Collaborative By Nature (#gecoenv) Liveblog

Today we are blogging live from the Collaborative By Nature: Interoperable Geospatial Approaches to the Environment event at the Welsh Government in Cardiff. The event, hosted by the Welsh Government and arranged by the IGIBS and NatureLocator projects is bringing together experts in the public and academic sectors on the sharing of environment research and data.

We will be live-blogging, taking pictures and capturing sound today and you will be able to find the slides and related materials here after the event. We will also be tweeting with the hashtag #gecoenv.

Please note that as this is a liveblog there may be errors, gaps etc. Please do let us know if you spot a correction or omission and please do leave comments here on the event itself.

Welcome to day – Chris Higgins, IGIBS Project/EDINA and Bill Oates, Welsh Government

First up Bill Oates is welcoming us to the Welsh Government. We are delighted to be working with EDINA on this project and we hope this will be the first of many collaborations of this kind around GIS. This project have allowed us to gain some insight and some technical skills and it is more important than ever that we reflect on what we do and this sort of collaboration with academia is a great way to do this.

Chris Higgings of the IGIBS project introducing today’s event with the timetable for the day and explaining the themes he set for speakers today: INSPIRE; Wales; JISC; Academic Sector; Future; UK Location Programme; Benefits. I am particularly interested in finding out how the academic and government sectors can work together

Chris Higgins, Workgroup Leader, EDINA, University of Edinburgh

Nicola Osborne, Social Media Officer, EDINA

Dave Kilbey, Nature Locator Project Manager, University of Bristol

Richard Lucas, Aberystwyth University

Steve Walsh, Reasearch Assisstant IGIBS, Aberystwyth University

Simon Hatherley, Researcher at Cardiff Metropolitan University (previously UWIC)

Steve Keyworth, Director, Environment Systems – I’m speaking later today.

Tim Pagella, Research Associate, Wales Environment Research Hub – also systems services division at Bangor University

Lucy Bastin, Senior Lecturer, Aston University

Steven Bromley, Development Officer, Groundwork Wales

Harry Gibson, Spatial data analyst, Centre for Ecology and Hydrology – particularly working on Hydrology side.

Steve Gardiner, GIS Analyst, Neath Port Talbot County Borough Council and also working on South Wales Transport

David Andrew, EMA, Flood Risk Assessment

Richard Fry, Researcher, University of Glamorgan and I also work on the WISARD Portal

Rob Berry, GIS Researcher, University of Glamorgan

Diana Waldron, Research Assistant, Welsh School of Architecture-  Low Carbon Research Institute

Tom Bassett, Building Physicist, Welsh School of Architecture, Cardiff University

James Morris, Planning Manager, Welsh Government, work broadly on climate change

Simon Agass, GIS Applications

Simon Agass, Corporate GIS Project Manager, Welsh Government

Bill Oates, Head of Geography and Technology, Welsh Government

Michel Koutrampas, EDINA, wrote software for IGIBS

I think we have a great cross section of speakers and participants today – hopefully we’ll have lots of opportunity for discussion.

Introduction – Nicola Osborne, JISC GECO Project

I gave a brief introduction to the JISC GECO project so will not be liveblogging for a wee while.

INSPIRE and Wales – Bill Oates, Welsh Government

As well as participating in the Welsh Governments work on the INSPIRE directive I’m also part of the group looking at the UK Location Programme. I would like to give an overview of how we’ve responded to INSPIRE in Welsh Government. And we will be interested in thinking about how those of us around the table can contribute. And when we make information available we need to make sure that we stimulate exploitation of the resources we have made available.

We are just at a milestone where we need to make INSPIRE services up and running. We have been looking first at the Annex I themes. Next year we will look at download and transformation through vector services. And we will be looking at Annex II and III themes. And we are a little challenged at the details of the technical requirements for services are not yet finalised – they probably won’t be until February 2012 and there were over 6000 comments about the specification so there really remain some wrinkles and it’s quite an ambition for the public sector across Europe to respond to.

So far there are over 1000 data sets already available. Some of the big organisations with a history of disseminating spatial data sets are taking the lead – the British Geological Survey, Centre for Ecology and Hydrology, Scottish Government, Barrow Borough Council (they have a great GIS officer), the Environment Agency and the Welsh Government are all included here when you look at who has made those services available. As of the end of September 35 providers had registered Annex I and II data sets. There is a strong environmental theme here – particularly strong sets around Protectected Sites (Annex 1.9), and Geology (Annex 2.4). But there may be included there single data sets that actually break down into a far broader collection of underlying data sets.

