and controls exerted on weather by the regional orography - both in the summer and winter periods. The IMPROVE project (Indian Monsoon Precipitation over Orography: Verification and Enhancement of understanding) is motivated to understand the effects of orography on Indian precipitation as part of the diurnal cycle of convection, as well as its role in extreme events. IMPROVE considers two focal regions. The Western Ghats, which intercept the monsoon flow across the Arabian Sea, receive some of the most frequent and heaviest rainfall during summer as well as being subject to extremes such as the 2018 Kerala floods. Meanwhile, the Himalayas play a vital role in separating dry midlatitude flows from tropical airmasses and are subject to extremes during the summer monsoon, as well as in winter due to the passage of western disturbances - cyclonic storms propagating on the subtropical westerly jet. This presentation summarizes the key results of IMPROVE. Firstly, we examine the impact of orography on the observed convective diurnal cycle and assess its simulation in models at a range of resolutions including convection-permitting scales. MetUM and WRF model experiments are used to identify key mechanisms and test their capability at simulating scale interactions between forcing at the large scale from the BSISO and newly identified regimes of on- and offshore convection near the Western Ghats. An additional aspect to this work is the construction of a two-layer analytical model to test the behaviour of sheared flow perpendicular to a ridge analogous to the Western Ghats. Secondly, the role of orography in extreme events is considered. For the Western Ghats, this focuses on the interaction between monsoon low-pressure systems and the southwesterly flow in enhancing local rainfall, a key mechanism in the Kerala floods of 2018. For the Himalayas, we focus on characterising interactions between tropical lows and western disturbances in enhancing the orographic precipitation, in which four key types of interaction are identified. IMPROVE works towards a deeper understanding of orographic rainfall and its extremes over India and uncovering why such mechanisms may be poorly represented in models.
Incorporating Vulnerability Information into Impact-Based Forecasts for Drought
Dr Vicky Boult, Research Scientist, Dept of Meteorology, University of Reading
Impact-based forecasting (IbF), has received considerable attention in the humanitarian sphere owing to its potential to save lives and livelihoods. The World Meteorological Organization notes that the fundamental difference between weather forecasting and IbF is the inclusion of vulnerability information. However, how vulnerability information is best incorporated is unclear, particularly in the development of IbF programs for drought.
Common challenges relate to the availability and granularity of vulnerability data, but more fundamentally, questions around the definition of drought, the purpose of IbF schemes, and the relationship between vulnerability and impact are seen as major stumbling blocks in the development of drought IbF. We draw on our experiences to address outstanding questions and propose a way forward.
Impactful Dry Spells in Southern Africa: A Case Study of Zimbabwe
Innocent Gibbon T. Masukwedza, PhD Student, University of Sussex
The maize crop is considered to be the regional staple food for southern Africa and is grown during the austral summer months. The majority of farmers in the region over-rely on rainfed agriculture and this increases the risk of food insecurity as there is still a lot to be understood regarding the drivers of rainfall systems over the region. Periods in which the maize crop receives inadequate moisture amounts are known to negatively impact its health and ultimately the final yield. This study uses a novel system (the Tropical Applications of Meteorology using Satellite data – AgriculturaL Early warning sysTem – TAMSAT-ALERT) in determining how maize is affected by impactful dry spells (IDS) during the crop development phases. This is achieved by examining the characteristics of an agriculturally relevant drought metric generated by TAMSAT-ALERT - the water requirement satisfaction index (WRSI). This metric was produced by driving the system with a local planting day definition used in Zimbabwe and having the WRSI derived at daily temporal time scales (from the planting date until the harvesting date). The study enables the moving away from the numerous generic meteorological definitions of dry spells which are mostly based on continuous days in which a place does not receive precipitation to an impact-based definition. Based on the study, a definition of IDS is proposed and the atmospheric drivers linked to such events are also examined. Since TAMSAT-ALERT is an operational forecasting system capable of providing this agriculturally relevant metric at daily timescales within the cropping season, results from this study show that the system has the potential to inform food insecurity anticipatory actions.
Recent and Future Changes in Wet and Dry Season Characteristics
Dr Caroline Wainwright, Post-Doctoral Research Assistant, University of Reading, NCAS
Climate change will result in more dry days and longer dry spells, however, the resulting impacts on crop growth depend on the timing of these longer dry spells in the annual cycle. Using an ensemble of CMIP5 and CMIP6 simulations, and a range of emission scenarios, here we examine changes in wet and dry spell characteristics under future climate change across the extended tropics in wet and dry seasons separately. Delays in the wet seasons by up to two weeks are projected by 2070-2099 across South America, Southern Africa, West Africa and the Sahel. An increase in both mean and maximum dry spell length during the dry season is found across Central and South America, Southern Africa and Australia, with a reduction in dry season rainfall also found in these regions. Mean dry season dry spell lengths increase by 5-10 days over north-east South America and south-west Africa. Furthermore, observations show an increase in mean dry season dry spell lengths over the past 30 years over parts of South America. On the other hand, future projected changes in dry spell length during the wet season are much smaller across the tropics with limited model consensus. Mean dry season maximum temperature increases are found to be up to 3C higher than mean wet season maximum temperature increases over South America, Southern Africa and parts of Asia. Longer dry spells, fewer wet days, and higher temperatures during the dry season may lead to increasing dry season aridity, and have detrimental consequences for perennial crops.
Analysis of Rainfall Trends and Extreme Precipitation During the Minor Season of Ghana
Mohammed Braimah, Meteorologist, Ghana Meteorological Agency
Over the past few years, policy makers and the general public have raised concerns of increasing cases of rainfall during the minor season of Ghana. Linear regression and Mann Kendall Test were used to establish trends of total rainfall amounts and frequency during the minor season. Descriptive statistical analyses generated coefficient of variation values between 28.329% and 70.892% for rainfall amounts and between 14.244% and 48.811% for the minor seasonal rainy days. Linear regression equations for most stations over southern Ghana shows positive slopes for both rainfall and number of rainy days, a signal of an increasing trend. Coefficient of determination values which were calculated shows that rainfall variability between 0.7% and 10.4%, and variability in total rainy days between 3.3% and 27% were explained by the linear regression. Mann Kendall test suggest that the frequency of rainfall in the minor rainy season between 1981 and 2018 increased significantly but the change in total rainfall amounts is non-significant. Analysis of minor seasonal rainfall anomalies shows that rainfall in the middle sector and the west coast experienced mostly above normal rainfall from 2014 to 2018.
Synoptic Features Associated with Thunderstorms over Southern Ghana during the Minor Rainy Season - A Case Study on 28th and 30th October, 2019
Vincent Antwi Asante, Assistant Meteorologist, Ghana Meteorological Agency
Severe and destructive thunderstorms that occurred over southern Ghana on October 28th and 30th, 2019 has called for investigation into the synoptic features related to these rainstorms for impact in knowledge. The study was conducted by analyzing observed stations data, products from EUMETSAT – NWC SAF and ECMWF as well as numerical model simulation from WRF. It was observed that build-up of low pressure systems, significant drop in the 24-hours pressure difference, moisture influx and build ups to the mid-levels of the atmosphere, cyclonic vortexes and flow at the 925hPa, 850hPa and 700hPa levels caused atmospheric instability and hence the initiation, sustenance and movement of the thunderstorms.
Earth System Modelling
Release of UKESM1.1 An Updated Version of the UKESM1 Model
Colin Jones, Professor, NCAS, University of Leeds
We present an updated release of the UKESM1 model, UKESM1.1, in which a number of targeted improvements have been introduced. These include a more advanced treatment for the dry deposition of sulfur dioxide that takes into account whether the depositing surface is wet or dry. A number of model tunings have also been revisited, including those for atmospheric dust, the albedo of snow on sea ice and convective wave forcing of the Quasi Biennial Oscillation (QBO). Finally, a number of model bugs have been corrected. The updated model shows a significantly improved simulation of the historical surface air temperature when run using CMIP6 forcing. Improvements are also seen in other key variables such as historical ocean heat uptake, Arctic sea ice and top of atmosphere radiation fluxes. By contrast, there is little change in either the transient climate response (TCR) or effective climate sensitivity (ECS) compared to UKESM1.0. In this presentation, we outline the main developments introduced to UKESM1.1 and the key simulation improvements we find.
