The Social Metwork in 2021

James Fallon – j.fallon@pgr.reading.ac.uk
Brian Lo – brian.lo@pgr.reading.ac.uk 

Hello readers, and a happy new year to you all! (ok we’re a bit late on that one… but regardless we both wish you a great start to 2022!)

We have really enjoyed reading and publishing all of the posts submitted over the last year, and following all of the research conducted by PhD students in the Reading department of Meteorology. There have been posts on conferences, papers, visitors, and much more, finishing with the heralded Met dept. Pantomime!

Thank you to everyone who has contributed to the Social Metwork over this past year. We are really looking forward to new submissions this term – and whether you are a veteran contributor or have never written for a blog before, we really encourage you to get in touch and write for us 🙂

In the last year, the blog has had over 7900 visitors from all over the world!

Map of visitors to the social metwork in 2021

Compared to 2020 we have seen Brazil and Australia enter the top 10 highest number of readers at 144 and 118 views respectively. And leading with the top 3 number of views are United Kingdom (6495), United States (1134) and China (361).

In case you missed any posts, or want to look through your favourites, here are all of the posts from 2021:

January
https://socialmetwork.blog/2021/01/15/why-renewables-are-difficult – Adriaan Hilbers

https://socialmetwork.blog/2021/01/22/forecasting-space-weather-using-similar-day-approach – Carl Haines

https://socialmetwork.blog/2021/01/29/air-sea-heat-fluxes-at-the-oceanic-mesoscale-the-impact-of-the-ratio-of-ocean-to-atmosphere-grid-resolution – Sophia Moreton

February
https://socialmetwork.blog/2021/02/05/main-challenges-for-extreme-heat-risk – Chloe Brimicombe

https://socialmetwork.blog/2021/02/12/the-role-of-climate-change-in-the-2003-european-and-2010-russian-heatwaves-using-nudged-storylines – Linda van Garderen

https://socialmetwork.blog/2021/02/19/weather-variability-and-its-energy-impacts – James Fallon & Brian Lo

https://socialmetwork.blog/2021/02/26/high-resolution-dispersion-modelling-in-the-convective-boundary-layer – Lewis Blunn

March

https://socialmetwork.blog/2021/03/05/do-local-or-non-local-sources-of-moisture-contribute-to-extratropical-cyclone-precipitation – Sophie Cuckow

https://socialmetwork.blog/2021/03/12/dialogue-concerning-the-obvious-and-obscurity-of-scientific-programming-in-python-a-lost-script – Brian Lo

https://socialmetwork.blog/2021/03/19/quantifying-arctic-storm-risk-in-a-changing-climate – Alec Vessey

April

https://socialmetwork.blog/2021/04/23/support-vector-machine-for-classification-of-space-weather – Carl Haines

https://socialmetwork.blog/2021/04/30/coding-lessons-for-the-newly-initiated – Daniel Ayers

May
https://socialmetwork.blog/2021/05/14/the-egu-experience-2021-a-phd-student-perspective – Max Coleman & Chloe Brimicombe

https://socialmetwork.blog/2021/05/21/better-data-information-with-metadata-application – James Fallon

June
https://socialmetwork.blog/2021/06/04/how-to-write-a-phd-thesis-during-a-global-pandemic – Kaja Milczewska

https://socialmetwork.blog/2021/06/18/cmip6-data-hackathon – Brian Lo & Chloe Brimicombe

July
https://socialmetwork.blog/2021/07/02/the-effect-of-surface-heat-fluxes-on-the-evolution-of-storms-in-the-north-atlantic-storm-track – Andrea Marcheggiani

https://socialmetwork.blog/2021/07/09/forecast-verification-summer-school – Lily Greig

https://socialmetwork.blog/2021/07/16/helicopter-underwater-escape-training-for-arctic-field-campaign – Hannah Croad

https://socialmetwork.blog/2021/07/23/connecting-global-to-local-hydrological-modelling-forecasting-virtual-workshop – Gwyneth Matthews & Helen Hooker

August
https://socialmetwork.blog/2021/08/13/diagnosing-solar-wind-forecast-errors – Harriet Turner

September
[ no submissions ]

October
https://socialmetwork.blog/2021/10/01/fluid-dynamics-summer-school – Charlie Suitters

https://socialmetwork.blog/2021/10/08/ensemble-data-assimilation-with-auto-correlated-model-error – Haonan Ren

https://socialmetwork.blog/2021/10/15/machine-learning-complement-or-replacement-of-numerical-weather-prediction – Emanuele Gentile

https://socialmetwork.blog/2021/10/22/ncas-climate-modelling-summer-school – Shammi Akhter

November
https://socialmetwork.blog/2021/11/05/climate-science-and-power – Gabriel M P Perez

https://socialmetwork.blog/2021/11/19/cop-climate-action-studio-2021-and-a-visit-to-the-green-zone-glasgow – Helen Hooker

https://socialmetwork.blog/2021/11/29/2021-academic-visiting-scientist-tim-woolings – Isabel Smith

December
https://socialmetwork.blog/2021/12/03/cloud-radiation-interactions-and-their-contributions-to-convective-self-aggregation – Kieran Pope

https://socialmetwork.blog/2021/12/14/panto-2021-hybrid-edition-semi-lagrangian-rhapsody – Natalie Ratcliffe

Panto 2021: Hybrid edition – Semi-Lagrangian Rhapsody! 

Charlie Suitters – c.c.suitters@pgr.reading.ac.uk
Hannah Croad – h.croad@pgr.reading.ac.uk
Isabel Smith – i.h.smith@pgr.reading.ac.uk
Natalie Ratcliffe – n.ratcliffe@pgr.reading.ac.uk

The pantomime has been one of the highlights of the year for the last 3 decades in the Met department. This is put on by the PhD students, and usually performed in person at the end of the Autumn term. Despite the ongoing COVID-19 pandemic, the panto is going from strength to strength, with a virtual instalment in 2020, and adapting to the hybrid format this year. It’s amazing to see the department tradition continue.  

This year the four of us (Charlie Suitters, Hannah Croad, Isabel Smith and Natalie Ratcliffe) agreed to organise the panto. It was clear that the panto this year would need to cater for both people joining in person and virtually, and with the lingering uncertainty of the covid situation in the UK, we came to a group decision to pre-record the performance in advance. This would provide the best viewing experience for everyone, and provided a contingency if the covid situation worsened. In hindsight, this was a good decision. 

This year’s panto was called Semi-Lagrangian Rhapsody, an idea based on the story of the band Queen. On Thursday 9th December 2021 we screened our pre-recorded pantomime in a hybrid format, with people watching both in the Madejski lecture theatre on campus and at home via Teams (probably in their pyjamas). Our story begins with our research group, Helen Dacre, Keith Shine, and Hilary Weller, on the lookout for a fourth member. In an episode of Mets Factor, the group sit through terrible auditions from Katrina and the Rossby Waves, Wet Wet Wet, the Weather Girls and Jedward (comprised of John Methven and Ed Hawkins), before finally stumbling upon Thorwald Stein (aka Eddy Mercury). The research group QUEEN (Quasi-Useful atmosphEric Electricity Nowcasting) is formed. Inspired by an impromptu radiosonde launch on the MSc field trip and skew-Ts (Chris knows!), QUEEN develop a Semi-Lagrangian convection scheme for lightning. Our narrator, SCENARIO administrator Wendy Neale, tells the story of the ups and downs of QUEENs journey, culminating in a presentation of their Semi-Lagrangian Rhapsody to the world at the AMS conference.  