Now there are 19 providers and 131 Annex 1 and II View Services available. Some of the groups with delay have included the Northern Irish data providers but they are moving on. Transport has also been a slower area to catch up – lots of challenges around operational data that may need abstracting from a long used system and provide it as a WMS but there has been progress there and those data sets are coming along.

So how did the Welsh Government make their services available. Well we have been able to deliver INSPIRE for very little investment through use of Amazon Web Services. And we’ve carried forward the IGIBS work on digital rights management. This has been supported by the availability of the National Mapping Agencies who have invested in open source technologies. So for WMS the Ordnance Survey supporting the development of GeoServer. The use of open source is something we want to do more of but we want to identify support for that work as we use those tools. This is a data model that works well for us. We were doing innovative stuff that didn’t work on the WAG data server so we decided to test Amazon Web Services for the year and we think we’ll continue on with that model as it works well. Elsewhere organisations have outsourced their requirements so ESRI have been doing some work in this area – the Northern Irish data is likely to be based on ESRI, and the Environment Agency have used their contract with IBM to deliver these types of services. Lots to consider here – is the diversity of solutions a strength or a weakness – we shall see how that plays out.

In terms of implementing the View and Access services there are some issues, particularly with how the services have been comprised – these are to do with the INSPIRE implementing rules which are very domain specific and the committees that have drawn up those guidelines are domain specialists. But many of the providers work across many domains so the way in which these specficications are provided can be tricky for applying to broader data sets.

We have established 4 virtual data areas along thematic areas (Administrative units; Protected Sites; Natural Risk Zones; Transport Networks) – some data is and some is not INSPIRE. And we’ve also experimented with automatic metadata generation – we have tried to stand up the service and have the tools generate the metadata from it. The BGS have classified most of their data as INSPIRE and they have a single View service for all of their data – they provide each theme through layers.  The Ordnance Survey have a single on-demand service for each product and they have not resegmented for INSPIRE.

The other particular concern is the issue of how we protect our services – what you can see here is that in the INSPIRE network there is a layer here – GeoRM Layers – and that remains an empty space at the moment. There isn’t an INSPIRE concensus on the right way to approach rights management. Geo rights management is important for a number of reasons. So far the approach has been to look at exemptions, use of simple watermarks, and a DRM wrapper around services – username & password in service request, ip address restrictuions, end-to-end identity mabagement and services protection. IGIBS fits into the last section of that approach.

We have made some progress on exemptions around derived data and that has seen some significant success with the PSMA licence. But the vector services are a much trickier issue  – it is specifically prohibited under the PSMA licence. So we have important work to do here. The benefits for rights holders of protection is not only to do with the rights holder and their commercial interest but we also want to know who accesses what – customer intelligence – and some data sets we don’t want to share with the public for Data Protection reasons, because they are not yet completed, or because we need to evolve those services so that the data is right – such as an electronic land rights register – before making it public. And there are lots of data sets that should be private under normal business but at certain times they need to be made available quickly.

So does it work? Yes!

I was in GIS as a consultant before becoming a civil servant and I can vouch for the amount of time it takes to collate geo data. So to demonstrate this I managed to combine multiple datasets from England and Wales in an Open Source product in an hour – really shows the time savings and future shape of how data that is interoperable and available and how easy that becomes. We hope that in the future we don’t just get asked for data sets but also for specific data. And the benefits of providing data on screen rather than other methods are significant – real time savings, our planning team thinks that first stage of data transformation being available saves them about a half a post of work. The next stage is to make the data available and connected to other types and sets of data. One area where the academic sector has a real role to play in educating GIS specialist and pushing that cultural shift.

The challenges ahead? Well we still have a bulk of people to publish their data. An awful lot of those it’s not about upskilling and training those services but increasing the role of service broker organisations – a sort of bureau for the public sector. In terms of running those bureau type services. There are quite a few business issues. One of the key things about the November 9th deadline was that we needed to adhere to the Quality of Service guidelines as per the implementation guidelines. We only have a draft of what might be done there – we have to legally comply but we do not have the confirmed information on what we comply with. But making those services available will help us work out how to deliver and style services in the future. And we have to be ready for WFS (Web Feature Services) next year. This is a bigger issue in terms of transforming data to different schemas. There’s a bit of a skills gap there – not neccassarily part of the day to day work of the GIS profession. Perhaps some of you here can help us equip the next generation of GIS professionals to do that type of GIS data transformation.