Comparison of the Surface Mass and Energy Balance of UKESM and MAR Forced by UKESM over the Greenland Ice Sheet: Present and Future
Charlotte Lang, Research Scientist, University of Reading
We have compared the climate and surface mass (SMB) and energy (SEB) balances of the Greenland Ice Sheet (GrIS) simulated by the Earth System Model UKESM and the regional climate model MAR forced at its boundaries by the former in an effort to identify processes and parameterisations in the land-surface/snow models of both models that might cause a different evolution of the snowpack properties when forced by the same climate.
Over the present ERA (1981 – 2010), the amount of melt compares very well for both models, at the exception of the area along the South west margin of the GrIS where melt is more important in UKESM due to a lower albedo. As snowfall is not downscalled in UKESM, the orographic precipitation on the South east margin is underestimated, resulting in a much lower SMB in UKESM than in MAR in that region.
At the end of the Century, the runoff increase with respect to the historical period and integrated over the GrIS is 32% smaller in UKESM than in MAR and is a consequence of a smaller melt increase (25%) and a larger amount of refreezing of the melt water in the snowpack (32%). As a result, the SMB simulated by UKESM decreases less than the MAR one (34%). This difference in melt evolution is due to a lower increase in net shortwave radiation as well as sensible and latent heat fluxes in UKESM, resulting in a lower amount of net energy absorbed by the snowpack in summer.
The summer near-surface temperature increase in 2100 is the same for both models (+9°) and the smaller increase in melt and net absorbed energy in UKESM can therefore not be attributed to a a difference in near-surface temperature increase but rather to a smaller sensitivity of the energy fluxes to the same temperature increase in UKESM.
Congo Basin Deforestation: How Does it Impact Vegetation and What Could This Mean for Land-Atmosphere Interactions in the Region?
Coralie Adams, PhD Student, University of Manchester
Despite its climatic importance as one of the large tropical forests, the Congo Basin is severely understudied compared to other tropical regions. Interactions between Congo Basin deforestation and rainfall are complex and often depend on the aggregate properties and spatial distribution of the remaining vegetation, rather than the amount of deforestation. Congo Basin deforestation is predicted to increase, and the distribution of deforestation drivers is also expected to change; it is, therefore, essential that we understand how deforestation impacts the climate. Previous studies have focused on future projections of large-scale deforestation, which are often unrealistic, and therefore analysis of how present-day small-scale deforestation affects the Congo climate has not been conducted. Observational studies looking at the impacts of Congo Basin deforestation on climate are very few and an assessment of how deforestation impacts surface properties and climate has not been done before. This highlights the need to investigate the chain of processes for how realistic deforestation impacts surface properties and climate.
In this study, we will use observational data to investigate how deforestation impacts surface properties and climate in the Congo Basin. Initial results will be presented from analysis of how deforestation affects vegetation properties by analysing vegetation indices retrieved by the MODIS satellite (EVI, NDVI, and LAI). Deforestation year and forest cover data were obtained from the Global Forest Change dataset 2000 - 2019. We find that deforestation in the region is typically due to small-scale agricultural clearing which has a short fallow cycle and therefore regrowth occurs very quickly as the vegetation indices often recover to their predeforestation values in as little as a year. However, the lack of ground-based meteorological stations and the prevalence of clouds in the region makes it difficult to obtain high-quality remote sensing data; in the austral summer, an average of 58% of data points are missing. The fast regrowth, coupled with the lack of high-quality remote sensing data, makes assessing deforestation’s impact on vegetation properties challenging. Understanding how deforestation impacts vegetation, and the subsequent process of regrowth, is essential to frame how deforestation impacts other surface properties and climatic variables. The next steps for this research include looking at other surface properties, like ET and LST, and how these are impacted by deforestation and subsequent regrowth.
Developing a Seamless Global Coupled Model in Partnership: The GC Programme
Dr Charline Marzin, Science Manager, Met Office
The UK Met Office has developed over the years an ambitious global coupled (GC) model development strategy to improve the Unified Model (UM)-based systems for applications across several timescales, from weather forecasting to sub-seasonal to seasonal and climate predictions. This is now formally governed within the GC programme directly involving key academic partners and international UM partners. The model development and evaluation activities have been adapted to focus on the fully coupled ocean atmosphere global model, and is currently delivering the first coupled configuration to be used operationally for numerical weather prediction at the Met Office as well as seasonal and climate applications. An ambitious seamless testing strategy has been developed to provide an extensive range of tests across timescales for proposed atmosphere, land, ocean and sea ice model changes throughout the model development process. A thorough quality assurance framework has also been devised to enable interaction with a wide range of experts, users and stakeholders of the global modelling systems. The GC programme aims to nurture and develop effective collaborations and prepare for next generation modelling systems. Some recent examples of how performance was improved through the release of the GC4 configuration will be given, as well as examples of ongoing developments towards the next GC5 release and the future next generation modelling system, LFRic.
Future Precipitation and Surface Melt on Ice Sheets in the UK Earth System Model: Implications for Projecting Sea-Level
Dr Robin S. Smith, Senior Research Scientist, NCAS, University of Reading
Our current limited ability to predict the future of the Greenland and Antarctic ice sheets under climate change means that ice sheets are the most uncertain term in projections of sea-level rise, with the worst-case predictions pushing global sea-level to a metre or more above the 20th century baseline by 2100. Several key ice-climate processes involve self-reinforcing feedbacks that, if triggered, could be practically impossible to reverse, making it critical that we use a coupled atmosphere, ocean and ice system to evaluate the possibility of such high impact scenarios. In the UK, our national Earth System Model is uniquely capable of simulating the physics of ice sheets in the context of globally evolving climate. Here we will report on our modelling of the atmospheric influences on the ice sheets under 21 century climate change, the balance of increasing precipitation with increased melt in key regions of Greenland and Antarctica and the overall implications for global sea-level rise. We will also identify avenues for future work important in improving our understanding of sea-level rise and climate change in these vulnerable regions.
Chemical Process Studies
Large Simulated Future Changes in the Nitrate Radical and Implications for Oxidation Chemistry
Dr Scott Archer-Nicholls, Postdoctoral Research Associate, University of Cambridge
The nitrate radical (NO3) plays an important role in the chemistry of the lower troposphere, acting as the principle oxidant during the night. Previous model simulations suggest that the levels of NO3 have increased dramatically since the pre-industrial. Here, we present what the authors believe are the first projections of the evolution of the NO3 radical from 1850-2100 using a global Earth System model: the UKESM1. Our model results highlight divergent trajectories for NO3, with some scenarios and regions undergoing rapid growth of NO3 to unprecedented levels over the course of the 21st Century, with others showing sharp declines. The local increases in NO3 are driven not only by local changes in emissions of nitrogen oxides but have an important climate component, with NO3 being favoured over N2O5 in future warmer climates. Changes in emissions of biogenic VOCs, such as monoterpenes, also play a role, with these varying depending on land use and climate. The changes in NO3 lead to changes in the oxidation of monoterpenes and their production of secondary organic aerosol, further contributing to particulate matter pollution regionally. This work highlights the potential for substantial future growth in NO3 and the need to better understand the formation of SOA from NO3 to accurately predict future air quality and climate implications.
Assessing Uncertainty in the Oxidation of DMS in Earth System Models
Bea Cala, MPhil Student, University of Cambridge
DMS (dimethyl sulfide), a gas released from phytoplankton, is the largest natural source of sulfur in the atmosphere. Its oxidation products impact the climate through aerosol and cloud formation, influencing the Earth's radiation budget. Natural aerosols like this contribute to large uncertainties in our understanding of changes in the climate system.
