 What I will basically do now Apologies for my my voice that I partly lost I would like to provide you an overview of what the World Weather Research Program is doing as especially what are the main priorities because this is important to set up the framework where the sub-seasonal to seasonal project is is is considered and also because To provide you a view of how this strange machine is working because I mean sometimes we have a kind of Quite a bureaucratic interpretation of what these big organizations are are doing and while we should in some sense to be quite To be a little bit closer to what the citizens and the med services and in general the services They are doing and providing at the country at the country level So I hope I will I will so it will be quite a little bit more general Respect to what Andrew was saying that I would like just to provide you this reference then to maybe have a quick discussion Okay So the title is it's not my title is these interesting review up here in nature This year is the quiet revolution of numerical weather prediction I invite you to try to find it to download it is from Peter Bauer Alan Torpe and Gilbert Brunet and I think it's an interesting the review of of how the The weather prediction community is working and probably in the near future and just now we are More and more we are closer to the to the climate community in this seamless approach or try to to work much More together. Okay, so I think let's start with the some let's say maybe stupid questions So Should I bring an umbrella tomorrow is probably is a question that is well answered or sufficiently answered today Although yesterday I was flying from from Rome to here and the captain was saying oh probably there's the risk of fog Fortunately the forecast will be wrong So he had he had a completely different view of respect to my answer here. That's that was funny You have to remember that aviation aviation is the main driver for med services, especially in developing countries And most of the revenue of the med services in developing countries are coming from aviation. So This is that's I think the problem is first to move to another set of questions And this is and this is a strongly related to what Adrian shown Before so we are moving from weather forecast to impact weather forecasts So to try to provide much more information on the impact side not just on the rainfall Which is just I mean Influencing the land surface, but really going down and providing information on what are the impacts so one of the Interesting question was I mean the that's no other in one of the last meeting are are we able to predict? water quality Not just the rainfall, but the water quality in a complex urban network for instance in a developing countries in omega cities So this is a challenge. So here just to to some some key question that could be interesting So how to plan next 10 days traffic in Shanghai? one a tropical cyclone is approaching or What actions for a 23rd percent landfall probability of a typhoon by a three-week prediction so you have a prediction which Provide you three weeks to pre-prepared, but the probability is quite low So what is cost-effective? or What health protocol should be run to be prepared for the next send and last or which is one other strong Azard which is affecting a lot health In terms of not not only in Africa, but in all the Middle East and other in other countries in Asia as well So I think more or less what we are what the world weather research program should do Within the context of the world the meteorological organization need to shape a little bit. What is the met future? so what is the future of meteorological service because more and more we are Competing and we are you know in a in a world which is it's changing quite quickly. So we have private sector coming We have other services appearing, but the main question remains so What what is really necessary and how we can move or bridge to this future? so the war the meteorological organization especially for the world world weather research program is Defined four challenges. So in the next ten years, we will work around four challenges first of all the water of course the water availability, but the water forecast in many sense in a city for instance The water quality or the risk of a flash flood. So water in in is in The integrated water cycle I would say but certainly this is the first element then technology Of course technology matters a lot In this quite a revolution on America weather prediction you can imagine that the first element that matters is the is the availability of computational I mean the availability of big computers, but it's not the only technological element how to measure Either and show on the mobility map based probably on mobile phone network and Using mobile phone network actually you can measure rainfall You can infer more than measure, but you can infer rainfall maps and this is a way Technology is important for our future also because the cost and the maintenance cost of met many Observational networks is quite high. So there's another threat for the future. The other challenge is urbanization in In 2050 We will I mean 60 70 percent of the population will work in cities will sorry will live in cities So this is is a is a big issue. So may most of the services should be targeted to the urban areas and of course the third the fourth challenge is the The I impact whether or extremes and how we should tackle these extreme issue with a with a multidisciplinary prediction capability and I think the example of the malaria is one but There are other examples in the abstract or or in other sector like energy for instance how to manage a complex grid Electric grid in the future where now in Paris they are discussing to increase the solar Renewal energy etc. So you will have more and more a mixture of energy sources and which of course is quite complicated to to manage. So Let's stay with these four challenges and The main of course activities of the World Weather Research Program is first of all to increase the predictive scale So this is just one example What happened in the last let's say four or five decades We gained one day per decade. So if you You you you if you consider yourself as a as a civil protection manager, I mean 40 years ago You would that able to use weather forecast, but just for the third or the fourth day Because the skill for the fifth or sixth day was not enough good for you. Let's make this example So today you can use the seventh and the and the off and the eighth day of the forecast So we we are gaining one in terms of skill We are gaining one day per decade more or less in the last 30 40 years Of course, this is is not a trend. We should we have some predictability limit, etc. But still it's a tangible I mean success of the of the big weather Research of the weather research