So the overall story is a pretty good one but there is quite a long way to go in terms of how we roll forward. What I said at the recent AGI conference was that to me… I think many in the public sector see INSPIRE as a set of deadlines to meet but I very much think of it as a culture change for how we work with geospatial data. Infraction risk is not a big issue here – the EU isn’t prepared for that – but it’s much more an opportunity to improve and make more efficient out service delivery. It’s a way to do business better not a liability for us. Programmes like JISC GECO are absolutely vital to helping us deliver this.

Q&A

Q1 – Richard Lucas) What are those main missing training aims? And with what approach?

A1) Certainly in terms of the skills and capabilities within my team here – I have a staff of 20 – the direction of travel is away from Desktop GIS. So deemphasizing the role of Desktop GIS and finding a way to provide the underlying data. The ability to transform data, to provide the web services and the ability to work with that data themselves through generating apps. GIS is no longer about developing desktop software, it’s about a lightweight client and probably in the web browser whether Open Layers, Flex etc. It’s JavaScript. The academic environment provides a fertile space for that sort of activity – how much is curriculum based and how much of that is students being in the right place to through ideas around I’m not too sure about. It’s tough to keep up to date with all the latest trends and technology though.

Q2 – Tim Pagella) I’m relatively new to this. What struck me that there are 11 elevation data sets and I was wondering as a user how coordinated that delivery is so that I can make an informed choice about what is available, at what resolution, and which I should use.

A2) The INSPIRE directive are very much geared to that. I’ve talked about delivering access and view services but there is a whole other area of specifications around metadata and the use of geo portals. data.gov.uk is the UK Geo Data portal and shortly there will be a special spatial data area based around spatial data sets. There’s been a collission of two different worlds working with the data.gov.uk guys. They’ve come from the Open Knowledge Foundations, from CKAN. In the spatial world we are used to systems nearer the traditional library catalogue software world and the OGC standards has it’s roots in the library technology domain. So putting this freewheeling and the structured librarian approach together has been quite challenge. The method to find spatial data sets on data.gov.uk isn’t perfect yet but getting there. There remain issues around the “handshaking” process for harvesting metadata from data providers uploading their metadata but most of those has been fixed. There are also plans for an EU portal harvesting national geo portals.

Q3 – Steve Keyworth) You talked about the Welsh Government data centre and your Amazon web services. There is lots of talk about a sort of GLoud government cloud. What’s your general understanding

A3) Well it’s an even more crowded space – as there is also thought around a spatial data government cloud for Wales as well. INSPIRE fits well into this space though. The discussions we have had with underlying vendors is that they want to move rapidly to get government certification for the cloud services they provide. It provides an opportunity for the academic sector to participate as well – what you will see is a coalescing around a relatively small number of platforms available in a cloud type service – whether it’s a portal or whether its effectively a JavaScript API. You can see an underlying ecosystem of widgets around the GCloud and that’s a really important way to consider how we deliver spatial data in the future.

NatureLocator – Dave Kilbey, University of Bristol

I have been managing one of the JISC Geo funded projects which is a very short 8 or 9 month project but it hangs off a longer running project called Conker Tree Science and that’s a NERC-funded fellowship that Micheal Pocock’s been working on this for a number of years. When the funding call came up we saw an opportunity to create something new that ties into the existing system he was using. I don’t know how many of you are ecologically minded but there are many damaged trees around and you may have spotted that this is a problem with Horsechestnut trees and the problem is moths. The Moths lay there eggs early in the season and they lay their eggs and the larvae mine through the leaves causing extensive damage that gets worse throughout the year. The trees look autumunal in August but that is the damage. We thought we could help Micheal out with his project. People were going out and recording where they’d seen moths with a webform and low res images – there is a fungal infection that looks similar to moth damage – so we said why don’t we do a phone app so they can take the picture as they see the leaf, the location can be grabbed automatically, and a damage score can be assigned. Damage changes at different times in the season depending on the level and time of infection. We also want to know the ground coverage between trees as well before being submitted to us. The app was simple, easy to use and aimed at anyone – a scientist would be great but we wanted loads of the general public to take part no matter how little experience.

We were still developing the app at the end of June and the season starts in July but we were delighted that there was over 12,000 downloads of iPhone and Android apps between 1st July and 1st October 2011. 5200 record uploaded, of which most were valid (a few had to be stripped because of lack of pictures).

But opening things to the public… well…

• 2 bottoms
• 1 that I won’t disclose as too distressing
• 1 nine-year olds birthday party
• Assorted naked feet

You have to be a bit careful and we wanted to weed those images out so that the public wouldn’t be exposed to them. We are now looking at automating the future version.