Currently, the implementation of gas-phase DMS oxidation chemistry in schemes such as UKCA StratTrop and CRI-Strat is insufficient, especially regarding key species for aesorol formation such as SO2 (sulfur dioxide) and MSA (methanesulfonic acid). Additionally, our understanding of DMS oxidation has been questioned by the recent discovery of a new DMS oxidation product, HPMTF (hydroperoxy methylthioformate). Currently, little is known about the fate of HPMFT in the atmosphere and its impacts. The data suggests that a majority of DMS is oxidized via the formation of HPMTF, a mechanism never before considered in models.
Through a box modelling approach, focusing on conditions representative of the remote marine boundary layer, a number of different DMS oxidation chemistry schemes are contrasted and evaluated and a possibility for improvements is proposed. Further, it is suggested which key reactions need to be added to reflect HPMTF chemistry and the effects in the box model are shown and discussed. This work highlights the large uncertainty in the representation of DMS chemistry in earth system models and the needs for more observational studies to constrain it.
AMCLIM – A Framework for Climate-Dependent Ammonia Emission Modelling
Jize Jiang, PhD Student, University of Edinburgh
Ammonia (NH3) is the primary form of reactive N (Nr), which has significant impacts on the environment, not only damaging ecosystems and water systems, but also affecting air quality and climate. Ammonia emission is mainly from agriculture and is found to be strongly influenced by climate through temperature and water interactions. Existing estimates used fixed emission factors (EFs) to calculate NH3 emissions, which may introduce large uncertainty because of the limited consideration of climatic effects. We developed a framework, Ammonia-CLIMate (AMCLIM), to quantify NH3 emissions whilst considering the effects of environmental factors and human management based on the understanding at the process level. This process-based emission scheme simulates and predicts the temporal variations of NH3 by following the nitrogen evolution pathways. Major simulated practices include livestock housing, manure management, and land applications of fertilizers. The overarching goal is to establish NH3 emission inventory for multiple agricultural sectors and to provide inputs to atmospheric models.
Improvements to the Representation of BVOC Chemistry-Climate Interactions in UKCA with the CRI-Strat 2 Mechanism: Incorporation and Evaluation
James Weber, PhD Student, University of Cambridge
We present the first incorporation and evaluation of the CRI-Strat 2 chemistry mechanism in the global chemistry-climate United Kingdom Chemistry and Aerosols (UKCA) model. CRI-Strat 2 comprises a state-of-the-art isoprene scheme, optimised against the MCM v3.3.1, which includes isoprene peroxy radical isomerisation and HOx-recycling. This increases low altitude OH in tropical forested regions by 75-150% relative to the standard UKCA mechanism StratTrop, leading to an improvement in model performance for surface OH, isoprene and monoterpenes. CRI-Strat 2 will enable a re-evaluation of the impact of BVOCs on the chemical composition of the atmosphere.
Ozone Budgets in UKESM AerChemMIP Experiments
Dr Paul Griffiths, NCAS Research Scientist, NCAS, Cambridge University
A grand challenge in the field of chemistry-climate modelling is to understand the connection between anthropogenic emissions, atmospheric composition and the radiative forcing of trace gases and aerosols. The AerChemMIP model intercomparison project, part of CMIP6, focuses on calculating the radiative forcing of gases and aerosol particles over the period 1850 to 2100.
We present an analysis of the trends in tropospheric ozone budget in the UKESM1 and other models from CMIP6 experiments. We discuss these trends in terms of chemical production and loss of ozone as well as physical processes such as transport and deposition. Where possible, AerChemMIP attribution experiments such as histSST-piCH4, will be used to quantify the effect of individual emissions and forcing changes on the historical ozone burden and budget. For future experiments, we focus on analogous experiments from the SSP3-70 scenario, a ‘regional rivalry’ shared socioeconomic pathway involving significant emissions changes.
The Role of Teleconnection Patterns in the Variability and Trends of Growing Season Indices Across Europe
Dr Philip Craig, Research Scientist, NCAS, University of Reading
Teleconnection patterns affect the weather and climate on both interannual and decadal timescales which in turn affects various socio-economic sectors such as agriculture. We use 3 climate indices (growing season onset (ogs10), growing season rainfall (gsr) and growing season temperature (ta_o)) to assess the interannual variability and trends over 1950-2017 associated with four teleconnection patterns (North Atlantic Oscillation (NAO), East Atlantic pattern (EA), Scandinavian pattern (SCA) and East Atlantic/West Russia pattern (EAWR)) using linear regression to extract the signal of each teleconnection pattern and their contribution to interannual variability. Trends towards an earlier growing season onset are found across most of Europe in low-lying regions. The NAO dominates interannual variability in North-West Europe with anomalies in excess of 10 days per NAO index and the EA dominates the continent with a trend towards the positive EA phase driving an earlier growing season onset of 1.1-1.7 days/decade in five regions. The EA and SCA gsr signals have north/south splits of orientation: positive EA is linked to increased gsr in northern regions and reduced gsr in Southern Europe, and vice versa for SCA. Correlations between gsr interannual variability and the teleconnection contributions are strongest in the Mediterranean regions and South Scandinavia with maxima of 0.41 and 0.46 respectively. Decreasing ta_o trends in Romania are explained by poor data coverage causing 3 high altitude stations to exert too much influence on the gridded data after 1961. The net effect is that Romanian ta_o is about 1.5°C cooler than expected compared to trends from surrounding countries. Improved spatial and temporal data coverage will benefit the EOBS dataset and prevent such erroneous trends.
The Impact of Land-Ocean Contrast of the Seasonal to Decadal Variability of the Northern Hemisphere Jet Stream
Samantha Hallam, Post-Doctoral Researcher, Irish Climate Analysis Research Unit, Maynooth University, Ireland
Seasonal to decadal variations in Northern Hemisphere jet stream latitude and speed over land (Eurasia, North America) and oceanic (North Atlantic, North Pacific) regions are presented for the period 1871 – 2011 from the Twentieth Century Reanalysis dataset.
Significant regional differences are seen on seasonal to decadal timescales. Seasonally the ocean acts to reduce the seasonal jet latitude range from 20° over Eurasia to 10° over the North Atlantic where the ocean meridional heat transport is greatest. The mean jet latitude range is at a minimum in winter (DJF) particularly along the western boundary of the North Pacific and North Atlantic, where the land-sea contrast and SST gradients are strongest. The 141-year trends in jet latitude and speed show differences on a regional basis. The North Atlantic has significant increasing jet latitude trends in all seasons, up to 3° in winter. Eurasia has significant increasing trends in winter and summer, however, no increase is seen across the North Pacific or North America. Jet speed shows significant increases evident in winter (up to 4.7ms-1), spring and autumn over the North Atlantic, Eurasia and North America however, over the North Pacific no increase is observed.
Long term trends are generally overlaid by multidecadal variability, particularly evident in the North Pacific, where 20-year variability in jet latitude and jet speed are seen, associated with the Pacific Decadal Oscillation which explains 50% of the winter variance in jet latitude since 1940.
The results highlight that northern hemisphere jet variability and trends differ on a regional basis (North Atlantic, North Pacific, Eurasia and North America) on seasonal to decadal timescales indicating that different mechanisms are influencing the jet latitude and speed. This is important from a climate modelling perspective and for climate predictions in the near and longer term.
Influence of Arctic Sea Ice Cover on Zonal Wind over Western Europe
Dr Peter Cook, Research Scientist, NCAS Climate, University of Reading
Recent years have seen increased windspeed and storms in the midlatitudes over Western Europe. Arctic sea ice cover has declined during this period and may influence the zonal wind by changing the meridional gradients in air temperature. Here 70 years of ERA5 reanalysis is used to determine the monthly mean values of zonal wind over the North Atlantic and Western Europe, and the area of Arctic sea ice around Greenland, Svalbard and the Barents Sea. Reduced sea ice is correlated to increased wind in the midlatitudes (around the British Isles) but reduced wind to the north and south (around Scandinavia and Spain), particularly in winter and spring, due to changes in the meridional air temperatures. However a cooling of sea surface temperature (SST) in the Atlantic sub-Arctic Gyre (south of Greenland) leads to a similar pattern of air temperatures and winds, and the wind fields, sea ice and SST all show a common 8-year cycle. Linear regression is used to determine the contributions of SST and sea ice cover to the influence on zonal wind, and these show different spatial patterns. The results imply that windspeeds and storms will increase over Western Europe in the long term as sea ice is reduced, while periods with cooler SST will further enhance the wind and storms.