community and the way we are providing services. So this is just an example of predictability. This is a Six day forecast So the contour lines are the The black line are the 500 ectopascal geopotential high while the dash the dashed line is the is the analysis of For the six day forecast corresponding to the sixth day forecast and the difference you can see is the Actually is the error So you can see that there's this kind of pattern and especially a big error over Europe But you can track. I mean some sort of wave coming from the Pacific up to the cross crossing the US and and arriving over Europe and actually If you try to make some numerical exercise Exercises and you you can actually determine determine that the the error source is here, especially over In the upper troposphere where this big Error in the measurements for the wind for the upper level wind. So if you are increasing the I mean the observation here you get this second forecast where you are Dramatically reducing the the error actually so just to explain to Simplify that one of the main tasks of the World Weather Research Program is try to improve The the predictive is killed and especially for several times scale ranging from one day to one month This is our objectives, of course and S to S is one of of the project, but There are several ways you would like to improve the you can try to improve this This forecast so in the last 20 years we developed several techniques in terms of data simulation capacity to increase the initial the the to increase the number of observations that are used in the model to To provide a forecast in terms of ensemble forecasting so not just a deterministic forecast This has been something I mean coming from the last 20 years of Research activities, so we are more and more working in terms of ensemble forecasting Which is actually is quite interesting because it's not only providing you a better Capacity to detect for instance extremes although with the with the low probability maybe but it is also Actually better linking you with the other decision-making and other Sectors because you are sharing the responsibilities So if you are just providing a forecast for let's say the next five days saying that in Trieste there will be just 30 20 millimeter per day in six days actually you are providing a deterministic forecast So anyone else will use this number while if you are telling you are providing a forecast in terms of probability So what is the probability to exceed some threshold in in in ten days actually the The people or the manager or other sector they should interpret this probability and they should use this So in some some way you are sharing the responsibility, which is an interesting element in the last 10 15 years of the weather over the weather forecast Community and that the way we are increasing the interaction with other sectors also So of course you have earth observation I would remember that now we are using we are receiving around seven millions of data per day from the satellites, but we are using just 20 percent of this data and Of course we we are more and more improving the the use of this data Why of course because each kind of each sensor needs to be Elaborate in order to trans translate this information into initial condition for the moment model, which is not a direct Let's say process and they need it needs a lot of work in terms of data simulation of course complexity So we are moving towards a Couple system for weather forecasting I will provide you some example especially for the for the polar prediction, but subsizonal to season is one example we are of course using most most part of the models providing the Monthly times monthly forecast to the subsizonal to seasonal database. They are coupled models. So also for these Weather forecast system. We are using weather. I mean couple system. Then of course we are moving to to another Evolution, which is the resolution? Now I think the European Center is now in January is The the resolution will be around 12 15 kilometers and At regional or national scales in most part of the world we have one kilometer scale forecast prediction and We are moving especially for special for the aviation down to 500 meters or something so this is really an Area where the the the step forward and we there's been a really increasing Work in the in the last 10 10 years Okay, the other element is of course bridging with technology. I already mentioned this This element which is the the mobile phone network providing you because of course you have power Depending on the rainfall rates the system is the network the mobile network is setting different power for transmitting power So you can infer the the rainfall rates based on this information Except if the mobile Phone company will provide you with the data because there are tricky I mean the element behind if because there are some threshold defined by law and in some cases they are Actually, not they are going They are they are they are They they is not expected that they they should go hop and to to overcome this threshold So this is one one issue and of course technology means also computing power This is just a funny example, but this is the first any a computer used for the first weather forecast 5th January of 1949 it's funny the first the first 20 or do you know how long it took to make a 24-hour forecast? That's funny because it's called the first forecast it took 24 hours, so it's it's And this was 30 tons and your eye watch could be order of 30 grams or something So it's interesting measuring the technological gap in terms of weight from Yeah, and But this is more or less that the same power of the same computational power We had we did with the anyak at the beginning after the Second World War So there's a strong and now there's a there's a bottleneck of course because While 10 years ago the big companies producing cheap producing computers They were first referring to weather climate sector now. They are referring to your Maybe your young younger brothers Playing with the PlayStation. So they are investing much more on games than on other sectors and so the different structure of the cheap of the technology of course affects how we can use our codes and our models and This is of course needs a lot of I mean there's a lot of work behind I think you see and WF is is working a lot in this in this along this direction and Of course is more and more a multidisciplinary prediction. So We are less and less stopping to just focusing on our info and more and more focusing on the impact-based variables parameters to be provided To to to in terms of services. So this is one of the has been probably quite It's been discussed quite quite a while during this this week. So of course I Would like just to to show you what is the structure of the how our world Research program is structured to to tackle these