Natalie suggested the idea for the panto, and we all agreed that it was a great idea – especially with the potential for lots of Queen songs! Once we had our storyline, next came the script writing. This was a daunting task, but working as a team we managed to produce a decent first draft in one intensive script-writing week, full of amazing terrible meteorology puns. Whilst writing the script we decided on the best Queen songs for the plot (and for reasons that we cannot explain/remember, a Rebecca Black song too). Now it was time to alter the lyrics, which was a lot of fun! Only once we had written the songs did we actually consider the complexity of Freddie Mercury’s voice and how we, a bunch of non-musically talented PhD students, were going to attempt to do these songs any justice. It was too late to go back though, and we had to break the news to the band. Thankfully they were up to the challenge! 

From week 6 onwards, we were able to start recording scenes; we were lucky that we were able to film in-person in and around the Met Department. We were still able to include students who weren’t in Reading at the time by writing in virtual parts into the panto. This worked perfectly well given the very hybrid nature of life currently anyway. 

Like last year, we wanted to start earlier as we knew that we needed to be finished at least a week – preferably more – before the big night to give time to edit everything in time (there were still a couple of late nights just before the big night). The final late night session did lead to the incredible slow-mo shot of Nicki Robinson (Charlie) turning around in Bohemian Rhapsody, so there is something that can be said about late-night-induced-insanity!  

Come week 10, we had nearly finished all of our filming and only had the songs left to record. We arrived at the London Road music rooms not yet having heard any of the band’s rehearsals. They sounded amazing. Many thanks to James and Gabriel who had been organising the band throughout the term. Then we started singing and immediately reduced the quality! But with a bit of practice around the piano, we started to improve, though the beginning of Bohemian Rhapsody was still a little questionable… With lots of pizza, we managed to record all of the songs in two nights! The band did an amazing job to put up with our musical incompetence (we are so very sorry). 

Over the next week, our three video editors worked hard to put the whole panto together and I hope you agree that they did a good job. This all led up to the big night where we were able to offer a small pre-panto reception in the Met coffee room before the panto started (somewhat attempting to mirror the normal pre-panto buffet). Apart from one slip up in scene 4 (my apologies hehe – Natalie), the screening went nearly perfectly with very few hybrid IT complications. Additionally, we had the return of an in-person performance of Mr Mets by our own Jon Shonk, and a heartwarming singing performance from the staff, organised by Chris Holloway and Keith Shine. Not only were we gifted this, but we were able to enjoy an in-person after-party in the coffee room with DJ Shonk. Of course there were a few Queen songs scattered in the mix, though we realised we struggled to remember the original lyrics and were only able to sing the panto versions! Following the story of Queen may have been a good idea, but have we forever ruined their songs for ourselves forever now? Quite possibly… 

And on that bombshell, we’d like to thank everyone who was involved in this panto, whether that be those who we convinced to act, sing, play in the band, help organise the event or even just come along to the screening. The whole process of creating this panto was exhausting, but so incredibly fun. I (Natalie) am so glad I did it and had a great time, but I now understand the ‘I’ve done my time’ sentiment of the previous organisers. (Hannah) Organising the panto was a lot of work, but so much fun (see bloopers). This has been a really rewarding experience, to see it all come together on the night, and to contribute to a fantastic department tradition. 

This year we sold tickets for the in-person showing and asked for donations to the David Grimes Trust from those viewing from home. Thank you to everyone who has already donated. Your generosity is greatly appreciated. We have managed to raise £170 for the David Grimes Trust. If you would like to donate still, please find our email with details on how to do so from Hannah Croad. 

Thank you to everyone who watched Semi-Lagrangian Rhapsody on Thursday, we hope you had a fun evening whether you watched at home or in-person! 

Cloud-Radiation Interactions and Their Contributions to Convective Self-Aggregation 

Kieran Pope – k.n.pope@pgr.reading.ac.uk

Convective self-aggregation is the process by which initially randomly scattered convection becomes spontaneously clustered in space despite uniform initial conditions. This process was first identified in numerical models, however it is relevant to real world convection (Holloway et al., 2017). Tropical weather is dominated by convection, and the degree of convective aggregation has important consequences for weather and climate. A more organised regime is associated with reduced cloudiness, increased longwave emission to space (Bretherton et al., 2005), and a higher frequency of long-lasting extreme precipitation events (Bao and Sherwood, 2019).

Because of its relevance to weather and climate, self-aggregation has been the focus of many recent studies. However, there is still much debate as to the processes that cause aggregation. There is great variability in the rate and degree of aggregation between models, and there remains uncertainty as to how aggregation is affected by climate change (Wing et al., 2020). Previous studies have shown that feedbacks between convection and shortwave & longwave radiation are key drivers and maintainers of aggregation (e.g. Wing & Cronin 2016), and that interactive radiation in models is essential for aggregation to occur (Muller & Bony 2015).

This blog summarises results from the first paper from my PhD (Pope et al., 2021), where we develop and use a framework to analyse how radiative interactions with different cloud types contribute to aggregation. We analyse self-aggregation within a set of three idealised simulations of the UK Met Office Unified Model (UM). The simulations are configured in radiative-convective equilibrium over three fixed sea surface temperatures (SSTs) of 295, 300 and 305 K. They are convection permitting models that are 432 × 6048 km2 in size with a 3 km horizontal grid spacing. The simulations neglect the earth’s rotation, so they approximately represent convection over tropical oceans within a warming climate.

Our analysis framework is based on that used in Wing and Emanuel (2014) which uses the variance of vertically-integrated frozen moist static energy (FMSE) as a measure of aggregation. FMSE is a measure of the total energy an air parcel has if all the water (vapour and frozen) was converted to liquid, neglecting its velocity. Variations in vertically-integrated FMSE come from perturbations in temperature and humidity. As aggregation increases, moist regions get moister and dry regions get drier, so the variance of vertically-integrated FMSE increases.

The problem with using FMSE variance as an aggregation metric is that it is highly sensitive to SST. A warmer atmosphere can hold more water vapour via the Clausius-Clapeyron relationship. This means there is a greater difference in FMSE between the moist and dry regions for higher-SST simulations, so the variance of FMSE is typically much greater for higher SSTs. To account for this problem, we normalise FMSE between hypothetical upper and lower limits which are functions of SST. This gives a value of normalised FMSE between 0 and 1.

Wing and Emanuel (2014) derive a budget equation for the rate of change of FMSE variance which shows how different processes contribute to aggregation. By rederiving their equation for normalised FMSE , we get:

\displaystyle \frac{1}{2}\frac{\partial\widehat{h'}_n^2}{\partial t} = \widehat{h'}_nLW'_n + \widehat{h'}_nSW'_n + \widehat{h'}_nSEF'_n - \widehat{h'}_n\nabla_h\cdot\widehat{\textbf{u}h_n}

where \widehat{h} is vertically-integrated FMSE, LW and SW are the net atmospheric column longwave and shortwave heating rates, SEF is the surface enthalpy flux, made up of the surface latent and sensible heat fluxes, and \nabla h \cdot \widehat{\textbf{u}h} is the horizontal divergence of the \widehat{h} flux. Primes (') indicate local anomalies from the instantaneous domain mean. The subscript (_n) denotes a normalised variable which is the original variable divided by the difference between the hypothetical upper and lower limits of \widehat{h}. The equation shows that the rate of change of \widehat{h'}_n variance (left hand side term) is driven by interactions between \widehat{h}_n anomalies and anomalies in normalised net longwave heating, shortwave heating, surface fluxes and advection.