Mechanisms of Internal Atlantic Multidecadal Variability at Two Different Model Resolutions
Michael Lai, PhD Student, University of Reading
The Atlantic Multidecadal Variability (AMV) is linked to climate impacts across the world and continues to modulate the changing climate. However, the processes behind the AMV are still not fully understood, due to a lack of observations and diversity in AMV simulated by models. Hence, a physical understanding of what drives the differences in simulated AMV is needed to improve decadal prediction.
This study broadly characterises and compares the key processes governing internal AMV in two resolutions of HadGEM3-GC3.1: N216ORCA025, corresponding to ~60km in the atmosphere and 0.25° in the ocean, and N96ORCA1 (~135km / 1°). Both models simulate AMV with a timescale of 70-100 years, which is related to low frequency ocean and atmosphere circulation changes. In both models, ocean advection dominates polar and subpolar AMV, whereas surface heat fluxes associated with cloud changes drive subtropical AMV. However, details of the ocean circulation changes and its impact on ocean heat transport differ between the models. The ocean circulation differences appear to be driven by differences in subsurface density propagation. The drivers of subsurface density anomalies also differ between models. In N216, the NAO is the dominant driver, while changes in Arctic-Atlantic exchanges of salinity-controlled density anomalies are the dominant driver in N96. These results further highlight that internal AMV mechanisms are model dependent and motivate further work to better understand and constrain the differences.
Decadal Variability of the East Asian Summer Jet and its Relationship with Sea Surface Temperature Variations
Dr Matt Patterson, Postdoctoral Research Assistant, University of Oxford
Variability of the East Asian summer jet stream (EAJ) has a significant impact on the climate of East Asia, primarily through its modulation of East Asian precipitation. For instance, a southward shifted jet in the period 1980-2000 was associated with flooding and drought over southern and northern China respectively.
In recent decades the impact of sea surface temperatures (SSTs) in the tropical Indian and Pacific oceans on the EAJ have been studied in considerable detail, however much less is known about the drivers of EAJ variability on decadal or multi-decadal timescales. Investigating this problem is made more challenging by the temporal limitations of reanalysis datasets.
In order to establish whether SSTs can provide a source of skill in predicting decadal variations of the EAJ, we analyse long pre-industrial control runs of the CMIP6 models as well as a large ensemble of atmosphere-only model runs. We find that variations in tropical Pacific SSTs are a strong driver of decadal EAJ variations, though the tropical Atlantic also plays a significant role in some models.
Equatorial Wave Activity and Teleconnections from ENSO in Observations and GloSea5
Gui-Ying Yang, Senior Research Scientist, NCAS Climate, University of Reading
Equatorially trapped waves are fundamental components of the tropical atmosphere; understanding equatorial wave activity is key to understanding tropical atmospheric variability. Evaluating model forecasts of equatorial wave activity is hence important for improving weather forecasting in the Tropics beyond a few days ahead; it is also likely to be crucial for climate prediction. For sub-seasonal to seasonal forecast, it is important to evaluate model ability to represent the seasonal cycle of equatorial wave activity and its relationship with ENSO. We investigate seasonal equatorial wave activity and its relationship with ENSO both in ERA5 and the Met Office Global Seasonal Forecast System version 5 (GloSea5). ERA5 is used as a benchmark to evaluate the ability of GloSea5 to predict equatorial wave activity. It is found that the seasonal cycle of Kelvin wave activity and its ENSO variations are not well captured by GloSea5 seasonal forecasts. On the other hand, GloSea5 performs better for the spatial pattern of seasonal cycle and ENSO variations for the three westward-moving equatorial waves: westward-moving Rossby-gravity wave and n=1 and 2 equatorial Rossby waves. Further evaluation of the ability of GloSea5 to predict tropical convection shows the poor performance for tropical eastward-moving convection which are often coupled with Kelvin waves, suggesting that the relatively poor GloSea5 performance for the Kelvin waves may be associated errors in simulating tropical eastward-moving convection.
Atmospheric Composition: Emissions and Measurements
Using Radon-222 to Determine 'Baseline' Concentrations of Trace Gases at the Weybourne Atmospheric Observatory, UK
Leigh S Fleming, PhD Student, University of East Anglia
In order to accurately assess long-term natural and anthropogenic emission influence and short-term pollution events on the atmosphere it is necessary to first identify baseline atmospheric conditions, i.e. air masses which contain constituent species at concentrations considered representative of background values due to having little influence from localised sources. However, it is difficult to separate baselines from localised sources using measurements of the gas species alone. In this study we have used a 3-year radon time-series from Weybourne Atmospheric Observatory to derive monthly baseline concentrations for multiple atmospheric gas species (CO2, O2, CH4, O3, N2O) using a two-step method. Firstly, selecting a radon threshold to identify the air masses with the least-terrestrial influence, and secondly, outlier removal. The resultant air masses were then evaluated using HYSPLIT and NAME back trajectories to identify the origin of the air masses and determine if there was any interaction with the land-surface. The resultant baselines are then evaluated and compared to those calculated using a statistical baseline (rfbaseline), and a meteorologically defined baseline.
Atmospheric Emissions from the UK Oil and Gas Industry
Shona Wilde, Postdoctoral Researcher, Wolfson Atmospheric Chemistry Laboratories, University of York
Accurate reporting of atmospheric emissions is at the core of assessing and mitigating environmental concerns resulting from industrial activities. Such reports are a crucial source of data which underpin the construction of national emissions inventories. Emission inventories for oil and gas (O&G) production are typically developed using "bottom-up" (BU) methods, which are based on scaled up component-level emissions.
Alternatively, top-down (TD) methods can be used, which involve measuring enhancements downwind of a point-source or region using aircraft, satellite or tower measurements to produce estimates which are representative of total atmospheric emissions. TD studies in regions of O&G production are commonly used to validate the BU estimates used in emission inventories. However, there are often discrepancies between results obtained by the two methods.
In the UK, emissions from upstream O&G sources are entirely operator-reported. Annual emissions estimates for atmospheric pollutants from all fixed and mobile installations are compiled within the Environmental Emissions Monitoring System (EEMS). The validation of such estimates is rarely attempted and the few studies that do are primarily concentrated around methane (CH4) with little focus on volatile organic compounds (VOCs). This work demonstrates the application of a TD methodology through which aircraft measurements are used to obtain emission rates of CH4 and speciated VOCs from O&G platforms in the North Sea. Airborne measurements were conducted on-board the Facility for Airborne Atmospheric Measurements’ (FAAM) BAe-146 research aircraft. Measurements of CH4 and ethane (C2H6) were made in-situ, whilst VOCs were measured using discrete whole air samples coupled with post-flight analysis by gas chromatography with flame ionization detection (GC-FID). Pollutant fluxes were calculated for each downwind plume using a mass balance approach.
Facility-level estimates of CH4 and total VOC ranged from 2.26 1.1 tonnes day-1 to 31.5 4.2 tonnes day-1 and 0.40 0.05 tonnes day-1 to 138.4 9.0 tonnes day-1, respectively, highlighting a widespread variability in the emission rates from individual platforms. Enhancement ratios of C2H6 to CH4 also displayed both spatial and temporal variability, facilitating the identification of unique emission sources within the offshore sector. Atypically high C2H6/CH4 ratios indicated the presence of an emission source rich in hydrocarbons relative to CH4, which was subsequently attributed to the venting of tank vapour during oil loading onto shuttle tankers. Comparison of the measurement-based emissions to the daily facility-reported estimates revealed a general underestimation in the BU reporting. The most significant disparity was observed for estimates concurrent with shuttle tanker operations, where the measured and reported values of total VOC differed by more than an order of magnitude, implying a significant gap in reported emissions from this source. Overall, this work potentially uncovers a significantly underestimated source of UK O&G emissions and demonstrates the need for further observational-based studies to understand the discrepancy between TD and BU approaches.