Public involvement was already part of the Conker Tree Science project, NERC are very keen to get the public involved in science. It opens up a world of potential. And it puts at least one potential recorder in nearly every 10K square. BUT you have to ask your questions very carefully, clearly and unambiguously. These systems can be subject to abuse. The National Trust is using an app of this sort and are not moderating submissions at all which I think is increadibly risky.

Interestingly looking at the National Biodiversity Network map of Cameraria ohridella (the horsechestnut tree moth) and their current map shows clusters of damage. Their map represents years of data collection. This really shows the power of crowd collection as the number and scale of records we have been able to collect in just 4 months has been increadible. It’s not a distribution map of the moth but of those taking part BUT it does give us a good idea of where the moth is. We have no verified records from Scotland – they are either erroneous or of healthy trees which we are also interested in.

We used Google’s App Engine and it lets you see the data in different way – you can do heatmaps of distribution. We did well for publicity given that we didn’t have time for publicity. Stephen Fry spotted the app and Tweeted about it and that was a huge boost. Someone at iTunes was

Farming Today – the pinnicle of my career. We got on the One Show as well but someone working on the app made a few tweaks that inadvertantly crashed the app so we had all that publicity was that people were excited only to immediately find the app crashed/not working.

The app is great but we also wanted a landing page to maintain a presence for the app. We also did a few other things. We thought we must put results up so that people can see what they have contributed and how it has been used. But we also decided to crowd source the data validation – we want to see if the public are any good at collecting the data and assigning the damage scores. We decided to open that up again. That’s also proved quite successful. One chap has gone through about 3000 records. There were 4 people going neck and neck on our leader board for record validation. But this relies on public good will, you can’t rely on people to do this too much. The leaderboard was a wee bit of gamification to engage people because other incentives won’t neccassarily produce good results. So far we have had about 400 people contributing to the validation and will do some stats on how effective that process has been.

And we’ve also created a damage map that shows spread and seasonality of the moth. We think the heavier damage will push further north and south. It would have been impossible for a team of biologists to do this would have been impossible. Using smart phones has been fantastic, very easy, and we’ll be doing this again. The project was funded with £65,000 from JISC and the institution put in an equal contribution. We had a team including 9 people – 2 programmers, 1 designer, a project manager, 1 usability expert and 2 biologists in the steering group. It delivered “accurate” crowd sourced data and public engagement to a far greater extent than we had expected. The app produced 6 times the data – and far more useful and accurate data – than the previous website. And it really helped increase the scope and reach of the project and massively enhanced public awareness of the problem.

The future? Well it’s not just a biologically constrained piece of software. We are looking to enhance the app so that whoever is delivering the app can access the data more easily (currently via a programmer and via scripts). We want to look at user feedback – for other damage – as well as ways for the app being able to feed you a task to keep your eyes peeled for a particular environmental phenomena in the area you are based in. And the more platforms the merrier – we went for the dominent platforms in our demographic but we are actively looking at BlackBerry and Windows Mobile. We are actively looking into developing this as a service – the environment agency are interested. It’s a great tool for plants but also largish animals. It’s a great tool for recording.

We have an Android phone and an iPad app here so you can see the app in action.

Q&A

Q1 – Richard Lucas) Is this the first of its kind?

A1) There are similar ones out there. We looks at one called epicollect, intended for epidemiology and biological recording. The work to adapt the code would have been more expensive than developing a new one. There’s one at CH looking at monitoring harliquin ladybirds. So it’s not groundbreaking but we wanted to make it generic and really usable. We think this could be great for invasive species like knotweed for instance

Q2) In terms of platforms used to deliver this – how straightforward would it be to add other data feeds for this. I’m assuming the development toolkit channels you to use Google Maps say? Perhaps you could deliver specific maps for specific types of area?

A2) We used open source tools, Google maps, etc.  We had a lot of problems delivering maps – if 200 people using the app simultaniously Google would cut off. We actually reduced the mapping interaction but people want instant feedback – an instant pin on a map – so we are thinking about that

Q2) I think we would be thinking about, as people with data sets to provide, how we could integrate data with the app.

A2) You would need to speak to my colleagues about that.

Q3) You had some outlyers there in Ullapool. Did anyone go and check? And did you have any means of comparing results with a model of species distribution in order to come up with a model of uncertainty – there’s been some work at Nottingham around that?