We use the variance of \widehat{h}_n as our aggregation metric. Hovmöller plots of \widehat{h}_n are shown in Figure 1 for each of our SSTs. In these plots, \widehat{h}_n is averaged along the short axis of our domains. The plots show how initially randomly-distributed convection organises into bands which expand until the point where there are 4 to 5 quasi-stationary bands of moist convective regions separated by dry subsiding regions. This demonstrates that once our domains become fully-aggregated, the degree of aggregation appears similar. Figure 2a shows time series of each of the variance of \widehat{h}_n, and shows that the variance of non-normalised \widehat{h}_n is ~4 times greater for our 305 K simulations compared to our 295 K simulation. Figure 2b shows time series of the variance of \widehat{h}_n. From this, we can see the convection aggregates faster as SST increases, yet the degree of aggregation remains similar via this metric once the convection is fully aggregated. Values of \widehat{h}_n variance around 10-4 or lower correspond to randomly scattered convection, whereas values greater that 10-3 are associated with strongly aggregated convection.

Figure 3: Maps of (a) cloud condensed water path, (b) vertically-integrated FMSE anomaly, (c) longwave heating anomaly, (d) shortwave heating anomaly. Snapshots at day 100 of the 300 K simulation.

To understand the processes contributing to aggregation, we have to look to Equation 1. We mainly focus on the two radiative terms on the right hand side. The terms show that regions in which the radiative anomalies and the \widehat{h}_n anomalies have the same sign contribute to aggregation. We can start to get an intuitive understanding of this concept by looking at maps of these variables. Figure 3b-d show maps of \widehat{h'}_n, LW' and SW'. We can see SW' and \widehat{h'} are closely correlated since SW' is mainly determined by the shortwave absorption by water vapour. Clouds have little effect on the shortwave heating rates, with ~90% of the shortwave heating rate in cloudy regions being due to absorption by water vapour. LW' is closely linked to cloud condensed water path (Figure 3a). This is because the majority of our clouds are high-topped clouds which, due to their cold cloud tops, are able to prevent longwave radiation escaping to space, so they are associated with positive longwave heating anomalies.

The sensitivity of the budget terms to both aggregation and SST can be seen in Figure 4. This figure is made by creating 50 bins of \widehat{h}_n variance and then averaging the budget terms in space and time for each bin and for each SST. Where the terms are positive, they are helping to increase aggregation. Where they are negative, the terms are opposing aggregation. The terms tend to increase in magnitude since every term has \widehat{h'}_n as a factor, which increases with aggregation by definition.

Figure 4: Terms in Equation 1 vs normalised FMSE variance for each SST

In general, we find the longwave term is the dominant driver of aggregation, being insensitive to SST during the growth phase of aggregation. Once the aggregation is mature, the longwave term remains the dominant maintainer of aggregation, however its contribution to aggregation maintenance decreases with SST. The shortwave term is initially small at early times but becomes a key maintainer of aggregation within highly-aggregated environments. This is because humidity variations are initially small, so there is little variation in shortwave heating. Once the convection is aggregated, moist regions are very moist and dry regions are very dry, so there is a large difference in shortwave heating between moist and dry regions. The variations in shortwave heating remain very similar with SST, meaning shortwave heating anomalies contribute the same amount to non-normalised \widehat{h} variance. Therefore, shortwave heating contributes less to aggregation at higher SSTs because they contribute to a smaller fraction of \widehat{h} anomalies. The radiative terms are balanced by the surface flux term (negative because there is greater evaporation in dry regions) and the advection term (negative because circulations tend to smooth out \widehat{h'}_n gradients). The decrease in the magnitude of the radiative terms with SST is balanced by the surface flux and advection terms becoming more positive with SST.

To understand the behaviour of the longwave term, we define different cloud types based on the vertical profile of cloud, assigning one cloud type per grid box in a similar way to Hill et al. (2018). We define a lower and upper level pressure threshold, assigning cloud below the lower threshold to a “Low” category, cloud above the upper threshold to a “High” category, and cloud in between to a “Mid” category. If cloud occurs in more than one of these layers, then it is assigned to a combined category. In total, there are eight cloud types: Clear, Low, Mid, Mid & Low, High, High & Low, High & Mid, and Deep. We can then find each cloud type’s contribution to the longwave term by multiplying the cloud’s mean [Equation] covariance by its domain fraction.

To see how the cloud type contributions change with aggregation, we define a Growth phase and Mature phase of aggregation. The Growth phase has \widehat{h}_n variance between 3\times10^{-4} and 4\times10^{-4} and the Mature phase has \widehat{h} variance between 1.5\times10^{-3} and 2\times 10^{-3}. The contribution of longwave interactions with each cloud type to aggregation during these two phases is shown in Figure 5a, with their mean LW'\times\widehat{h'} covariance and fraction shown in Figures 5b & c.

Figure 5: Mean (a) contribution to the longwave term in Equation 1, (b) normalised longwave-FMSE covariance, (c) cloud fraction for the Growth phase (dots) and Mature phase (open circles). Data points for each category are in order of SST increasing to the right.

We find that longwave interactions with high-topped clouds and clear regions drive aggregation during the Growth phase (Figure 5a). This is because high clouds are abundant, have positive longwave heating anomalies and occur in moist, high \widehat{h} environments. The clear regions are the most abundant category, have typically negative longwave heating anomalies and tend to occur in low \widehat{h} regions, so their LW'\times\widehat{h'} covariance is positive. During the Growth phase, there is little SST sensitivity within each category. During the Mature phase, longwave interactions with high-topped cloud remain the main maintainer of aggregation however their contribution decreases with SST. This sensitivity is mainly because there is a greater decrease in high-topped cloud fraction with aggregation as SST increases. This also has consequences for the LW'\times\widehat{h'} covariance of the clear regions. As high-topped cloud fraction reduces, the domain-mean longwave cooling increases. This makes the radiative cooling of the clear regions less anomalous, resulting in an increasingly negative LW'\times\widehat{h'} covariance during the Mature phase as SST increases.

There is great variability in the degrees of aggregation within numerical models, which has important consequences for weather and climate modelling (Wing et al. 2020). With cloud-radiation interactions being crucial for aggregation, understanding how these interactions vary between models may help to explain the differences in aggregation. This study provides a framework by which a comparison of cloud-radiation interactions and their contributions to convective self-aggregation between models and SSTs can be achieved.