Polyisotopologue Ratios: A New Tool to Constrain Greenhouse Gas Budgets
Grant Forster, Research Fellow, NCAS, University of East Anglia
Through a recently funded NERC Highlight Topic project, POLYGRAM (POLYisotopologues of GReenhouse gases: Analysis and Modelling), we will use very high-resolution isotope ratio mass spectrometry (IR-MS) to explore the frontiers of isotopic research into carbon dioxide (CO2) and methane (CH4). POLYGRAM focusses on eight isotopic signatures, comprising both ‘traditional’ bulk stable isotopes ratios and polyisotopologue ratios for both CO2 (13C/12C, 18O/16O, 17O/16O and 13C18O16O/12C16O2) and CH4 (13C/12C, D/H, 13CH3D/12CH4 and 12CH2D2/12CH4). The improving ability to measure polyisotopologue ratios in atmospheric CO2 and CH4 is a promising new tool to further our understanding of how greenhouse gases are cycled in and out of the atmosphere, supplementing established work using the more common stable isotopes.
POLYGRAM will establish a small, but comprehensive, global atmospheric sampling network to examine latitudinal and longitudinal variations for these isotopic signatures, carry out field campaigns to determine the isotopic signatures of important CO2 and CH4 sources, and conduct laboratory experiments to estimate the reaction rates for CH4 isotopologues when they are oxidised and destroyed in the atmosphere. In addition, we will carry out atmospheric transport modelling for both gases to better interpret and understand our measurements. In this talk, we will give a brief overview of the POLYGRAM project and highlight the role that the UK’s Weybourne Atmospheric Observatory will play in establishing quasi-continuous in situ atmospheric measurements of the Δ17O signature of CO2, also known as the “triple oxygen isotope” or “17O excess”. This activity will provide us with a novel data set that will help us understand and develop the potential role that Δ17O measurements can play in quantifying and assessing the terrestrial processes driving CO2 fluxes.
Long-Term Trends in Tropospheric Composition at the Cape Verde Atmospheric Observatory
Matthew J. Rowlinson, NCAS Research Associate, NCAS, University of York
Long-term observations of climate composition are essential to detect and understand changes occurring in the Earth’s atmosphere. Since 2006 the Cape Verde Atmospheric Observatory (CVAO, 16° 52' N, 24° 52' W), a World Meteorological Organisation-Global Atmosphere Watch (WMO-GAW) global station in the Atlantic Ocean, has measured a wide range of trace gases and aerosols. The CVAO time series represents one of the few long-term records of composition of the sub-tropical marine boundary-layer. Analysis of trends between 2006 and 2020 shows that ozone concentrations have increased by 3.1 ppb over the entire period, although with an accelerated rate of increase of 0.7 ppb yr-1 since 2016. We present an analysis of long-term trends in surface concentrations and seasonality at Cape Verde. We utilise varies observations made at the site (CO, CH4, CO, VOCs etc), as well as model comparisons and sensitivity analysis using the GEOS-Chem chemistry transport model to understand the processes driving the trend.
Greenhouse Gas Flux Measurements onboard the FAAM Research Aircraft: Capability and Future Applications
Dr Stéphane Bauguitte, Chemistry Specialist, FAAM Airborne Laboratory, NCAS
Since 2019, FAAM has implemented 10 Hz in-situ airborne measurements for carbon dioxide and methane, with carbon monoxide to be added to this capability in 2021.
We present an update of FAAM's greenhouse gas measurements instrumentation performance, and a summary of related concentration and spatial resolution uncertainties. We discuss the potential future scientific opportunities presented by this development work.
We illustrate how FAAM fast chemistry observations during the recent MOYA campaign in Africa were used to derive eddy covariance emission fluxes of methane from wetlands, or to derive emission factors from biomass burning plumes.
The Rapid Rise of Methane: The Importance of Tropical Emissions
Euan G. Nisbet, Professor, Greenhouse Gas group, Dept. of Earth Sciences, Royal Holloway, University of London
Methane’s 2020 rise was about 14.7 ppb, according to the initial NOAA estimate. Growth may be revised later, depending on calculation of de-seasonalised global trends, but this is the strongest annual growth since detailed observation began in the early 1980s. Methane in the air increased rapidly in the twentieth century, then stabilised in the early years of this century but from 2007 began increasing again. In 2014, growth accelerated. Geographically, much of the post-2007 growth appears to have been led from the Tropics. The estimated 2020 rise is even greater than the previous record 14 ppb rise in 1991, an extraordinary year of climate and volcanic events, and fossil fuel industry chaos amid the end of the Soviet Union.
There has also been a major change in methane’s 13C/12C ratio. For two centuries prior to 2007, the 13C content of the methane burden had been increasing, consistent with emissions from fossil fuel industries and biomass burning. But since 2007, the most obvious explanation of the dramatic new growth is emissions from biological sources, which have lower 13C/12C ratios. These include both natural wetlands and anthropogenic sources including cattle, landfills and sewage.
Extremely rapid changes are taking place in the Tropics. Human populations are increasing rapidly, with intensification of crop and cattle farming, and growth of new waste-generating megacities. Although it is possible methane’s remarkable 2020 rise is due to unremarkable changes (more cows, more gas), it is also a troubling possibility that helping drive the rise is climate change feedback – warming feeding warming.
Our own current work in the NERC MOYA project has determined isotopic values in strong methane fluxes over tropical papyrus wetlands in the Nile, Congo and Zambezi basins, herbaceous wetlands in Bolivian southern Amazonia, and over fires in African woodland, cropland, and savannah grassland. Measured δ13CCH4 isotopic signatures were in the range -55 to -49‰ for emissions from equatorial Nile wetlands and agricultural areas, but widely -60 ± 1‰ from Upper Congo and Zambezi wetlands. These new isotopic values will help place isotopic constraints on global methane budget models.
The work highlights the need to understand tropical greenhouse gas emissions in order to meet the goals of the UNFCCC Paris Agreement. For CO2, the growth in the atmosphere is roughly in line with Paris expectations, but for CH4 the divergence is increasingly severe: methane needs attention.
Measurements of Air Quality
QUANT: A Traceable Assessment of Low-Cost Sensors in UK Environments
Sebastian Diez, Research Associate, Wolfson Atmospheric Chemistry Laboratories, University of York
Low-Cost Sensors (LCS) are creating a step-change in our understanding of air pollution, playing crucial roles in evaluating the health impacts of exposure (outdoors and indoors) and the effectiveness of solutions. Enabling these technologies in UK urban environments is thus a key priority for both the scientific and environmental policy communities. The potential of these devices is currently limited by the measurement paradigm in which they are being used, and new ways of deploying and analyzing the data are therefore needed. In this project, we evaluate the performance of a range of LCS technologies across multiple UK urban environments and develop methods that will enable their integration into the UK's air quality monitoring infrastructure and use in atmospheric chemistry research. For the deployment of the LCS devices, 3 field sites were chosen: Manchester, London, and York, all providing extensive reference measurements across a range of chemical environments representative of UK urban atmospheres. In this presentation we will show initial results from measurements made between December 2019 and August 2020, using both out-of-box and field calibrated data to characterize the uncertainties on real-world LCS measurements. This work will ultimately enable applications of these technologies to be optimized based on LCS data quality.
Measurement and Simulation Comparison for Aerosol Chambers
Dr Simon P. O'Meara, Postdoctoral Research Associate, NCAS, University of Manchester
A reasonable quantitative comparison between measurement and simulation is essential for effective interpretation of observations.
A measurement comparison tool has been added to the PyCHAM (CHemistry with Aerosol Microphysics in Python) software for aerosol chambers. PyCHAM already included a box model for aerosol chamber experiments (O’Meara et al. (2021)). The problem before this work was the arduous task of independently modifying model and/or instrument output to allow reasonable quantitative comparison.
