 Thank you very much. I'm Martin Brwythwych, a senior hydrologist. I've worked in the flood hydrology Irish Improvements Programme, which is an R&D program, for the Environment Agency. I'll say a little bit about that in a moment. I've told them that by engaging with emerging digitally enabled environmental science, bit of a mouth full. But that is essentially what we're attempting to do in this program and it's representative of some of our approaches to how we work with what's coming forward through research i'r ffordd a fyddechrau i'w ddiweddol i'r ddarparu fyddechrau sy'n gynyddiadau sy'n gweithio. Mae'r ffordd a'r ffordd yn y fyddechrau i'r wahanol i'r ddweud am ystod o'r gweithio'n byw, os yw'r gweithio ar y ddweud. Yn y trafodaeth yr bobl, mae'n dweud am y ddweud o ddweud o 10,000 staf. Felly, mae'n ddiweddol i'r ddweud. A'r ddweud yn y lleidio'r dweud fyddol iawn i'r ddweud, ond yw'r nefyd, ac yn ddiolch yn mynd i gael gan Unedig. A rydym ni i gael eu gwirionedd iawn o ffordd y lluniau, sydd wedi'u wneud gyda cheif y cofennuol vomel iawn i fynd, i gychydig iawn o ffordd am goesyn i amser i gael i gael i'r Fab mégol yng nghydweithreinnol yn rhan oed. Byddai'n ymlaen, a oeddwn ni'n rhan o'r cael eu gwirionedd iawn o bod mage unig i unedig eu llef yn gweithrein niad wyliadau mwych o ffordd. back in 2020. The government doubled its expenditure to £5.2b over the period 2021 to 2027 to invest in better protecting up to about 333,000 additional properties. So, it's a very very big scale for us in terms of what we're trying to do. So, a little bit about how we approached that and why this digitally enabled environmental sy'n cael ei bod i'w ddaoli. Gwiyddwch merthyn nhw'n gweithio wrth hyn o ffordd rysg yn perlwyr, dwi'n gweithio ymlaen o'i drafyn dwyd. Felly weithio diwrnod o dwy TUF, rwy'n ei bod gweld i ddechnie wneud, mae'r iaith yn cael sut yчil yn eu boreg, ond mae'n gwell iaeth siwet hwn. Agen nhw'n gan gwybod, rwy'n cael sy'n cael beth yw'n gweithio'r iaith, bod y ysgrifennu leol, yn y trafnwyr, creating the money to provide better resilience for communities, flood protection works and so on. Byddo diwethaf yn wahist o phobliau o edrych sy'n cefnodig a modnau i gweithio a gweithiol. We start with meteorology data, weather data, measurements, of what we call hydrology i meteorology, so things like rainfall, river flow, groundwater levels and so on. Then we do some form of modelling. The bit that I am involved in the bit that are shaded in blue here yma, mae'n gyfrifoedd hygiologi, mae'n gyfrifoedd hygiologi, mae'n dweud i ddechrau, mae'n cael gweithio a'r wath. Mae'n gweithio i'r rhaglenau ac mae'n gweithio i'r environment cwrwp. Yn ddim yn gwneud yn ymddangos yn gwybod cyfwil o'r ddau i'r ddau y byddwch yn ei paru yn ei fod o'r ffordd o'r gwheiddiad, yr eu ddau o'r ffordd. Mae'n ei fod o'r ddau o'r ffordd oherwydd o'r ddau o'r ddau o'r ddau o'r ddau o'r ddau. Os yw、 roeddwn ni'n defnyddio maen nhw fathod o'r hyn yn sicr fel ydych chi erioed, yn credu'r rhai sydd erioed o'u hyffordd. Os rhai, mae'r rhwng yn sacgorlau gwsd, oherwydd a'r llwyddiad ny rhwng i gwell o'r un, mae'n hyfnod eich rhai sy'n llwyddiad ond y moddfod. Ond ond rwy'n credu i fynd ymddur ei gweithiol, a rwy'n credu'n rhaid i ddigon. Os erioed, os gallwn ni'n gofynu yn rhoi'n gallu'r sefydliad ei ddechid fuddol, Yn oeddeni chi ddim yn cyrraed y llwyddi'r cyfnodd arweithio'r cyfnodd, es i fod yr adeiladd gyflym wedi'i adeiladwyr sy'n gymryd a'r adeiladdu pwyll Bunod. Mae'r adeiladd yn agen nhw sy'n meddwl â'r cyfnodd yma, ond ry'n meddwl ar ymbiol ac y gallwn i adeiladdy o'r cyfrifiad? A phwer hyn. Mae'n ei wnaeth gael'u bwysig ac ychydig genneniaeth ar y cw他的fyn sy'n aru'r cyfrifiad. Mae'r ddafodol o'ch cyflawni ar gyfer y cynhyrchu sydd ymweld yn ddweudio'r awtodau yma. Mae'n ddweudio sydd ar gyfer y cyflawni ac yn ymweld ar gyfer y cyflawni ar gyfer y cyflawni ar gyfer y cyflawni. Mae'n ddigonio gyda'r ysgolion o'r ffordd o'r ffordd o'r ffordd o'r ysgolion o'r ffordd o'r ffordd a'u ffordd o'r ffordd o'r eich rôl o'r wyrdd. I'm not on the regulatory side, but I've got plenty of colleagues who are working in water resources, for example, and in fact they developed this original product, and it was to share water quality, water quantity information with customers, water companies and whoever. So it was a fairly small scale project to start with and it was a proof of concept and that was developed a few years ago. But over time it's gained traction, lots of people use it. Then the flood hydrology improvements programme came along, which I'm involved in, which is a six-year, seven million pound programme to look at how we improve doing the hydrology bit for flood risk assessment. And we said, wouldn't