Page Break 

REFERENCES 

Bao, J., & Sherwood, S. C. (2019). The role of convective self-aggregation in extreme instantaneous versus daily precipitation. Journal of Advances in Modeling Earth Systems11(1), 19– 33. https://doi.org/10.1029/2018MS001503 

Bretherton, C. S., Blossey, P. N., & Khairoutdinov, M. (2005). An energy-balance analysis of deep convective self-aggregation above uniform SST. Journal of the Atmospheric Sciences62(12), 4273– 4292. https://doi.org/10.1175/JAS3614.1 

Hill, P. G., Allan, R. P., Chiu, J. C., Bodas-Salcedo, A., & Knippertz, P. (2018). Quantifying the contribution of different cloud types to the radiation budget in Southern West Africa. Journal of Climate31(13), 5273– 5291. https://doi.org/10.1175/JCLI-D-17-0586.1 

Holloway, C. E., Wing, A. A., Bony, S., Muller, C., Masunaga, H., L’Ecuyer, T. S., & Zuidema, P. (2017). Observing convective aggregation. Surveys in Geophysics38(6), 1199– 1236. https://doi.org/10.1007/s10712-017-9419-1 

Muller, C., & Bony, S. (2015). What favors convective aggregation and why? Geophysical Research Letters42(13), 5626– 5634. https://doi.org/10.1002/2015GL064260 

Pope, K. N., Holloway, C. E., Jones, T. R., & Stein, T. H. M. (2021). Cloud-radiation interactions and their contributions to convective self-aggregation. Journal of Advances in Modeling Earth Systems13, e2021MS002535. https://doi.org/10.1029/2021MS002535 

Wing, A. A., & Cronin, T. W. (2016). Self-aggregation of convection in long channel geometry. Quarterly Journal of the Royal Meteorological Society142(694), 1– 15. https://doi.org/10.1002/qj.2628 

Wing, A. A., & Emanuel, K. A. (2014). Physical mechanisms controlling self-aggregation of convection in idealized numerical modeling simulations. Journal of Advances in Modeling Earth Systems6(1), 59– 74. https://doi.org/10.1002/2013MS000269 

Wing, A. A., Stauffer, C. L., Becker, T., Reed, K. A., Ahn, M.-S., Arnold, N., & Silvers, L. (2020). Clouds and convective self-aggregation in a multi-model ensemble of radiative-convective equilibrium simulations. Journal of Advances in Modeling Earth Systems12(9), e2020MS0021380. https://doi.org/10.1029/2020MS0021380 

2021 Academic Visiting Scientist – Tim Woolings 

Isabel Smith – i.h.smith@pgr.reading.ac.uk

Every year, the Met PhD students at the University of Reading invite a scientist from a different university to learn from and talk to about their own project. This year we had the renowned Professor Tim Woolings, who currently researches and teaches at the University of Oxford. Tim’s interests generally revolve around large scale atmospheric dynamics and understanding the impacts of climate change on such features. We, as Met PhD students, were very excited and extremely thankful that Tim donated a week of his time (4th-8th of October) and travelled from Oxford for hybrid events within the Met. building. Tim told us of his own excitement to be back visiting Reading, after completing his PhD here, on isentropic modelling of the atmosphere, and staying on as a researcher and part of the department until 2013.  

The week started with Tim presenting “Jet Stream Trends” at the Dynamical Research Group, known as Hoskin’s Half Hour. A large number of PhD students, post-doctorates and supervisors attended, which was to be expected considering Tim has a book dedicated on Jet streams. After a quick turnaround, he spoke at the departmental lunch time seminar on “The role of Rossby waves in polar weather and climate”. Here, Tim did an initial review on Rossby wave theory and then talked about his current fascinating research on the relevance of them within the polar atmosphere. The rest of Tim’s Monday consisted of lunch at park house with Robert Lee and the organising committee, Charlie Suitters, Hannah Croad and Isabel Smith (within picture). Later that evening Tim visited the Three Tuns pub with other staff members, for an important staff meeting! The PhD networking social with Tim on Thursday was a great evening where 15 to20 students were able to discuss Tim’s research in a less formal setting within Park House pub.  

Tim Woolings (2nd left) and the visiting scientist organising committee

Tim’s Tuesday, Wednesday (morning) and Thursday consisted of virtual and in-person one on one 15-minute meetings with PhD students. Here students explained their research projects and Tim gave them a refreshing outsider perceptive. On Wednesday afternoon, after Tim attended the High-Resolution Climate Modelling research group, he talked about his career in PhD group (A research group for PhD students only, where PhD students present to each other.). Tim explained how his PhD did not work as well as he had initially hoped, and the entire room felt a great weight of relief. His advice on keeping calm and looking for the bigger picture was heard by us all.  

On Friday the 8th, a mini conference was put on and six students got to the “virtual” and literal stage and presented their current findings. Topics ranged from changes to Arctic cyclones, blocking, radar and Atmospheric dust. The conference and the week itself were both great successes, with PhD students leaving with inspiring questions to help aid their current studies. All at the University of Reading Department of Meteorology were extremely grateful and we thoroughly enjoyed having Tim here. We wish him all the best in his future endeavours and hope he comes back soon! 

COP Climate Action Studio 2021 and a visit to the Green Zone, Glasgow  

Helen Hooker h.hooker@pgr.reading.ac.uk 

Introduction 

SCENARIO DTP and the Walker Academy offered PhD students the opportunity to take part in the annual COP Climate Action Studio (COPCAS) 2021. COPCAS began with workshops on the background of COP, communication and interviewing skills and an understanding of the COP26 themes and the (massive!) schedule. James Fallon and Kerry Smith were ‘on the ground’ in the Blue Zone, Glasgow in week 1 of COP26, followed by Gwyn Matthews and Jo Herschan during week 2. Interviews were arranged between COP26 observers, and COPCAS participants back in Reading who were following COP26 events in small groups through livestream. Students summarised the varied and interesting findings by writing blog posts and engaging with social media.

Figure 1: COPCAS in action.   

Motivation, training and week 1 

Personally, I wanted to learn more about the COP process and to understand climate policy implementation and action (or lack thereof). I was also interested to learn more about anticipatory action and forecast based financing, which relate to my research. After spending 18 months working remotely in my kitchen, I wanted to meet other students and improve formulating and asking questions! I found the initial training reassuring in many ways, especially finding out that so many people have dedicated themselves to drive change and find solutions. During the first week of COP26 we heard about so many positive efforts to combat the climate crisis from personal actions to community schemes, and even country wide ambitious projects such as reforestation in Costa Rica. A momentum seemed to be building with pledges to stop deforestation and to reduce methane emissions.

Green Zone visit 

Figure 2: Green Zone visit included a weekend full of exhibitors, talks, films and panel discussions plus a giant inflatable extracting COvia bouncing!

During the middle weekend of COP26, some of us visited the Green Zone in Glasgow. This was a mini version of the Blue Zone open to the public and offered a wide variety of talks and panel discussions. Stand out moments for me: a photograph of indigenous children wearing bamboo raincoats, measuring the length of Judy Dench’s tree, the emotive youth speakers from Act4Food Act4Change and the climate research documentary Arctic Drift where hundreds of scientists onboard a ship carried out research whilst locked into the polar winter ice-flow.  

COPCAS Blog 

During COPCAS I wrote blogs about: a Green Zone event from Space4climate, an interview by Kerry Smith with SEAChange (a community-based project in Aberdeenshire aiming to decarbonise old stone buildings) and Sports for climate action. I also carried out an interview arranged by Jo with WWF on a food systems approach to tackling climate change.

Ultimately though, the elephant in the large COP26 Blue Zone room had been there all along…

Interview with Anne Olhoff, Emissions Gap Report (EGR) 2021 Chief scientific editor and Head of Strategy, Climate Planning and Policy, UNEP DTU Partnership.

Figure 3: Source: UNEP Emissions Gap Report 2021 updated midway through week two of COP26 accounting for new pledges. 

Time is running out, midway through the second week of COP26, the United Nations Environmental Partnership (UNEP) presented its assessment on the change to global temperature projections based on the updated pledges so far agreed in Glasgow.  