A3) Actually the picture helps us be certain about correct records. But some were either erroneous or healthy trees. We haven’t got that far with analysis yet, we’ve been about collecting the data. We also have someone at the university working on software that automatically detects specific types of diseases and damage so we may link into that in the future. Unfortunately the validator for the moment was me.

We are just headed to lunch but I’ve just had a great offer from Bill Oates, Head of Geography and Technology, Welsh Government – if you post any questions here about Wales and INSPIRE – or anything else on his presentation earlier today – he will answer them. So please do get posting your questions!

Interoperable Geographic Information for Biosphere Reserves – Chris Higgins, EDINA

This project has been a partnership between EDINA, the Institute of Geography and Earth Sciences (IGES), University of Aberystwyth and the Welsh Government. We wanted to look at how different datasets around the same areas can be made available within academia and how the Welsh Governments data can also feed into and be combined with this work.

The project is based in the Dyfi Biosphere – a UNESCO

SDI – this has elements including INSPIRE, the UK Location Programme, UK Academic SDI and some key EDINA/JISC contributions to the UK Academic Location Programme.

The main aim of the project is to improve UK SDI Interoperability. The public sector wants their data used well, they want highly skilled and experienced graduates and the universities want to make the best use of their research and to skill those graduates up.

We have four deliverables that include a Best Pracice model for using UK academic SDI at the department level – Steve Walsh at Abersystwyth University; a working prototype for the WMS Factory tool – an easy way to turn your data into a web service.

We think that the use of spatial data interoperability standards will be a huge benefit of this work.

Steve’s report on INSPIRE sets the scene of the increasing amount of data available and the importance of disseminating and justifying the public funding that supports academic research, and how interoperable data can come into Data Management Plans, how that can feed into publication plans. I tried to look at what can be done and what we have done at Aberystwyth. We looked at the Dyfi river catchment areas and found 139 data sets around it and only a tiny proportion were actually available to use. The JISC is interested in research data management and at the moment it is a bit of a mess but it is being addressed. You’ll all be familiar that the outputs of projects can sit gathering dust. They might spend millions gathering data and can then lie dorment. The report will be available for download on the IGIBS blog.

The WMS Factory Tool – or view service if you want to use the INSPIRE jargon. The idea is that a user sitting there doing his project on the BioSphere, he is generating shape files but he wants to quickly see his data in a variety of maps, he can quickly instantiate a web maps service and bring in government data without the need to transform or preprocess that.

Access Control continues to be an unresolved issue within the global SDI community. There is no global method nor an interoperable way of sharing controlled data globally. In the UK we do have the UK Access Management Federation – EDINA runs this service and it is based on Security Assertion Markup Language, we’ve also managed to map this to OGC standards and we’ve been attempting to transfer some of that knowledge into this project.

So now over to Micheal Koutrumpas for his demo.

It is not just a WMS factory tool but also an INSPIRE view service and a security tool. And we showed how access control can be applied to OGC Standards. To prove that this solution is a working one. And this demo is available already (http://dlib-mumra.ucs.ed.ac.uk/igibs/) and the code is open source.

A user can come with their data, any projection and can upload it. The data is then automatically probed for metadata – the user can also edit that metadata to make any corrections but clicking on a field – say keywords, language etc. There is a field for Shibboleth Protection – that allows you to create a protected view service and that allows combinations of data to take place. You can also title your data. For each layer detected you can set colours, labels, etc. So in this demo we will add 4 layers here. You can obviously reuse the generic WMS service from a web or your desktop client – we have actually made a version of OpenJUMP which is able to use Shibboleth authentication so you can use it with protected WMS. Once the settings have been chosen you can click “Create WMS Server” and that will give you a URL you can use, share, etc. And/or you can go to view that URL. In this case we are using a protected WMS. You login via the UK Federation (single sign on so you only need to do this one) and you get to the generic client that uses OS free maps. The layers included are those selected by the user. As a viewer of that data we can click on any layer here and see it on the map.

Steve: We have used data here from within the Dyfi catchment area, boundaries, forestry holdings, areas that show the right to cut peat in the Dyfi Nature Reserve etc.

Micheal: So here you can add the welsh data sets – again all under the same login and shown on the same map. This is combining the welsh government data – and this required the Welsh Government putting an authentication layer for shibboleth on top of that data – with the uploaded/viewed data. So that’s it.

Q&A

Q1 – Bill Oates) So to be clear you showed two tools: the first is to publish data, the second is the client bit. What if you came back and wanted to get into that view client and that earlier data set along with a new one. So you can either view existant data or add in a WMS from elsewhere.