PyCHAM has developed to include measurement characteristics that cannot be corrected for (e.g. particle size limits), which allows transformation of model output into the expected corresponding instrument output. To date the instruments provided for are the: condensation particle counter, differential mobility particle sizer and scanning mobility particle sizer. We are also working on methods for predicting instrument response functions such that comparisons between expected and measured e.g. mass spectral signatures can be made. This potential development not only bridges the gap between measurements and models but also allows us to evaluate the sensitivity of instrument response functions to changes in processes and properties.
The new functionality provides a quicker method for interpretation of aerosol chamber measurements. Whilst we focus here on presenting the new tool, we also demonstrate its use through analysis of a proposed new gas-phase chemistry mechanism.
The result is a software tool that allows measurement data to be readily compared with simulation results to test process understanding.
The project has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement no. 730997, which supports the EUROCHAMP2020 research programme. S.P. O’Meara received funding support from the Natural Environment Research Council through the National Centre for Atmospheric Science.
PM 10 and PM 2.5 Emission Factors for Non-Exhaust Particles from Road Vehicles: Dependence upon Vehicle Mass and Implications for Battery Electric Vehicles
David Beddows, Research Scientist, NCAS, University of Birmingham
Governments around the world are legislating to end the sale of conventionally fueled (gasoline and diesel) internal combustion engine vehicles (ICEV) and it is assumed that battery-electric vehicles (BEV) will take their place. It has been suggested - for example by Timmers and Achten (2016, 2018) - that due to their increased weight, non-exhaust emissions of particles from BEV may exceed all particle emissions, including exhaust, from an ICEV. To address this from a UK context, we have examined the vehicle weight dependence of PM10 and PM2.5 emissions from abrasion (brake, tyre and road surface wear) and road dust resuspension and made a comparison of the two vehicle types. We found that the outcome is critically dependent upon the extent of regenerative braking relative in place of friction brakes on the BEV, but overall the comparison shows only modest changes to the total local emissions of particles from a passenger car built to current emissions standards.
Changes in Ambient Air Quality and Atmospheric Composition and Reactivity in the South East of the UK as a Result of the COVID-19 Lockdown
Dr Kevin P. Wyche, Research Centre Director, Centre for Earth Observation Science, University of Brighton
The COVID-19 pandemic forced governments around the world to impose restrictions on daily life to prevent the spread of the virus. This resulted in unprecedented reductions in anthropogenic activity, and reduced emissions of certain air pollutants, namely oxides of nitrogen. The UK ‘lockdown’ was enforced on 23/03/2020, which led to restrictions on movement, social interaction, and ‘non-essential’ businesses and services. This study employed an ensemble of measurement and modelling techniques to investigate changes in air quality, atmospheric composition and boundary layer reactivity in the South East of the UK post-lockdown. The techniques employed included in-situ gas- and particle-phase monitoring within central and local authority air quality monitoring networks, remote sensing by long path Differential Optical Absorption Spectroscopy and Sentinel-5P’s TROPOMI, and detailed 0-D chemical box modelling. Findings showed that de-trended NO2 concentrations decreased by an average of 14-38% when compared to the mean of the same period over the preceding 5-years. We found that de-trended particulate matter concentrations had been influenced by interregional pollution episodes, and de-trended ozone concentrations had increased across most sites, by up to 15%, such that total Ox levels were roughly preserved. 0-D chemical box model simulations showed the observed increases in ozone concentrations during lockdown under the hydrocarbon-limited ozone production regime, where total NOx decreased proportionally greater than total non-methane hydrocarbons, which led to an increase in total hydroxyl, peroxy and organic peroxy radicals. These findings suggest a more complex scenario in terms of changes in air quality owing to the COVID-19 lockdown than originally reported and provide a window into the future to illustrate potential outcomes of policy interventions seeking large-scale NOx emissions reductions without due consideration of other reactive trace species.
Clouds and Moisture
Evaluating Memory Properties in Different Convective Parameterisation Schemes
Dr Chimene Laure Daleu, Postdoctoral Research Scientist, University of Reading
A series of high‐resolution three‐dimensional simulations of the diurnal cycle of deep convection over land are performed using the new Met Office NERC cloud‐resolving model. This study features scattered convection. A memory function is defined to identify the effects of previous convection in modifying current convection. It is based on the probability of finding rain at time t0 and at an earlier time t0−Δt compared to the expected probability given no memory. The memory is found to be strongest at grey‐zone scales of 4–10 km, there is a change of behaviour for spatial scales between 10 and 15 km, and it is reduced substantially for spatial scales larger than 25 km. At grey‐zone scales, there is a first phase of the memory function which represents the persistence of convection and it is maintained for about an hour. There is a second phase which represents the suppression of convection in regions which were raining 1 to 3 hr previously, and subsequently a third phase which represents a secondary enhancement of precipitation. We used the results from these high‐resolution simulations to evaluate parameterizations. More specifically, we assessed the ability of CoMorph (the new Met Office convective parameterization scheme which currently under development at the Met Office) and the 6A Mass Flux scheme (convection scheme currently used with the Met Office Unified Model) in capturing these three phases of the memory properties via their feedbacks onto the resolved state.
A Global Climatology of Lagrangian Coherent Structures Shaping the Atmospheric Moisture Distribution
Gabriel Perez, PhD Student, University of Reading
Large-scale mixing in the atmosphere redistributes moisture as organised bands or filaments. Studies have shown that large-scale filaments or bands of moisture and rainfall form along attracting Lagrangian Coherent Structure (LCSs) - material skeletons associated with strong attraction of air parcels. LCSs can be identified as ridges of the Finite-time Lyapunov Exponent (FTLE), a Lagrangian measure of deformation among neighbouring trajectories after a finite advection time scale. In this study, we compute the FTLE on 2-day back trajectories and identify LCSs using ECMWF's ERA5 reanalysis data between 1980 and 2009. The temporal averages of the FTLE and LCS occurrence show spatial features that broadly correspond to rainfall features such as the ITCZ and subtropical convergence zones. In the extratropics, regions of precipitation associated with fronts are co-located with strong FTLE features; these extratropical FTLE features are even stronger where fronts interface with subtropical anticyclones. This novel methodology offers a powerful process-based diagnostic to identify large-scale flow features with the potential to organise rainfall bands. This is the first long-term global climatology of FTLE and LCS occurrence in literature and it allowed us to derive interesting insights and links between circulation and rainfall. We intend to apply this process process-based diagnostics to evaluate the performance of circulation models in simulating such rain bands.
In-Situ Measurements of Cirrus Clouds on a Global Scale
Dr Gary Lloyd, Research Fellow, NCAS, University of Manchester
Observations of high-altitude cirrus clouds are reported from measurements made during the routine monitoring of cloud properties on commercial aircraft as part of the In-Service Aircraft for a Global Observing System. The increasing global scale of the measurements is revealed, with 7 years of in situ data producing a unique and rapidly growing dataset. We find that cloud fractions measured ≥ 10 km at aircraft cruise altitude are representative of seasonal trends associated with the mid-latitude jet stream in the Northern Hemisphere, and the relatively higher cloud fractions are found in tropical regions such as the Inter-Tropical Convergence Zone and South East Asia. Both stratospheric and tropospheric data were used to calculate the cloud fractions routinely experienced by commercial aircraft. Further work is needed for a direct comparison with previous studies that limit cloud fraction calculations to tropospheric data only. The characteristics of these clouds are discussed and the potential different formation mechanisms in different regions assessed.