Pledges made prior to COP26 via Nationally Determined Contributions (NDCs) put the world on track to reach a temperature increase of 2.7C by the end of the century. To keep the Paris Agreement of keeping warming below 1.5C this century, global greenhouse gas emissions must be reduced by 55% in the next eight years. At this point in COP26, updated pledges now account for just an 8% reduction – this is 7 times too small to keep to 1.5C and 4 times too small to keep to 2C. Updated projections based on COP26 so far now estimate a temperature rise of 2.4C by 2100. Net-zero pledges could reduce this by a further 0.5C, however plans are sketchy and not included in NDCs. So far just five of the G20 countries are on a pathway to net-zero.

Anne’s response regarding policy implementation in law: 

“Countries pledge targets for example for 2030 under the UN framework for climate change and there’s no international law to enforce them, at least not yet. Some countries have put net-zero policies into law, which has a much bigger impact as the government can be held accountable for the implementation of their pledges.” 

Following my own shock at the size of the emissions gap, I asked Anne if she feels there has been any positive changes in recent years: 

“I do think we have seen a lot of change, actually…the thing is, things are not moving as fast as they should. We have seen change in terms of the commitment of countries and the policy development and development in new technology needed to achieve the goals, these are all positive developments and here now, changing the whole narrative, just 2 years ago no one would have thought we’d have 70 countries setting net-zero emission targets…we are also seeing greater divergence between countries, between those making the effort to assist the green transition such as the UK, EU and others, and those further behind the curve such as China, Brazil and India. It’s important to help these countries transition very soon, peaking emissions and rapidly declining after that.”   

I asked Anne how countries on track can support others: 

“A lot of the great things here (at COP) is to strengthen that international collaboration and sharing of experiences, it’s an important function of the COP meeting, but we need to have the political will and leadership in the countries to drive this forward.” 

Summary 

The momentum that was apparent during the first week of COP26 seemed to have stalled with this update. Despite the monumental effort of so many scientists, NGOs, individuals and those seeking solutions from every conceivable angle, the pledges made on fossil fuel reduction are still so far from what is needed. And at the final hour (plus a day), the ambition to ‘phaseout’ burning coal was changed to ‘phasedown’ and the financial contributions from developed nations pledged to cover loss and damage to countries not responsible for, but impacted now by climate change, have not been realised. I think this is the first time I have really felt the true meaning of ‘climate justice’. Perhaps we do need a planet law, as it seems our political leaders, do not have the will.

Overall, the COPCAS experience has been enjoyable, slightly overwhelming and emotional! It has been great to work together and to share the experiences of those in the Blue zone. It was also an amazing learning experience; I think I have barely touched the surface of the entire COP process and I would still like to understand more.

Climate Science and Power 

Gabriel M P Perez – g.martinspalmaperez@pgr.reading.ac.uk 

Introduction  

Climate science, especially climate-change science, is increasingly becoming a source of power in society and integrating politics. As an academic in meteorology, I started realising how possibly other scientists and I rarely think about how our profession fits in the power networks that constitute politics; on the contrary, it seems that we often think about our scientific outputs as something detached from the wheels of history.  

In this essay, I paint a picture of how climate science relates to the main sources of power in the civic sphere by building upon some of my recent readings in history of ideas and philosophy. I also discuss a few past and recent instances where alleged “apolitical” scientific discourse was moulded to support politics of domination and exclusion. By better understanding the relationships of power surrounding our science, we can be more confident that our scientific outputs will contribute more positively to society at large.  

The participation of climate-change science in politics is not exactly new: it has existed for at least three or four decades. However, up until the early 2010s, the hypothesis of anthropogenic global warming still faced a few challenges in the scientific realm. Perhaps the last of these challenges was the alleged 1998-2014 warming hiatus – climate scientists had to answer to the public and come to an agreement as to why the increase in global temperatures appeared to halt. Now, in the second decade of the 21st century, anthropogenic warming is a consensus and the most relevant contenders of climate science are found in the political realm. 

Historical background

In the seminal text “A discourse on the method of correctly conducting one’s reason and seeking truth in sciences”, Descartes proposes a method of pure reason to conduct scientific research. He proposes that the inquisitive individual should start by forgetting everything he learned and start deriving basic truths from simple logical statements and build up from that to provide scientific answers to more complex questions using “pure reason”. This text was one of the starting points of a scientific revolution and helped build the bases of modern science. Although epistemology quickly moved on from the Cartesian thought, it still affects the way we do science. After years in academia, trained scientists grow to believe that their scientific outputs are disconnected from the other spheres of society and the networks of power. This “apolitical” mindset will then affect, for example, our ideas, hypotheses and communications regarding climate change.  

The historian of ideas Michel Foucault in his book “The Order of Things”, deconstructs the idea that the scientific discourse is independent of the surrounding socio-economic environment. He argues that the scientific discourse is inherently tied to the “lenses” by which scientists of a certain time are capable of analysing the physical world. Foucault calls these lenses “epistemes”. The epistemes are the ways of thinking in each stage of history that define what is acceptable scientific discourse. Let us take Descartes’ Discourse as an example of that: the author lived in a highly religious time, and, although in Parts 1 to 3 of the Discourse he describes his method of “pure reason”, in Part 4 he employs his method to argue for the existence of God: this would not be acceptable scientific discourse in the 21st century. Therefore, even the brightest minds are subject to having their scientific discourse shaped by the epistemes of their time. 

For some scientists, accepting that our science is shaped by factors outside of the realm of pure reason may be uncomfortable. However, embracing our episteme and the historical forces that drive scientific paradigm shifts may aid us in producing and communicating science in ways that are more likely to impact society positively; this could also help prevent distortions and misuses of the power stemming from our science.

For example, the scientific consensus has been distorted in the past to provide an intellectual background to the darkest side of environmentalism: “ecofascism”, a political model that, in the early 20th century, used environmentalism to justify white supremacy and genocide of indigenous peoples (see the New Yorker article “Environmentalism’s Racist History” by Jedediah Purdy). Sadly, such distortions of environmental sciences are not buried in the past, on the opposite, they are gaining popularity in certain extremist groups (Lawton, 2019; Taylor, 2019) and even influencing today’s politics: the Portuguese ecofascist party was an important early supporter of the current Brazilian president Jair Bolsonaro, whose policies are ironically accelerating the deforestation of the Amazon rainforest (Pereira, 2020). In United States politics, we have recently seen in the media the eco-fascist “shaman” invading the White House after Donal Trump’s defeat. 

The power relations of climate science 

Power can be defined as the ability to have others do as you would. Evoking Foucault one more time, there are two kinds of power: the repressive and normalising power. Repressive power is a second-rate type of power that requires the use of force to control the actions of others.  Normalising power, on the other hand, is silent, non-aggressive, and much more effective than repressive power. Normalising power controls what other people want. If you have normalising power over others, they will do as you would because you have succeeded in making them want the same as you. As climate researchers, our scientific output is a growing source of normalising power. More and more people and governments want to do what climate science says is better for life on Earth. Therefore, “power” hereafter refers to “normalising power”. 