issues and to try to organize of work it's I Think it could be interesting to know how this this machine is working Of course, I've I told you that we have these four challenges the water cycle urbanization emerging technology and extremes and We have core projects Which are the subsizonal to seasonal the polar prediction and the high impact weather usually this Those projects are ten years long the previous one was Torpex which was Focusing I mean to improve predictability and now we have these three project more or less ten ten year long and They are our core projects. So and subsizonal to seasonal is one of them is not They are not the only project we have. So we have also another bunch of projects Which of course we are not defining as core project, but we have several demonstration project One is on aviation for instance, which is quite important and several on tropical cyclones Demonstration means that they are trying to bridge between the research and the operational and the operational services and of course we have pre operational project and one is sand and dust storm project, which is providing Real-time weather Sand and dust Forecast which is quite important element for the health sector as well and We have several working groups expert working with us and try to provide guidelines and try to provide a Framework to improve the predictability to improve the Multidisciplinary approach, etc. And this is the least of the of the working groups the the numerical Experimentation which is which is known as weakening. I'm trying to to have a kind of a free acronyms presentation Which is the opposite of what happened what usually happens in in WMO where you have I mean the first two months in WMO I spent my time just to to try to understand acronyms. That was a funny funny period That's it's a good good question So we have now casting them as a scale working group of course working on the short-times short-time scale We have tropical meteorology, which is really important because of course WMO is a united nation organization as ICTP under UNESCO so the first Question is I mean the first Important question is how to to help or to support who needs who need most and of course tropical areas you have most of the developing countries and Predictability and dynamic and assemble forecasting. We're trying to better Understand the predictability issues data simulation and observe a system Which is really critical because is one of the element that is Actually linking our activities also with the climate community verification working group. That is another important An important Working group social and economic application and weather modifications Okay, just I think I can try to To go quickly along the three main the three core projects So the sub-seasonal to seasonal you I think you know everything But I would like just to to remember you what Behind there there's certainly the fact that on these monthly timescale most of the decision Or most of the decision-makers they are really interested to this time scale because in terms of energy agriculture health sector several decisions are based on this information if available, of course and Just an example the this year drought in in in California, which is continuing actually and and affected the both the agriculture sector and the energy sector and This is one key example and I think One of the ideas Andrew mentioned to to really reinforce The activity of the sub-seasonal to seasonal project is to find partnership with other sectors to try to provide these exemplars to provide potential I mean example and the one Adrienne today about Uganda and about that the sector could be an interesting Example to to to be explored and I think we have a good connection with WHO and we other UN agencies working on the migration So that's it's could be an interesting area, but this is One of our projects the second one is the polar prediction You know there's a strong interest along the polar area a special the Arctic in the last why Because of course Me in many cases the driver is not the climate driver, but is an economic driver as well. So we We should be aware. I mean we are living in this world is complex word so one of the question is if the decaying eyes cover in the last 15 years will allow shipping routes Across the Arctic of course most of the Asian countries Will prefer to go through the Arctic? Shipping routes. So there's a strong interest from China, South Korea Singapore and other other countries there, but and of course there is a strong interest from the countries Sitting in into the Arctic so Russia Canada etc so of Course there's a strong interest in terms of science why because the coverage both for the Arctic and the Antarctica in terms of The observing system is quite low. So we actually we are not Observing the net the observational network is not up-to-date Respect to what we we need and this is important not only for the weather community for the climate community as well because Do you know you have several processes running at the very small scale. So one of the last I mean finding in terms of climate research for the Arctic is you have this In the in the ice sheet you have the formation of these small holes around the the Arctic and actually the amount of Shortwave radiation is penetrating down into the ocean Is that never never been measured up to the last two years and now you you discovered that? the shortwave Amount of energy that is penetrating down in the ocean is really higher than you you expected and this is of course will could accelerate the Melting phase or in the next ten years, but this is really related to the capacity to observe observe the surface but observe also the the What is happening? Beneath the ice the ice cover so there is a gap in terms of observation We starting to have a couple system or at least we have a couple system in terms of ocean and ice models This is a Canadian operational system, so it's running and providing 10 to 20 to 30 days of of Forecast for the for the area And this is just the sea surface temperature if I remember well, but is a couple system so it's providing information on the ice extent and in terms of verification, it's interesting because What what has been done here is to ask the the shipping routing industry what would have been the best measure of to verify The the extent of the ice sheet and of course the The answer was quite simple So the best way is to measure the distance between the coast and the ice and this has been I mean set up as the new a new parameter to be to be verified by environmental Canada Which is the Met Services in Canada? Just to explain you how the verification also is related to the users Not just to what you believe to be the most important parameter for your dynamical system so this is another interesting question and There's potential skill. Yes. This is the persistence