A1) What we do here is that the WMS can easily be used with a catalogue service and add that WMS in. You can generate and publish that WMS. Then you can use any Desktop GIS client to access as many WMS as you want. If there are many you want to gather you can add all of those WMS in a catalogue. So you can use these WMS in existing systems. We want a WMS service not a catalogue or discovery or view service really. And you can customise that WMS

Q2 – Tim Pagella) Is there any check in place to stop inappropriate data getting posted up?

A2) The system cannot do that so that’s down to the user to do – and they can already share data wherever they want

Q3 – Lucy Bastin) Is there any way of indicating a license for data or the copyright for data if you can share it but only in certain ways?

A3) You might be best doing that by making a protected service. But you can also edit the metadata. And this software is open source so you can deploy and customise this software to include that field if you like.

Q4) Can you use protected layers with desktop clients like QGIS etc?

A4) The problem is that the desktop software needs modifying to handle the protected layers. We have already done this with OpenJUMP. Would be good to do this to QGIS (Quantum GIS) as well but you would need to make additions to the code. Some proprietary softwares are implementing Shibboleth <check list with Chris Cadcorp>

Q4) Surely WMS was picked as it’s so standard – surely changing that defeats the purpose?

A4) No, we are not changing anything here, that’s the point of how this works. This is WMS and we have not changed WMS. OGC are higher level application protocols, the authentication has to do with the services themselves.

A4 – Bill Oates) From a publisher perspective it’s straightforward to implement – we didn’t need to change those services, we just needed to add a Shibboleth layer on top of them. In the first draft of INSPIRE there was a great deal of concern about access control as the original draft would have required all desktop applications to be significantly amended to work. This is a great halfway house really. As publishers we have to spread our bets a bit as the user services come after we put that data out there. We hope there’ll be a single or small number of standards for access control. The most difficult part of the authentication was the identity management provider aspect. At the moment that is most politically different. For academia there is one UK Access Management system BUT in the government world there is a different identity provider in use, the Government Gateway. It would be relatively easy for us to participate in the Federation but it may not be compatible with GCHQ and other high security uses of Government Gateway.

Q4) I think our organization would have been behind INSPIRE earlier if it hadn’t been the need to publish openly. Our IPR managers don’t feel comfortable with the idea of insecure services.

A4 – Bill Oates) I think actually we are better off with the open requirement that we then look at applying control on rather than the other way around.

Wales Environment Research Hub – Tim Pagella, WERH

The Hub exists to provide coordination and promote collaboration in the environmental research sector, to strengthen the evidence base for practical politics and actions in support of the Wales Environment Strategy. Now subsumed by National Environment Framework. We are based in the Bangor in the Environmental Centre. WERH is supported by various stakeholders in Welsh environment research and we act as a sort of knowledge exchange community. Both between the public and academic sector and look to provide immediate evidence needs and future-orientated assessments.

So at the core of the sustainability and environmental evidance division of SEED we have loads of organisations and academic institutions who feed data in to inform policy. The hub sits between some of those organisations and the Welsh Government researc programme priorities. There is that issue of academia and policy which can be challenging.

The hub tries to build databases of where bits of research have been done in Wales and generating information on research which has been done in Wales. And then we disseminate that work to stakeholders through websites, reports and workshops across the country that raise awareness. We also have hotdesks in the Hub office that researchers can use.

I wanted to touch on priority work themes to try and capture the importance of geospatial approaches in what we’re doing.

1. Ecosystem Services and valuation – that is our main priority
2. Climate change – mitigation, impacts and adaptation
3. environmental monitoring in Wales
4. Land management

Ecosystem Services and valuation – that is our main priority – these are benefits people derive from nature. Where human activity has modified natural ecosystems the services derived from them has changed and most often been degraded. Provisioning services for instance; Regulating services – water usage/quality for intance; Cultural services; and Supporting systems – humans have turned natural ecosystems into agricultural usage space with implications on how the environment can or cannot sustain itself.

Geospatial data is an essential part of understanding those ecosystem services. The Welsh government is increasingly interested in managing our landscape and interacting with ecosystem services to make the best decisions about how to manage them.

So looking at a figure of the environment with key services. Green infrastructure is important in urban areas for health. More widely clean water and flood risk becomes important. The key point here is that where things are matters both in terms of services received, services generated and we need to consider both at a broad range of scales.