Power is present in all spheres of human relations (e.g., family, workplace and institutions). Here we will discuss power in the civic sphere, i.e., the power of having groups of people or societies do as you would. The civics educator Eric Liu suggests that power emanates from six sources. Here, I list four of these sources that I deem most relevant to climate science and discuss some ways that they relate to it: 

  1. Ideas Ideas, hypotheses and theories about the physics, impacts, mitigation and adaptation of climate change emanate almost exclusively from academia.  This directly places climate scientists from top institutions as raw sources of power that shape how people and governments think, behave and act towards climate change.  Combining David Hume’s proposition that ideas come from the impressions one has had throughout their lives and Foucault’s theory of epistemes, we may come to the conclusion that this source of power (i.e., scientific ideas) might not be as purely rational as one might have hoped. 
  2. Wealth Since ideas stem from academia, it is important to remember that most top institutions are in the wealthy nations of the Global North. Moreover, most scientists in these institutions were born and raised in the same wealthy nations. Naturally, the bulk of scientific outputs, both in terms of results and communications, are tied to the episteme, or “ways of thinking”, of this particular set of scientists.    Wealth is also related to science through the sources of research funding. Decisions regarding the allocation of research funds are often made by boards composed by either: 
  • Scientists in wealthy countries or 
  • Influential individuals outside academia 

A controversial example of B is the influence that Bill and Melinda Gates, through their foundation, exert over the World Health Organisation, having a disproportionate influence on scientific and public health decisions (Wadman, 2007). The issue is that Bill and Melinda’s suggestions are often not aligned with the public’s best interest or the scientific consensus, but rather with the personal motivations of these individuals. The power of wealth in climate-related negotiations is further evidenced when we notice that ideas such as climate debt (Warlenius, 2018) are typically ignored by current and former imperialist nations. These ideas were advocated by Global South agents in the widely ignored “World People’s Conference on Climate Change and the Rights of Mother Earth” held in Bolivia in 2010. 

  1. Numbers Climate activists, when numerous enough, have the power to pressure or convince governments and individuals to act according their beliefs. These beliefs are largely based on climate-change scientific literature. Scientists sometimes also take a more direct approach and practise environmental activism of some sort. 
  2. State action Governments are themselves a source of civic power but also subject to the other sources of power (i.e., ideas, wealth and numbers). Democratic states, as representatives of the people, have the power to directly and indirectly influence climate change by reducing (or increasing) emissions, funding climate research, educating the future generations, and many others. The governmental action, in its turn, is constrained by law in states under “Rechtsstaat” (or “the rule of law”). This raises the question: are lawmakers, prosecutors, judges and other agents well equipped to make decisions around climate change? In the next decades, it is not hard to imagine climate scientists being consulted regarding climate-change litigation in national or international courts. A few weeks ago, for example, the Brazilian president Jair Bolsonaro was accused of crimes against humanity at the International Criminal Court; his policies were said to be “directly connected to the negative impacts of climate change around the world”. 

Conclusion 

In this essay, I have outlined and attempted to disentangle a few existing and emerging power relations around climate change. I argue that as climate scientists we are sources of power in society. Therefore, we should be aware of our own “ways of thinking”, or “epistemes”, and remember that these are driven by external factors. Those external factors have the ability to shape our ideas, hypotheses and communications regarding climate change. Being aware of our role in the complex network of power known as politics could maximise the positive impact of the power stemming from our scientific outputs. Hopefully, this awareness could help prevent this power from being misdirected to support politics of domination and exclusion. Moreover, as the impacts of climate change are increasingly damaging to life on Earth, it is likely that in the next decade’s climate scientists involve themselves with litigation in national and international courts of justice. It is, therefore, timely for us to be aware of our roles in all levels of politics. 

References and further reading

Descartes, Rene. A discourse on the method of correctly conducting one’s reason and seeking truth in science. 1637

Foucault, Michel. The order of things. 1966.

Lawton, Graham. “The rise of real eco-fascism.” New Scientist 243.3243 (2019): 24.

Pereira, Eder Johnson de Area Leão, et al. “Brazilian policy and agribusiness damage the Amazon rainforest.” Land Use Policy 92 (2020): 104491.

Purdy, Jedediah. Environmentalism’s racist history. The New Yorker. 2015 https://www.newyorker.com/news/news-desk/environmentalisms-racist-history

Taylor, Blair. “Alt-right ecology: Ecofascism and far-right environmentalism in the United States.” The Far Right and the Environment. Routledge, 2019. 275-292.

Wadman, Meredith. “Biomedical philanthropy: state of the donation.” Nature 447.7142 (2007): 248-251.

Warlenius, Rikard. “Decolonizing the atmosphere: The climate justice movement on climate debt.” The Journal of Environment & Development 27.2 (2018): 131-155.

Eric Liu Ted Talk about civic power: https://www.youtube.com/watch?v=Cd0JH1AreDw

List of resources about climate debt: 

https://fascinated-soccer-ac0.notion.site/Climate-Debt-Resources-f7f24ce9a7344aa1a3e5290f853b6b4f

History of ecofascism: https://www.youtube.com/watch?v=FkhmP7yDWeY

NCAS Climate Modelling Summer School

Shammi Akhter – s.akhter@pgr.reading.ac.uk

The Virtual Climate Modelling Summer School covers the fundamental principles of climate modelling. The school is run for 2 weeks in the September of each year by the leading researchers from the National Centre for Atmospheric Science (NCAS) and from the Department of Meteorology at the University of Reading. I attended the school mainly because I have recently started using climate models in the second aspect of my research work and also because one of my supervisors recommended this to me.

What happened during the first week?

In the first week, lecturers introduced us to numerical methods used in climate models and we had a practical assignment implementing a chosen numerical method of our own in Python. We mostly worked individually on our projects that week. There were also lectures on convection parameterisation and statistical analysis for climate models.

What happened during the second week?

Figure 1: Earth energy budget comparison diagram between control (red) and flat earth (green) experiments produced by me in week 2.

In week two, with the assistance of NCAS and university scientists, we analysed climate model outputs. I personally was involved in the Flat Earth experiment- in which we tested the effect of changing surface elevation for terrain such as mountains, high plateaus on the climate. In this experiment, the perturbation is imposed by reducing the elevation of mountains to sea level. There were eight people in our team. As you may know, we PhD students have the occasional opportunity to do research collaboratively with other students in Reading Meteorology and encourage our teammates. For this reason, it felt very nice to me to work as the part of a research group. I was amazed by how we had been able to produce a small good piece of scientific work just within a matter of days due to our team effort. In Figure 1, I have presented a small part of our work which is the global energy budget comparison between a control experiment and the flat earth experiment (where the elevation of the mountains has been reduced to sea level). Along with our practical, we also attended some lectures on the ocean dynamics and physics, water in the climate system and land-atmosphere coupling and surface energy balance during this week.

How was it like to socialize with people virtually?

We used the Gather.town during the lunchtime and after work to socialize. I was a bit surprised though that I was the only student joining the Gather.town and as a result I always had to hang out (virtually) with NCAS and university scientists all the time. I rather consider it a blessing for me as there was no competition to introduce myself to the professionals. I even received a kind offer from one of our professors to assist him as a teaching assistant in his course in the department.

Concluding Remarks

I learnt about some of the basic concepts of climate modelling and I hope to use these things in my research someday. It was also very refreshing to talk to and work with other students as well as the scientists. While working in a group in week 2, I once again realized there are so many things we can accomplish if we work together and encourage each other.

Machine Learning: complement or replacement of Numerical Weather Prediction? 

