Institute for Environmental Analytics: How Data Drives Environmental Decisions

This presentation is by Jon Blower, CTO at the Institute for Environmental Analytics. It supported his talk at Swirrl's 2016 data conference 'Data-Driven Decisions'.

Transcript

1. DELIVERING VALUE FROM BIG DATA 1 DELIVERING VALUE FROM BIG DATA WHAT COULD ENVIRONMENTAL DATA DO FOR YOU? Jon Blower CTO Institute for Environmental Analytics
2. DELIVERING VALUE FROM BIG DATA WHAT IS THE INSTITUTE FOR ENVIRONMENTAL ANALYTICS? •  Established mid-2015 with c.£6m start-up funding from HEFCE •  Aim to be a leading worldwide centre of excellence for big data analytics focussed on environmental challenges. •  Expert team with knowledge of data science, algorithm development, software engineering and data visualisation. •  Growing partnership with diverse expertise (science, technology, remote sensing, supercomputing, ...) •  Focus on development of pre-commercial big data technology applications with industry and public sector •  Helping users to make decisions using environmental data •  We work with all kinds of environmental data, not just satellite data
3. DELIVERING VALUE FROM BIG DATA IEA PARTNERS
4. DELIVERING VALUE FROM BIG DATA Application Identify Merge Analyse Display Earth Observation Social Media Open Data Telecoms Internet of Things 4
5. DELIVERING VALUE FROM BIG DATA 5 BRIEF INTRO TO EARTH OBSERVATION
6. Benefits of Earth Observation Wide area observation capability Uniform - same sensor used at many different places in the world Rapid measurement – sensors can be targeted by operators at any point on earth, including remote and inhospitable areas. Continuity - single sensors or series of sensors providing long time series of data Many sources are Open Data
7. © DigitalGlobe, Inc. All Rights Reserved. Image © Satellite Applications Catapult Ltd 2014. Optical Imagery example
8. Radar Imagery example 8
9. DELIVERING VALUE FROM BIG DATA
10. DELIVERING VALUE FROM BIG DATA 10 SOME APPLICATIONS OF SATELLITE DATA
11. ESA UNCLASSIFIED - For Official Use Sentinel-2 – mapping urban/rural land use S2-A launched 23 June 2015, •  13 Bands (VIR, NIR, SWIR, red edge), •  290km swath, •  10-20-60m resolution, •  10d revisit. Image of Pavia acquired on 27 June 2015 at 10:25 UTC, Credit: Copernicus data (2015) / ESA 11
12. 12 Situational Awareness using thermal imaging •  High resolution infrared monitoring of thermal anomalies. •  Refrigerated and Air- con buildings are illustrated as dark. •  Warmed up cars show a clear white colour- indicating their high relative heat Hot Cold
13. DELIVERING VALUE FROM BIG DATA Flood monitoring with satellite radar
14. Ground   mo)on  from   L’Aquila   earthquake,   Italy,  2009     ENVISAT  SAR   (few  cm   wavelength)  
15. ESA UNCLASSIFIED - For Official Use EO to determine water stress Source: ESA / ITC / TU Wien / VU Amsterdam / DWD / KNMI / Planetary Visions 15
16. hHp://www.met.reading.ac.uk/tamsat/  
17. DMR-CMS-SUPSC03-PRE-11-E © DEIMOS Space S.R.L. Precision agriculture - KORE 17 Augmenting a precision agriculture service using Satellite and UAV Earth Observation imagery to provide cost effective crop management advice
18. DELIVERING VALUE FROM BIG DATA SUSTAINABLE TUNA FISHING (“FREE SCHOOL FISHING”) •  EO data •  Chlorophyll •  Sea Surface Temperature •  Turbidity •  RSR •  Supplemented by: •  Wind Speed •  Ocean Current •  Daily (or as close as possible) for 2015, Maldives EEZ, As high spatial resolution as possible. •  Ecosystem model uses above data to predict fish abundance
19. DELIVERING VALUE FROM BIG DATA 19 MORE ENVIRONMENTAL DATA (MOSTLY OPEN)
20. DataSpring: A different approach to data access www.esspartnership.org www.thedataspring.org In addition to ‘traditional’ service delivery, we’re working across government to develop DataSpring, a one-stop shop web service providing direct access to authoritative cross-government environmental and EO data for developers and innovators based on consistent and open data standards. Reducing barriers to innovation, stimulating growth and supporting decision making Our data will be easy to find; easy to understand; easy to use, share and integrate
21. DELIVERING VALUE FROM BIG DATA gizmodo.com
22. DELIVERING VALUE FROM BIG DATA Fishing Bathymetry Renewable energy Boundaries Biodiversity Aquaculture potential Gravity
23. Image credit: ZullyC3P, Wikipedia CC-BY-SA iSpex Array of Things, Chicago
24. DELIVERING VALUE FROM BIG DATA BIG DATA DIVERSE DATA DYNAMIC DATA COMPLEX DATA NUANCED DATA IMPERFECT DATA
25. ESA UNCLASSIFIED - For Official Use From Big Data … to insight Credit: www.mashable.com | Sven Sachsalber 25
26. DELIVERING VALUE FROM BIG DATA 26 HOW CAN WE DEAL WITH THIS?
27. THE JASMIN DATA CLOUD •  Data-processing facility for environmental data •  In-place access to Sentinel and other environmental data
28. “The  uniqueness  of  SHEM  is  that  it  gives  us  highly  accurate  es9mates  of  what  a  stream   gauge  would  normally  be  sending  out  in  place  of  a  failed  gauge  when  a  flood  is  taking  place.”                  Tim  PeDy,  PhD  candidate,  University  of  Alaska,  Fairbanks   Streamflow  hydrology  es)mate  using  machine  learning  (SHEM)  project  is  a  Cortana   Intelligence  Suite  experiment  that  creates  a  predic)ve  model  that  can  act  as  a  proxy  for     streamflow  data  when  a  stream  gauge  fails.  
29. DELIVERING VALUE FROM BIG DATA SUMMARY •  Many useful satellite datasets are now Open Data •  Can still be hard to access, understand and use •  Good example of a silo! •  But many possible applications for public and private sectors •  Need to turn data into services •  A lot of supporting capability is needed •  Other (contextual) data sources •  Expert knowledge •  IT infrastructure/expertise •  We want to make all of this simpler for the end user! •  What are your challenges? How could we help? •  Could we think of a feasibility study or demonstrator?
30. DELIVERING VALUE FROM BIG DATA The Institute for Environmental Analytics is a unique flagship centre, formed in January 2015 with £5.6m from the HEFCE Catalyst Fund and coordinated by the University of Reading. j.blower@the-iea.org @Jon_Blower