The Hub has been involved in writing the Welsh section of the UK National Ecosystem Assessment – capturing the current position in Wales. It was a fairly rushed process. We did want quite a lot of maps in the NEA but quite a lot had to be removed. The discussions today are very very relevant for us. Looking at our diagram of delivery of ecosystem services we can see that each system costs and impacts on other systems. You can visualise the relationships in simplified form. The key thing to note is that there are lots of uncertainties – these are starting rather than end points. But it’s a starting point for finding cost effective changes. We have been looking at comparing, for instance, Social Deprivation and Environmental Quality in Wales – there is a known impact of the environment on wellbeing so this is a very useful comparison to make.

There is now an attempt to operationalise the ecosystems approach in Wales via A Living Wales: Natural Environment Framework. We are in a very responsive situation, it’s a very live situation. We have frequent demand for maps and GIS is part of this NEF. Certainly in terms of engagement and communicating with other stakeholders those discussions about what can and cannot use have real importance for us. GIS is also a key part of the evidence base and as part of communicating with stakeholders. Important to understanding how and why changes to ecosystems are being made.

So in a spatial sense we are looking to interoperate large data sets at a number of different scales in quite a complex fashion.

We have recently done some work on spatial tools available to use more widely: WERH Report: A review of the potential use of spatial tools for decision making in ecosystem service provision. So also looking at the output from the InVest project for instance.

And coming back to the flows in the system between ecosystem services we wanted to look at how and where we can map these. There are 52 ecosystem mapping approaches in the last five years. We are now at the point of research looking at insitu provision of ecosystem servicesand areas such as the flows, beneficaries etc. are very understudied. There is also work focusing on National and Regional scales but there is a strong need to implement policy at local scales for effective ecosystem management.

We are keen to map some data but also to use maps to communicate. So that’s where we’re coming from in terms of the priorities of the Hub. This work you’ve been doing is particularly relevant to us. Knowing about data.gov.uk is really useful and I’d be very keen to follow on to the IGIBS work and the Hub may also be able to play a role in spreading the word about this work. And finally I am here representing Dr Shaun Russell, Director of WERH and he’ll be happy to hear from you with any comments or questions you may have.

Q&A

Q1 – Chris H) How can the Hub help with spreading the message and increasing capacity in Wales

A1) The hub itself is very small but we can create a fora, we can put workshops on. It’s the networking component here. And we’ve certainly led workshops to increase connections between academia and policy.

Q2 – Bill Oates) Is there a strong enough concensus about research services that we can create tools that would let us exploit data coming on stream and build upon that

A2) Well personally my feeling is probably not because even the typologies of ecosystem services are fairly vigerously debated. That’s not to say that more couldn’t be done but at certain levels we’ve still yet to reach concensus. We are being pushed to produce things very very quickly particularly around things that academics are traditionally slow to respond to. But there is a really strong remit as well since we cannot wait to make changes – we can’t wait!

Q2 – Bill Oates) Based on that work, the InVEST survey, etc. Do you think it’s realistic to suggest that they will ever be available. The research goalposts are a moving target, will there ever be a moment at which that is finalised. Is any tool likely to provide policymakers in the space with the kind of confidence they’re after… or will it just be creating pretty pictures for reports

A2) I think, again personally, I think we can. The issues are very much related to scale, the scale at which we make decisions. I work with local level tools with farmers where we come up with strategies for dealing with different issues.

Q3) Is that separate from policy?

A3) To some extent

Q4 – Bill Oates) We are used to thinking nationally so perhaps the lesson for us is that the data will be best exploited at more local scales

Comment) Yes, that’s where the interaction comes in really.

Environment Systems – Steve Keyworth, Environment Systems

Environment Systems is an SME (Small/Medium Sized Enterprise) based in Aberystwyth and I’m giving “a view from other the hedge” on policy and how business – SME sized business – can get involved. We have about 23 staff. We specialise in earth observation and GIS. Although we are in the private sector we are in a number of camps. We’ve all played a bit of Twister – we have feet in many camps. We work a lot for the public sector in the UK but also in the private sector throughout the UK and further appeal. So I shall try and think about how we export some of this work as well as what we do here. We certainly don’t work alone – we have a very good relationship with Aberystwyth University – we are sponsoring 4 masters students, we part sponsor postgrad and PhD students. A lot of my colleagues have MScs or PhDs – from Aberystwyth but also Edinburgh, UCL etc. I am interested in the full circle of this, the other products of the academic system.