based in terms of root me square error And this is basically a forecast for the ice extent that is based on persistence and This is just based on the system I've shown you before and the root me square error is much lower for a specific season in using the the couple system ocean and NC ice But why should we invest a lot of money to have networks there where? Few guys are living there. So what? So this is I don't know how many from you are are from from the Nordic regions But probably no one here is coming from from the Arctic or the Antarctica But the answer is good. Yes This is for a specific season I would say Yeah I But this is a weekly forecast, so I think it's a for a specific period So it's not for average Over all the all the all the year So I think it's you are you are filtering out your your system Sorry, so here you have an experiment so done For a 30-day forecast which is the the sub seasonal timescale we are looking here and This is the difference in the is a kind of indicator of potential predictability Which is quite interesting because is if you run your prediction system without any I mean The state of the art prediction system and then you run the same system just Of course, this is an I'm cast experiment so if you run for the past and you And you relax your your system to the analysis over this part of the Arctic What you get of course you get a a Trivial answer that you are increasing the predictability where you are relaxing your system. That's fine Okay, but what is interesting that you are increasing predictability far away and this is a little bit sometimes we should explain in terms of of the link with the with with the jet or with The shape or the phase of the planetary waves But certainly it's interesting that for day 6 10 and on for the 11th or 30 you have some Signature of potential predictability, of course The indicator is not so high But it's interesting to say that if you are increasing the number of observation of the Arctic This is just to simplify these results. Actually you are getting Some positive results far away from from the Arctic This is just to show you the the working group working on the polar prediction And this is the most important thing since we this field campaign the year of polar prediction that will be between 2017 and 2019 but centred on 2018 this will be a quite big effort Concerning also Arctic and Antarctica and this is one of the one of the I Mean of the challenge of the next two three years for the polar prediction Let's few words on the third project which is related in terms of in terms of challenges to the Subseasonal to seasonal although the time scale is is shorter is the high impact weather project Who is co-leaded by Brian Golding met office and David Johnstone from from New Zealand and this project of course is mainly focusing on on cities or urban regions, let's say Here is an example of risk associated to natural hazards For cities and you can see that this is quite democratic is covering It's not just for is for developing and developed countries. So there's in terms of a thread It's it's of course covering all all the world. So of course cities matter and this is the This is another indicator, which is quite important is coming from WTO is outdoor air pollution Is the risk associated to air pollution because this project is trying to bridge between the classical meteorological information and other impact based information and one is is for air quality as well so These are the main element of the the main topic of the project the main topics so urban flood this rapidly winter weather Wildfire your many ways and their pollution and extreme local wind. So the project will try to cover all these activities, but the way is doing is the way is is run is Is focusing on not only on the geophysical side of the how to improve our forecast How to provide better information for this specific sector, but how this information is actually used in the decision-making chain So how the weather information the weather Forecasts it transformed or translated into weather information. So this project is a multidisciplinary So social sociologists are part of the project and we are trying to understand for these specific sectors and area how this information is used and Actually, how this information can be better used. I told you at the beginning of the talk about this Sharing responsibility when the first time you came it off is moved from a determinacy forecast to assemble forecasting the civil protection the UK civil protection was quite Upset because of course receiving a millimeter per day was just saying okay I can issue an early warning because the UK met office said that There would be 100 millimeter per day in three days. Okay, it's fine. Well, if I have to interprets This date I have to share the responsibility So I have to take a decision based on my interpretation and of course this is is changing a little bit So here we are trying to better understand this process and to try to see and this is just a decision-making I mean how for a specific hazard Which is a Related to flesh flood and especially for river management and coastal floods. What are all the step in terms of decision-making and in terms of actions based on the on the lead time as well So just to explain you what what was the and of course at the end there is the decision I Take by individuals this is quite so I think that I will stop here I would just Just to let you know something what of course this is funny title that what successful people read before bed. So this is You can Google a seamless prediction WMO and you will win this book you can download is a PDF This book it comes out from the world weather open-size conference last year in more real and this is kind of Summary of the main challenges and there are several chapters dedicated to to out improve weather prediction Observation data simulation. I think it's it's an interesting book and is free of charge Of course, you cannot resell but it's free of charge so you can download and Is a WMO publication, but I think that more than 100 the scientists around the world They worked to produce in terms of authors and reviewers to produce this book So it's really a community book and that's it's an important step forward I would say and I would like to I don't know if the sound is working In the if I run a short movie I don't know Adrian if the Because I would like to show you is one minute movie. So it's really sure short Because then I would like to ask you at the end of the movie if if you can summarize in one two keywords what? What you what is your first feeling? What is come out from from based on the video on your on your mind because this video is It will be used in the new WMO website as a kind of Presentation of world weather research program activities. So let's see if the Nicely for discussion so comments, please