Decadal Trends in Surface Solar Radiation and Cloud Cover Over the North Atlantic Sector during the Last Four Decades: Drivers and Physical Processes
Buwen Dong, Principal Research Fellow, NCAS and University of Reading
Analyses of time evolutions in downward surface solar radiation (SSR) and cloud cover, based on satellite-derived product and two reanalyses, show some consistent decadal trends over the North Atlantic sector, especially over North America and Europe during the last four decades. These trends show a strong seasonality with large changes in boreal summer, characterized by increases in SSR and decreases in cloud cover from 1980s to 2000s. A set of timeslice attribution experiments using an atmospheric general circulation model (AGCM) forced with prescribed changes in sea surface temperature/sea ice extent (SST/SIE), anthropogenic greenhouse gases (GHGs) concentrations and anthropogenic aerosol (AA) emissions all together, or separately, are performed to assess the roles of different forcings on these observed trends. Model responses to all forcing changes reproduce main observed features over Europe and North America, including the seasonality of trends, suggesting a dominant role of forcing changes on recent trends in SSR and cloud cover. Individual forcing experiments indicate that recent decadal trends in SSR over Europe are very likely and predominantly driven by AA emission changes with additional influence of SST/SIE and GHGs changes. In contrast, changes in AA, SST/SIE, and GHGs are all important for model simulated decadal trends in SSR and cloud cover over North America. Responses of SSR to AA emission changes are strongly modulated by aerosol-radiation interactions with a weak influence by cloud fraction through aerosol-cloud interactions and atmospheric feedbacks. Responses in SSR to SST/SIE and GHGs changes are dominantly due to cloud cover changes. This process level understanding of different forcing factors on decadal changes in SSR and cloud cover is of crucial importance for understanding and increasing the credibility of future projections and delivering more robust information for relevant climate impacts and adaptation studies.
Seasonal Impact-Based Mapping of Compound Hazards
Dr John Hillier, Reader in Natural Hazard Risk, Loughborough University
Impact-based, seasonal mapping of compound hazards is proposed. It is pragmatic, identifies phenomena to drive the research agenda, produces outputs relevant to stakeholders, and could be applied to many hazards globally. Illustratively, flooding and wind damage can co-occur, worsening their joint impact, yet where wet and windy seasons combine has not yet been systematically mapped. Here, seasonal impact-based proxies for wintertime flooding and extreme wind are used to map, at 1° × 1° resolution, the association between these hazards across Europe within 600 years as realized in seasonal hindcast data. Paired areas of enhanced-suppressed correlation are identified (Scotland, Norway), and are shown to be created by orographically-enhanced rainfall (or shelter) from prevailing westerly storms. As the hazard metrics used are calibrated to losses, the maps are indicative of the potential for damage.
Stratospheric Processes 1
Radiative Temperature Changes Associated with 2016 QBO Disruption
Alison Ming, Leverhulme Trust Early Career Fellow, University of Cambridge
A significant part of the ozone variability in the tropical stratosphere is driven by the Quasi-biennial Oscillation (QBO). The QBO is characterized by a descending pattern of alternative bands of eastward and westward winds and dominates the variability of the tropical stratosphere. During 2016, the regular pattern of winds was disrupted by the appearance of westward winds near 40 hPa. We use a radiative code to calculate the radiative impact of this disruption due to the changes in ozone. We find that that the ozone changes lead to large temperature changes (3 K at 70 hPa) and argue that being able to capture the details of the radiative and dynamical interactions during this disruption is an important test for models trying to generate a QBO.
The Main Findings of the Quasi-Biennial Oscillation Initiative
Dr Scott Osprey, NCAS Senior Researcher, NCAS and University of Oxford
The Quasi-biennial Oscillation Initiative (QBOi) was conceived to assess the state of tropical stratosphere variability in current-day general circulation models. To achieve this, the details of 5 experiments were agreed upon by international modelling groups with an active interest in assessing and improving model stratosphere representation. These experiments included present-day forcing scenarios, hindcasts of the recent past and idealised climate change. Here we summarise the principal results from a coordinated analysis of these phase 1 experiments.
Across the 17 global climate models used common biases were found including climatological westward/eastward wind biases in the tropical upper/lower stratosphere. Relatively weak QBO wind amplitudes were found in the low to middle stratosphere compared to reanalyses with the model QBO winds having too narrow meridional extent, especially in the lower stratosphere. Under idealised climate change scenarios, the QBO period changed but not consistently across the multi-model ensemble. However, a consistent weakening of the QBO amplitude in the lower stratosphere was seen across models in doubled and quadrupled CO2 scenarios. Momentum forcing within the descending wind shears came from a combination of both parameterised gravity waves and resolved waves. The latter had contributions from Kelvin and Rossby-gravity wave modes within the eastward and westward wind shears, respectively. Although sizeable resolved waves were diagnosed in the lower stratosphere, lower than expected forcing was diagnosed within the main shear zones, most probably through a combination of course vertical resolution, implicit and explicit numerical dissipation. Hindcasts initialised from early November showed a strong response of the boreal polar night jet for the first forecast month, but this response weakened considerably from mid-winter onwards. The extratropical surface response (e.g. NAO) was as strong as observed, but showed appreciable variation within the multi-model ensemble.
Following phase 1 of QBOi it is recommended future directions should include better understanding the origin of the QBO model biases and how they might impact the polar night jet and surface responses to the QBO.
Total Column Ozone Projections under Different SSP Scenarios
James Keeble, NCAS PDRA, NCAS, University of Cambridge
Stratospheric ozone is a key component of the Earth system, and past ozone depletion has had significant impacts on global and regional climate. Following the implementation of the Montreal Protocol and its subsequent amendments, stratospheric chlorine concentrations are now declining, and stratospheric ozone is projected to recover over the course of the 21st century. However, future stratospheric ozone projections are strongly influenced by emissions of other anthropogenic greenhouse gases, including CO2, CH4 and N2O. Here, we evaluate total column ozone projections in CMIP6 models under a range of future Shared Socioeconomic Pathways (SSPs). For the CMIP6 multi-model mean, global mean total column ozone is projected to return to 1960s values by the middle of the 21st century under the SSP2-4.5, SSP3-7.0, SSP4-3.4, SSP4-6.0, and SSP5-8.5 scenarios, and under the SSP3-7.0 and SSP5-8.5 scenarios total column ozone values are projected to be ∼10 DU higher than the 1960s values by 2100. However, under the SSP1-1.9 and SSP1-1.6 scenarios, total column ozone is not projected to return to the 1960s values despite reductions in halogenated ODSs due to decreases in tropospheric ozone mixing ratios. This raises issues with the use of return to historic total column ozone values as an adequate metric for measuring ozone recovery.
Dynamics of the 2009 Split Stratospheric Sudden Warming
Prof Lesley Gray, NCAS Senior Scientist, NCAS, University of Oxford
In a recent ensemble study of the January 2009 Sudden Stratospheric Warming (SSW) using the Met Office Unified Model, a series of experiments were performed in which the zonal winds were relaxed towards ERA Interim reanalyses in different parts of the atmosphere (Gray et al. 2020, Nat. Comms). Relaxing the tropospheric winds to observations was not sufficient to determine the SSW evolution and the stratosphere produced an incorrect displacement (wave-1) SSW in early December. However, if the equatorial winds above 5 hPa were additionally relaxed towards the reanalyses the correct timing of the SSW was achieved, the correct split type (wave-2) was achieved and the standard deviation of the 50-member ensemble was almost zero, indicating high statistical confidence. This study has been extended to investigate the relevant dynamics in much more detail, by examining the stationary and transient wave-forcing contributions and using refractive index and EP flux diagnostics. The analysis confirms that correcting the easterly bias of the semi-annual oscillation is crucial to the correct evolution of the SSW. Evidence is also shown for wave-wave interaction in early winter, as well as a resonant wave-2 cavity in the lead-up to the SSW.
How do Stratospheric Perturbations Influence North American Weather Regimes?
Simon Lee, PhD Student, University of Reading
Under different lower-stratospheric vortex states, there are observed changes to the occurrence and persistence frequency of four recurrent large-scale wintertime weather regimes over North America. However, it is not clear how this occurs or whether an improved stratospheric forecast would improve regime predictions. By considering both the location of the regimes in a 3D principal component space and the linear relationship between each principal component and the lower-stratospheric zonal-mean winds, we show which regime transitions are likely to occur due to stratospheric perturbations. We then test the effect of polar stratospheric relaxation on regime predictions in a set of OpenIFS simulations. While there is only a modest improvement to the regime predictions, the overall effect of the stratospheric relaxation is consistent with that expected from the linear relationship. Our results agree with the observational changes in regime occurrence and provide a framework for understanding the stratospheric influence on North American regime behaviour. The approach can be applied to understanding forecast uncertainty and errors, and in assessing model regime representation.