To my mind we have to focus hard on the challenges – that’s where the business is. So I’m going to work our way down here. Some of the big challenges for society include increasing population, food security, energy security. I heard a great talk from Professor Sir John Beddington talking about population pressure – possible plateuing at 9 billion global, increasingly urbanised. By 2030 we need 20% more water, we need 40% more energy etc. We need to engage with those bigger issues on a larger stage. But that is a very tricky thing to quantify and bring down. We have Land Use Futures, The Future of Food and Farmning reports which are both foresight studies from the Government Office for Science looking out for issues in ten, twenty, thirty years time. For big challenges time and scale are big issues.

Stepping down a scale you need to be able to distill those issues down to, say, the multiple pressures on the land – renewable energy, water regulation, pollination (gardens), natural floor defence – in the case of the example image we’re seeing (of Aberystwyth) a playing field is a form of flood defence; climate regulation; food production; recreational opportunities; new build. That’s no exhaustive but is about localising those bigger issues. Wouldn’t it be great to have the time to get even more specific?

So Japanese Knotweed – an item on BBC News site about a couple whose house has massively impacted on house prices because Japanese Knotweed has invaded – real societal impact. Looking closer to wales there are cases where in South Wales mortgages have been refused for the first time because of the presence of Japanese Knotweed. This is a fast moving issue! We have been working on a method for detecting the presence of Japanese Knotweed (Fallopia Japonica). It’s not perfect but it’s a quick way to get information out there. Really interesting to consider crowdsourcing this sort of information and to better understand the best way to manage the issue.

So this is another project looking at updating the habitat map for Wales. The revised map will replace the original Phase 1 mapping which took 3 teams of 10 field surveyors some 15 years to complete. And there was a need to update this expensively produced data. It uses a time series rule based approach to classify areas of habitat. This is currently being worked on by Environment Systems but also being checked and interrogated by Aberystwyth University – looking to research questions of the future. There is a website for this work which will be updated over the coming months.

Project Ursula was a large industry led project which simple switched the lightswitch on this sort of research. Using aerial photography to collect data – on small unmanned drones. This work has been taken forward with Environment Systems and Callen-Lenz consulting looking to agricultural possibilities etc. I’m interested in “research with a small r not Research with a capital R”. This has been a hugely beneficial process for us but it is an investment for us.

So why are SMEs interested in collaborations of this sort? Well we have to be innovative and invest. We have to stay responsive. We have to deliver on the challenges faced by our customers. And we also have to make a profit and we shouldn’t be ashamed of that – we reinvest in research, in sponsoring PhDs and masters students etc.

So why do we want to work with academia? Well we want them to continue punching above it’s weight in terms of research – Royal Society stats show that the UK are 1% of the population but we get 3% of global funding for research, we publish 3.9% of papers, 11.8% of world citations and 14.4% of the world’s most highly cited papers.

I have some material here from Bret Victor, he’s ex-Apple and he has some really interesting ideas about the role of design and the importance of hand held devices. So I’m nicking a few of his images. We need to make sure we have the right tool for the job, it has to address it’s human need and amplifies human capabilities (e.g. a hammer). And the tool develops what we can do into what we want to do – great tools both deliver what we can do and what we want.

Some of the challenges we face are so big and so complex that there are huge combinations of data and information and sources that we can access and combine. But there are also huge challenges about the reliability, comparability of that data. But we do need increased access to data. I do like INSPIRE but it will only get us so far. But telling us what we have got so that we can identify what is still needed, what should be captured.

So just drawing together some conclusions. We musn’t just get stuck on the technical possibilities, we need to use tools wisely. Foresight studies are really helpful. As increased devolution takes place we need to see how Welsh challenge fit into European and global challenges – and those relationships change over time. Operate within the timescales adn risks that policy, service delivery or business use need.

We each have our strength in the private sector

Q&A

Q1) With all this data around are you involved in producing open source software?

A1) We mostly use propritary software and don’t do much software development in general. But we are keen to use and build upon tools and data produced elsewhere.

Q2 – Bill Oates) Do you think the transition from a research problem to a repeatable almost a commercial problem that an ecosystems approach helps make that leap? Is it likely to make that leap into production?

A2) Almost certainly because it looks as if the policy requirements but also even within industry there is interest and important on ecosystems approaches because of the values that are placed on them. The economic case is made but it shouldn’t be taken lightly that the transformation from a research demonstrator to a production system  – it can be a big leap requiring a lot of planning. The questions are posed hard and fast but some of the answers are known – but you also need space to make those those answers known, to understand what they may be.

Summary, conclusions – Nicola Osborne, JISC GECO Project