Emanuele Silvio Gentile – e.gentile@pgr.reading.ac.uk

Figure 1 Replica of the first 1643 Torricelli barometer [1]

Humans have tried, for millennia, to predict the weather by finding physical relationships between observed weather events, a notable example being the descent in barometric pressure used as an indicator of an upcoming precipitation event. It should come as no surprise that one of the first weather measuring instrument to be invented was the barometer, by Torricelli (see in Fig. 1 a replica of the first Torricelli barometer), nearly concurrently with a reliable thermometer. Only two hundred years later, the development of the electric telegraph allowed for a nearly instant exchange of weather data, leading to the creation of the first synoptic weather maps in the US, followed by Europe. Synoptic maps allowed amateur and professional meteorologists to look at patterns between weather data in an unprecedented effective way for the time, allowing the American meteorologists Redfield and Epsy to resolve the dispute on which way the air flowed in a hurricane (anticlockwise in the Northern Hemisphere).

Figure 2 High Resolution NWP – model concept [2]

By the beginning of the 20th century many countries around the globe started to exchange data daily (thanks to the recently laid telegraphic cables) leading to the creation of global synoptic maps, with information in the upper atmosphere provided by radiosondes, aeroplanes, and in the 1930s radars. By then, weather forecasters had developed a large set of experimental and statistical rules on how to compute the changes to daily synoptic weather maps looking at patterns between historical sets of synoptic daily weather maps and recorded meteorological events, but often, prediction of events days in advance remained challenging.

In 1954, a powerful tool became available to humans to objectively compute changes on the synoptic map over time: Numerical Weather Prediction models. NWPs solve numerically the primitive equations, a set of nonlinear partial differential equations that approximate the global atmospheric flow, using as initial conditions a snapshot of the state of the atmosphere, termed analysis, provided by a variety of weather observations. The 1960s, marked by the launch of the first satellites, enabled 5-7 days global NWP forecasts to be performed. Thanks to the work of countless scientists over the past 40 years, global NWP models, running at a scale of about 10km, can now simulate skilfully and reliably synoptic-scale and meso-scale weather patterns, such as high-pressure systems and midlatitude cyclones, with up to 10 days of lead time [3].

The relatively recent adoption of limited-area convection-permitting models (Fig. 2) has made possible even the forecast of local details of weather events. For example, convection-permitting forecasts of midlatitude cyclones can accurately predict small-scale multiple slantwise circulations, the 3-D structure of convection lines, and the peak cyclone surface wind speed [4].

However, physical processes below convection permitting resolution, such as wind gusts, that present an environmental risk to lives and livelihoods, cannot be explicitly resolved, but can be derived from the prognostic fields such as wind speed and pressure. Alternative techniques, such as statistical modelling (Malone model), haven’t yet matched (and are nowhere near to) the power of numerical solvers of physical equations in simulating the dynamics of the atmosphere in the spatio-temporal dimension.

Figure 3 Error growth over time [5]

NWPs are not without flaws, as they are affected by numerical drawbacks: errors in the prognostic atmospheric fields build up through time, as shown in Fig. 3, reaching a comparable forecast error to that of a persisted forecast, i.e. at each time step the forecast is constant, and of a climatology-based forecast, i.e. mean based on historical series of observations/model outputs. Errors build up because NWPs iteratively solve the primitive equations approximating the atmospheric flow (either by finite differences or spectral methods). Sources of these errors are: too coarse model resolution (which leads to incorrect representation of topography), long integration time steps, and small-scale/sub-grid processes which are unresolved by the model physics approximations. Errors in parametrisations of small-scale physical processes grow over time, leading to significant deterioration of the forecast quality after 48h. Therefore, high-fidelity parametrisations of unresolved physical processes are critical for an accurate simulation of all types of weather events.

Figure 4 Met Office HPC [6]

Another limitation of NWPs is the difficulty in simulating the chaotic nature of weather, which leads to errors in model initial conditions and model physics approximations that grow exponentially over time. All these limitations, combined with instability of the atmosphere at the lower and upper bound, make the forecast of rapidly developing events such as flash floods particularly challenging to predict. A further weakness of NWP forecasts is that they rely on the use of an expensive High Parallel Computing (HPC) facility (Fig. 4), owned by a handful of industrialised nations, which run coarse scale global models and high-resolution convection-permitting forecasts on domains covering area of corresponding national interest. As a result, a high resolution prediction of weather hazards, and climatological analysis remains off-limits for the vast majority of developing and third-world countries, with detrimental effects not just on first line response to weather hazards, but also for the development of economic activities such agriculture, fishing, and renewable energies in a warming climate. In the last decade, the cloud computing technological revolution led to a tremendous increase in the availability and shareability of weather data sets, which transitioned from local storage and processing to network-based services managed by large cloud computing companies, such as Amazon, Microsoft or Google, through their distributed infrastructure.

Combined with the wide availability of their cloud computing facilities, the access to weather data has become more and more democratic and ubiquitous, and consistently less dependent on HPC facilities owned by National Agencies. This transformation is not without drawbacks in case these tech giants decide to close the taps of the flow of data. During a row with the Australian government, Facebook banned access to Australian news content in Feb 2021. Although by accident, also government-related agencies such as the Bureau of Meteorology were banned, leaving citizens with restricted access to important weather information until the pages were restored. It is hoped that with more companies providing distributed infrastructure, the accessibility to vital data for citizen security will become more resilient.

The exponential accessibility of weather data sets has stimulated the development and the application of novel machine learning algorithms. As a result, weather scientists worldwide can crunch increasingly effectively multi-dimensional weather data, ultimately providing a new powerful paradigm to understand and predict the atmospheric flow based on finding relationships between the available large-scale weather datasets.

Machine learning (ML) finds meaningful representations of the patterns between the data through a series of nonlinear transformations of the input data. ML pattern recognition is distinguished into two types: supervised and unsupervised learning.

Figure5 Feed-forward neural network [6]

Supervised Learning is concerned with predicting an output for a given input. It is based on learning the relationship between inputs and outputs, using training data consisting in example input/output pairs, being divided into regression or classification, depending on the type of the output variable to be predicted (discrete or continuous). Support Vector Machine (SVM) or Regression (SVR), Artificial Neural Network (ANN, with the feed-forward step shown in Fig. 5), and Convolutional Neural Network (CNN) are examples of supervised learning.

Unsupervised learning is the task of finding patterns within data without the presence of any ground truth or labelling of the data, with a common unsupervised learning task being clustering (group of data points that are close to one another, relative to data points outside the cluster). Examples of unsupervised learning are the K-means and K-Nearest Neighbour (KNN) algorithms [7].

So far, ML algorithms have been applied to four key problems in weather prediction:  

  1. Correction of systematic error in NWP outputs, which involves post-processing data to remove biases [8]
  1. Assessment of the predictability of NWP outputs, evaluating the uncertainty and confidence scores of ensemble forecasting [9]
  1. Extreme detection, involving prediction of severe weather such as hail, gust or cyclones [10]
  1. NWP parametrizations, replacing empirical models for radiative transfer or boundary-layer turbulence with ML techniques [11]

The first key problem, which concerns the correction of systematic error in NWPs, is the most popular area of application of ML methods in meteorology. In this field, wind speed and precipitation observational data are often used to perform an ML linear regression on the NWP data with the end goal of enhancing its accuracy and resolving local details of the weather which were unresolved by NWP forecasts. Although attractive for its simplicity and robustness, linear regression presents two problems: (1) least-square methods used to solve linear regression do not scale well with the size of datasets (since matrix inversion required by least square is increasingly expensive for increasing datasets size), (2) Many relationships between variables of interest are nonlinear. Instead, classification tree-based methods have proven very useful to model non-linear weather events, from thunderstorm and turbulence detection to extreme precipitation events, and the representation of the circular nature of the wind. In fact, compared to linear regression, random trees exhibit an easy scalability with large-size datasets which have several input variables. Besides preserving the scalability to large datasets of tree-based method, ML methods such as ANN and SVM/R provide also a more generic and more powerful alternative for nonlinear-processes modelling. These improvements have come at the cost of a difficult interpretation of the underlying physical concepts that the model can identify, which is critical given that scientists need to couple these ML models with physical-equations based NWP for variable interdependence. As a matter of fact, it has proven challenging to interpret the physical meaning of the weights and nonlinear activation functions that describe in the ANN model the data patterns and relationships found by the model [12].