Atmospheric Patterns Classification and NWP Models Verification
Raffaele Salerno and Laura Bertolani, Senior Researcher, R&D Dept, Meteo Expert, Milan
Self Organizing Map (SOM) algorithm is used to study synoptic circulation over Southern Europe, evaluating the capability of NWP global models to predict its variability in order to relate synoptic circulation patterns to temperature and precipitation forecast quality over Italy. SOM is an iterative algorithm that ‘learns’ the patterns of the input data vectors and organizes them into nodes within the SOM space, arranging like patterns in neighboring nodes and the most unlike patterns in nodes farthest from each other. Daily observed and predicted weather types from five NWP global models, GFS from National Centers for Environmental Prediction, IFS from European Centre for Medium-Range Weather (ECMWF), Arpege from Meteo France, GEM from Canadian Meteorological Centre, ICON from Deutscher Wetterdienst, together with MIX, a multi-model ensemble, were recognized and classified by the SOM. This SOM-based classification built for our purposes produces a 12-weather-type set using daily 500 hPa and 700 hPa geopotential, sea level pressure, 850 hPa temperature and 700 hPa specific humidity.Here we present some examples of this operational activity, also showing how much the source of forecast errors may depend on large-scale dynamics rather than model's physical parameterisations. A quality index has been used to quantify the overall ability of models in predicting the circulation patterns, showing that MIX and ECMWF generally reached the best performance.
Defining Regimes of Cloud Controlling Factors using Machine Learning
Dr Alyson Douglas, Postdoctoral Research Assistant, University of Oxford
Cloud feedbacks, their controls, and their impact on the climate system remain extremely uncertain and unconstrained from observations. Exploiting over fifteen years of satellite observations, we train a simple machine learning model to understand how aerosols and the environment work to modulate cloud fraction and cloud optical depth, the two main controls of multiple cloud feedbacks. We use the learned relationships between the environment, aerosols, cloud optical depth, and cloud fraction to define regimes of cloud controlling factors. These regimes can be used to better understand the responses of cloud properties to the environment, including how clouds may respond in future climates. Our cloud controlling factors include the predominant cloud feedback constraints, such as sea surface temperatures and stability of the boundary layer. Our results show that our regimes have various signatures better understood when the relationships between the environment and cloud properties are used to separate distinct cloud regimes. Further, the signals are defined using the models of the cloud properties, a better step in isolating the cloud response from the environment as it is inherently controlled for by the models.
Stratospheric Ozone in Earth System Models: Future Modelling Strategies
Dr Beatriz Monge-Sanz, Research Scientist, NCAS and University of Oxford
The stratosphere is emerging as one of the keys to improve tropospheric weather and climate predictions. Including a realistic stratosphere within the next generation of Earth System Models (ESMs) will enable better exploitation of stratospheric sources of predictability for ESMs seamless applications.
The role the stratospheric ozone layer plays to improve weather predictions at different time scales has been shown by Monge-Sanz et al. (2021), highlighting the need for seamless models to include realistic prognostic ozone interactive with radiation. However, high resolution operational models cannot yet afford interactive full chemistry modules due to computational costs and processing times. It is therefore necessary to follow alternative strategies that enable seamless ESMs to include realistic interactions between stratospheric chemistry and physics at reduced computational costs.
In this study, we discuss the machine learning strategies we are using to create realistic models for stratospheric ozone suitable for seamless ESMs, and show preliminary results of their performance.
Applications of Agnostic Long Short-Term Memory (LSTM) Neural Networks in River Discharge Modelling
Kieran M. R. Hunt, Senior Research Scientist, NCAS, University of Reading
Physically-based river discharge models require large computational resources and catchment-scale hydrological data that can often be difficult to quantify. As such, LSTMs, a type of artificial neural network which can accurately simulate timeseries dependent on highly complex and non-linear processes, have become increasingly popular in streamflow modelling. However, choosing the correct model hyperparameters when setting up the LSTM remains challenging. In this paper, we show that this can mostly be overcome by using a large ensemble in which each member uses a different set of hyperparameters, increasing performance considerably over earlier methods. We apply this technique to four gauges in Pakistan, situated on the Jhelum, Chenab, Kabul, and Indus rivers, training the LSTM with catchment-averaged ERA5 surface fields. Kling-Gupta efficiencies for modelled discharge during the testing period (June 2019 to June 2020) were 0.916, 0.878, 0.950, and 0.953 respectively.
We then present a number of applications for the model. We show that the ensemble can be probed directly to show that model fidelity is most sensitive to the number of neurons in the first hidden layer and to the initial choice of starting weights. We show that the LSTM ensemble can be used to reconstruct historical discharge at the four stations beyond the range of publicly available observations, correctly simulating extreme flow during the 1988 and 2010 Pakistan floods. We demonstrate several methods by which sensitivity to input fields can be tested, showing that in general, modelled discharge is most sensitive to input runoff, soil moisture and seasonal mean flow, which have important implications for forecasting. Finally, we show how by adjusting the model, physical relationships between input variables and modelled streamflow can be deduced – looking at the relative contribution of snowmelt and rainfall to the seasonal cycle, then using an impulse response method, determine the shape of the storm hydrograph at each gauge.
Quantifying COVID-19 Related Fossil Fuel CO2 Reductions using Atmospheric Potential Oxygen (APO)
Penelope Pickers, Research Fellow, School of Environmental Sciences, University of East Anglia
The year 2020 saw significant reductions in fossil fuel CO2 (ffCO2) emissions caused by the COVID-19 pandemic and associated national lockdowns. Estimates of these reductions using “bottom-up” methods based on indirect activity data and emissions conversion factors find emissions reductions for the UK to be on the order of 17% during the first part of 2020 compared to emissions in 2019 (Le Quéré et al., 2020).
Attempts to detect and quantify COVID-19 associated emissions reductions using more direct “top down” methods (i.e. using atmospheric measurements) have largely been hampered by natural fluxes of CO2 between the atmosphere and the terrestrial biosphere and by atmospheric transport processes, both of which can mask ffCO2 signals. As stated by Le Quéré et al. (2020), “Despite the critical importance of CO2 emissions for understanding global climate change, systems are not in place to monitor global emissions in real time.”
Atmospheric Potential Oxygen (APO), calculated from the sum of atmospheric CO2 and O2 observations, is invariant to terrestrial biosphere exchange processes, making it an ideal tracer for ffCO2. We demonstrate the potential of APO as a continuous tracer for ffCO2 using data from the Weybourne Atmospheric Observatory on the Norfolk coast in the UK. By applying a machine learning algorithm to account for the impact of atmospheric transport processes, we can detect and quantify two distinct periods of ffCO2 reductions in the atmosphere associated with the two major COVID-19 waves: the first occurring in spring 2020, and the second beginning in Nov 2020. This APO-based assessment has the potential to provide high-resolution information (e.g. daily estimates) in near real-time, and is in good agreement with the spread of ffCO2 reduction estimates from three independent bottom-up UK emissions estimates.
A New Statistical Learning Approach to Constrain Global Cloud Feedback and its Impact on Climate Sensitivity
Dr Peer Nowack, Lecturer in Atmospheric Chemistry and Data Science, Climatic Research Unit, School of Environmental Sciences, University of East Anglia
Global warming drives changes in Earth’s cloud cover, which in turn have the potential to strongly amplify or dampen climate change. This ‘cloud feedback’ is the single most important cause of uncertainty in Equilibrium Climate Sensitivity (ECS) – the equilibrium global warming following a doubling of atmospheric carbon dioxide. Using data from Earth observations and climate model simulations, we here develop a novel statistical learning analysis of how clouds respond to changes in the environment. We show that global cloud feedback is dominated by the sensitivity of clouds to surface temperature and tropospheric stability. Considering changes in just these two factors, we are able to constrain cloud feedback to 0.43 W m-2 K-1 (5–95% confidence interval 0.08–0.78 W m-2 K-1), implying a robustly amplifying effect of clouds on global warming and only a 0.5% chance of ECS below 2 K. We thus anticipate that our new approach will enable tighter constraints on climate change projections, including its manifold socioeconomic and ecological impacts.