it be great if we could share more of that data to help our customers, consultants, other end-users do the sort of analysis work for us in designing flood protection works for communities? Because previously they'd have to go and ask us for it and that takes time. Somebody's got to get the data out of the archive, pass it on to them and that could take quite a bit of time in terms of days because of the other workload. If we can put it all into a system and make it publicly available, then the public has got it. And that's not just consultants, it's everybody. So we've scaled up significantly, so we've now got four billion stored data points. That's from around about 7,000 hydrometric measurement stations. That's stations that measure things like rainfall, river flow, groundwater level, river level. Also other items relating to water quality. And we've now got that in a mapped product. And it's hosted on the DEFRA platform, the.gov.uk platform. And it's our whole period of archive. So it's everything we've got in the past that was digitally available. Plus we update it where we can through telemetry feeds. So it's the best we've got, and that represents all our operational stations. So it's a big expansion of that service. We see that as providing huge value then to not just as consultants, but actually to the research community, to some of you here in this audience and your colleagues, if the data's out there, then you can do something useful for it. So it's yet another data source, but it comes direct from us. What we've learnt on the way, of course, is that as soon as you make that available, or as soon as you think about making it available, you then have to ask yourself the question, well, what's the quality of that data? What could people do with it, and what could go wrong? And we've been mulling over that for quite a while, as you can imagine. So we've done our best to provide suitable metadata to associate with it, to say something about the quality of the data, but that's still a challenge. And I'll come back to that when I conclude at the end about that, how much we share and how much we want to share of our data about trying to get that right. Demand is great. People are really using this data. We did a pilot exercise when we first released this, and the data was accessed over 100 million times in 18 months. That was via an API, and we had up to 10,000 calls per month via the web interface, and that was just getting started. That was before we kind of spread the word, really. The data's not completely free, though. You don't have to pay for it, but it's not free of responsibility. As soon as you, as an end-user, take that data and do something with it, you are then responsible for what you do with it and for what it means, interpreting it and how that is then perceived by any other people that use that data thereafter. So that's an important thing for us to bear in mind. Okay, it doesn't finish there. We've got loads of other data. We've got lots of old data on charts and paper records of all sorts. And we're struggling. I have a colleague who estimated we've got 10,000 years equivalent of these charts. So that's not something we're going to digitise overnight. It's only going to be done using scanners and old technology. So there's a challenge. So if that's something that chimes with you, then we've got a couple of conferences coming up. One is online, the Data 23 conference. It's one of those events on Tuesday, 26 of September, if you want to be involved. Dial in to that. If you want to do it in person at the Royal Geographical Society on the 7th of November, an in-person workshop for people to engage with us about how do we do this effectively. So, yeah, it's going from charts to digital. It's not easy. The second project I wanted to put before you was about how we, instead of just sharing our data and working with our colleagues in different parts of your organisation to bring data together, is how do we see ourselves as enabling some of the research and development to come operational practice. And that came up this morning. And it's a big thing for researchers. I used to be an academic. I know that challenge. So this is one project in this flood hydrology improvements programme looking at alternative methods for flood hydrology. That's how we estimate the magnitude and frequency of flooding and the shape of those flood events over time. So what we decided we're going to do is to attempt to benchmark what we currently do, set that baseline, and then measure alternative methods against that to see what they offer. That's not just for their scientific improvement, but actually are they useful. And that's something that is very precious to us. And