The second key problem, represented by the interpretation of ensemble forecasts, is being addressed by ML unsupervised learning methods such as clustering, which represents the likelihood of a forecast aggregating ensemble members by similarity. Examples include grouping of daily weather phenomena into synoptic types, defining weather regimes from upper air flow patterns, and grouping members of forecast ensembles [13].

The third key problem, which concerns the prediction of weather extremes, corresponding to weather phenomena which are a hazard to lives and economic activities, ML based methods tend to underestimate these events. The problem here lies with imbalanced datasets, since extreme events represent only a very small fraction of the total events observed [14].

The fourth key problem to which ML is currently being applied, is parametrisation. Completely new stochastic ML approaches have been developed, and their effectiveness, along with their simplicity compared to traditional empirical models has highlighted promising future applications in (moist) convection [15]

Further applications of ML methods are currently limited by intrinsic problems affecting the ML methods in relation to the challenges posed by weather data sets. While the reduction of the dimensionality of the data by ML techniques has proven highly beneficial for image pattern recognition in the context of weather data, it leads to a marked simplification of the input weather data, since it constrains the input space to individual grid cells in space or time [16]. The recent expansion of ANN into deep learning has provided new methodologies that can address these issues. This has pushed further the capability of ML models within the weather forecast domain, with CNNs providing a methodology for extracting complex patterns from large, structured datasets have been proposed, an example being the CNN model developed by Yunjie Liu in 2016 [17] to classify atmospheric rivers from climate datasets (atmospheric rivers are an important physical process for prediction of extreme rainfall events).

Figure 7 Sample images of atmospheric rivers correctly classified (true positive) by the deep CNN model in [18]

At the same time, Recursive Neural Networks (RNN), developed for natural language processing, are improving nowcasting techniques exploiting their excellent ability to work with the temporal dimension of data frames. CNN and RNN have now been combined, as illustrated in Fig. 6, providing the first nowcasting method in the context of precipitation, using radar data frames as input [18].

Figure 6 Encoding-forecasting ConvLSTM network for precipitation nowcasting [18]

While these results show a promising application of ML models to a variety of weather prediction tasks which extend beyond the area of competence of traditional NWPs, such as analysis of ensemble clustering, bias correction, analysis of climate data sets and nowcasting, they also show that ML models are not ready to replace NWP to forecast synoptic-scale and mesoscale weather patterns.

As a matter of fact, NWPs have been developed and improved for over 60 years with the very purpose to simulate very accurately and reliably the wind, pressure, temperature and other relevant prognostic fields, so it would be unreasonable to expect ML models to outperform NWPs on such tasks.

It is also true that, as noted earlier, the amount of available data will only grow in the coming decades, so it will be critical as well as strategic to develop ML models capable to extract patterns and interpret the relationships within such data sets, complementing NWP capabilities. But how long before an ML model will be capable to replace an NWP by crunching the entire set of historical observations of the atmosphere, extracting the patterns and the spatial-temporal relationships between the variables, and then performing weather forecasts?

Acknowledgement: I would like to thank my colleagues and friends Brian Lo, James Fallon, and Gabriel M. P. Perez, for reading and providing feedback on this article.

References

  1. https://collection.sciencemuseumgroup.org.uk/objects/co54518/replica-of-torricellis-first-barometer-1643-barometer-replica 
  1. https://www.semanticscholar.org/paper/High-resolution-numerical-weather-prediction-(NWP)-Allan-Bryan/a40e0ebd388b915bdd357f398baa813b55cef727/figure/6 
  1. Buizza, R., Houtekamer, P., Pellerin, G., Toth, Z., Zhu, Y. and Wei, M. (2005) A comparison of the ECMWF, MSC, and NCEP global ensemble prediction systems. Mon Weather Rev, 133, 1076 – 1097 
  1. Lean, H. and Clark, P. (2003) The effects of changing resolution on mesocale modelling of line convection and slantwise circulations in FASTEX IOP16. Q J R Meteorol Soc, 129, 2255–2278 
  1. http://www.chanthaburi.buu.ac.th/~wirote/met/tropical/textbook_2nd_edition/navmenu.php_tab_10_page_4.3.5.htm 
  1. Bishop, C., and Christopher, M., Pattern Recognition and Machine Learning, Springer 
  1. https://www.arup.com/projects/met-office-high-performance-computer 
  1. J. L. Aznarte and N. Siebert, “Dynamic Line Rating Using Numerical Weather Predictions and Machine Learning: A Case Study,” in IEEE Transactions on Power Delivery, vol. 32, no. 1, pp. 335-343, Feb. 2017, doi: 10.1109/TPWRD.2016.2543818. 
  1. Foley, Aoife M et al. (2012). “Current methods and advances in forecasting of wind power generation”. In: Renewable Energy 37.1, pp. 1–8. 
  1. McGovern, Amy et al. (2017). “Using artificial intelligence to improve real-time decision making for high-impact weather”. In: Bulletin of the American Meteorological Society 98.10, pp. 2073–2090 
  1. O’Gorman, Paul A and John G Dwyer (2018). “Using machine learning to parameterize moist convection: Potential for modeling of climate, climate change and extreme events”. In: arXiv preprint arXiv:1806.11037 
  1. Moghim, Sanaz and Rafael L Bras (2017). “Bias correction of climate modeled temperature and precipitation using artificial neural networks”. In: Journal of Hydrometeorology 18.7, pp. 1867–1884.  
  1. Camargo S J, Robertson A W Gaffney S J Smyth P and M Ghil (2007). “Cluster analysis of typhoon tracks. Part I: General properties”. In: Journal of Climate 20.14, pp. 3635–3653. 
  1. Ahijevych, David et al. (2009). “Application of spatial verification methods to idealized and NWP-gridded precipitation forecasts”. In: Weather and Forecasting 24.6, pp. 1485–1497. 
  1. Berner, Judith et al. (2017). “Stochastic parameterization: Toward a new view of weather and climate models”. In: Bulletin of the American Meteorological Society 98.3, pp. 565–588. 
  1. Fan, Wei and Albert Bifet (2013). “Mining big data: current status, and forecast to the future”. In: ACM sIGKDD Explorations Newsletter 14.2, pp. 1–5 
  1. Liu, Yunjie et al. (2016). “Application of deep convolutional neural networks for detecting extreme weather in climate datasets”. In: arXiv preprint arXiv:1605.01156. 
  1. Xingjian, SHI et al. (2015). “Convolutional LSTM network: A machine learning approach for precipitation nowcasting”. In: Advances in neural information processing systems, pp. 802–810.