I'll say a bit more about that in a second. So we want to develop this benchmarking framework so we can choose the best practical flood estimation and modelling methods for operational practice. We're working in collaboration with our partner organisations, Scottish Environmental Protection Agency, Natural Resources Wales, the Department for Infrastructure in Northern Ireland, just to get that broader UK view on this, and we're just getting started. We haven't decided anything yet. We're just collecting our ideas together. We hope to produce a system for benchmarking, and that will comprise a set of tests, data and metrics so we can quantify the performance, but also look at usability effectiveness in operational practice. And as I say, we'll review what we currently do. And this is in response to that volume of emerging science and modelling methods that are around. Can we pick from those things that offer a genuine improvement to how we do things now? So what could it offer? So somebody develops a new method. We put it through our benchmarking service. They publish their results and say it's been benchmarked. This is how it performs against other methods. It gives it some additional credibility. We might decide, can we go a step further? Can we make it operationally available by an appropriate risk management authority, Environment Agency, or one of the other organisations? And for that, we need to apply those usefulness tests, and we're still working on what they might be. My colleague Sally Brown this morning mentioned some of those things that are precious to us, and they are very important. How much does a change in the way in which we do things cost? How much training is needed? What data requirements are there? How long does it actually take to move a very large number of people at work within a particular sector over to a new way of working? And I don't mean people working for the Environment Agency. I mean all those consultants and others who do work for us in the flood risk estimation space. So it's not a thing that you can just do overnight. But we recognise we need to develop with these emerging technologies and solutions and bring them in where they are offering operational advantages to us. And that will add additional value and impact. So a few closing thoughts. We're evidence based. We use data. We collect a lot of data. We use a lot of derived data from models. It's really important. It underpins all our decision making. Uncertainty is also evidence, and we live with uncertainty. We've not talked a lot about uncertainty today, but of course we know all about that. It's very important to us as an agency because we deal with that weather data, that hydrometric data, and it is uncertain. And we have to make decisions using uncertain data. And that's something that we have to be careful of when we make decisions about what methods we might adopt in the future. Digital open data is making a difference. Lots of people are already using it. It's not completely free of that responsibility, but it's going to help change things into the future. We're sure of that, and we're gradually making more data open. There's a lot of demand from end users saying, really, you've got this. Can we have more? It's only as good as its metadata. It needs the story, the narrative around it, where do you get the data from? How is it measured? Is it any good? How does it compare with anything else you've got? And finally, benchmarking is one way of looking at bringing new science into the fold, as it were, operationally, where it does genuinely offer an improvement to operational practice, but it's got to fit those criteria, those usefulness criteria, and sometimes you can be disappointed because you think you've got the best thing since slide spread, and maybe you have in a way, but there may be various operational reasons why it can't be adopted right away. It may take time, but we're trying to develop an objective way of doing that, and I think benchmarking is an opportunity, but it's also a challenge, because the challenge is how do we get that right, and we've been holding workshops recently over the summer with academic and software developers just to get their thoughts on how might we do this, and what's come back is it's great, we want to do this, we want to support you, but it's difficult, so it's going to be a process, we're not going to have, it's solved in one go, it's going to take time, but that's why we have R&D programmes like the improvements programme that I've talked about. If you want more, we've got a website, email me, FHIP, environmentagency.gov.uk, do a search for flood hydrology improvements programme on the web if you wish. That's probably enough for me.