IUKWC Workshop Nov16: Developing Hydro-climatic Services for Water Security – Session 3 – Item 2 C_Prudhomme
1. Need for a set of consistent hydrological
projections for water-related climate change
assessment in GB
Climate services for the water sector: EDgE a
prototype for Europe
Christel Prudhomme
2. Problem statement: what is out there?
• Uncertainty in climate
change modelling
• Large volume of data
• Choice of model (CMIP5 ~
20 Global Climate Models)
• Complex climate information
• Water-related projections
(drought; flood; water
availability indices) absent
...
HadRM3-PPE
25-km grid
future climate
forcing
10,000 monthly change factors (P, T, ..)
25-km grid; 3 ES; 9 TH
5km reference historic climate
Daily 1961-2014; monthly from 1910
3. UK experiment: Future Flows products
• Ensemble of nationally consistent
hydrological & groundwater level daily
projections (1951-2099)
• Regulators and policy
• EA: ‘Case for change’ evidence to Government
• Defra: White water paper
• Water companies: must use Future Flows for Water Resource Planning
• Research and development
• Water abstraction licensing reform (Risk solution)
• Water Framework Directive River status assessment (SEPA)
• Noro virus through shellfish (CEFAS)
• Electrical earthing installations and heat pump potential (BGS)
• Aquatic bird population (U B’ham)
• Soil saturation and runoff processes (Cranfield U)
5. pace
On behalf
of for the European
Commission
EDgE vision
Hydrological
Modelling
For Climate
Impact
Indicators
Stakeholder
View
Web Delivery
System
Define
Case
Studies
Design Implement
An hydro-climate service
co-designed with
stakeholders
to break down barriers for
users of different
backgrounds
to access and understand
state-of-the art hydrological
predictions
6. pace
On behalf
of for the European
Commission
Stakeholder view: Focus groups
Norway UK Spain
What information do you use
now?
What would you prefer to have?
7. pace
On behalf
of for the European
Commission
Common needs
High resolution
information about
future climate
drought indices
Norway SpainUK
Metrics
Skilful seasonal hydrological forecasts
Hydrological indices
Potential evapotranspiration
Water temperature
Norway SpainUK
Interface
Simplicity
No post-processing
Interactive maps
Maps and graphs for download
Maps of thresholds or triggers
Choice of scenario
Different spatial resolutions
EDgE results in wider context
Skill & uncertainty information
Norwegian Spanish
8. pace
On behalf
of for the European
Commission
State-of-the-art modelling framework
CMIP5 (ISI-MIP): 5-GCMs ensemble GFDL-ESM2M, HadGEM2-ES, IPSL-CM5A-LR, MIROC-ESM-CHEM, NorESM1-M
Seasonal Forecast: 4-NWP ensemble NMME: CanCM4, GFDL(FLOR) (12 real.); ECMWF: ECMF, LFPW (15 real.)
Hydrology: 3/4-model
ensemble Noah-MP, mHM, PRC-Glob,
VIC routed by mRM
High
Resolution
Products at 5km resolution/ spatial areas
Time series: P, T, PET, Q, SWE (daily CP; monthly
SF)
Up to 33 indicators of river flow, GWR, SWE, SM
and climate
9. pace
On behalf
of for the European
Commission
SF SCIIs: Drought Event Aug. 2003
ecmf_sm_quantile_dist_08_2003_lead_from_1month_1quantile
%
e.g. Soil Moisture
Drought
Likelihood for SM to be
in lowest quintile
Uncertainty
representation?
Skill representation?
Access to data for low
skill predictions?
10. pace
On behalf
of for the European
Commission
Data store and web interface
The EDgE demonstrator is a web-based application to:
• Deliver information on climate services
• Allow different types of users to interact with the data
• Help users to make basic decisions
• Allow users to download information
11. pace
On behalf
of for the European
Commission
Co-generation: e.g. uncertainty representation
Indicators Interface
Case Studies
Capturing the Stakeholder View
Evidence Gathering & testingFocus Groups
Live demo and
Exercises
Display
examples
Voting sheets for
preferred options
12. pace
On behalf
of for the European
Commission
The EDgE demonstrator
Live demo
On chrome
http://edge.climate.copernicus.eu/Apps/#map
http://edge.climate.copernicus.eu/Apps/#climate-change
13. pace
On behalf
of for the European
Commission
Next steps
Hydrological
Modelling
For Climate
Impact
Indicators
Stakeholder
View
Web Delivery
System
Define
Case
Studies
Design Implement
Case studies: assessment of
value/utility
Low skill vs access to data
Operationalisation roadmap
Feedback on interface
Lessons learnt from other
services
http://edge.climate.copernicus.eu/ edge@ceh.ac.uk
chrp@ceh.ac.uk
15. pace
On behalf
of for the European
Commission
Metrics
Changes in soil moisture
Changes in groundwater
recharge
Changes in potential and
actual evapotranspiration
Changes in precipitation
Changes in snow water
equivalent
Changes in appropriate
drought indicators, including
drought duration
Area under drought
Changes annual high flow
(e.g. Q10 – the flow
exceeded 10% of the time
on average)
Changes in annual
maximum flood
Changes in annual low flow
(Q90 or Q95)
Changes in seasonal flow
Changes in mean annual
flow
Probability distributions of
future flows, for both
seasonal forecasts and for
climate change
Notas do Editor
I am Christel Prudhomme, and together with Glenn Watts, Luis Samaniego, Justin Sheffield and Matt Fry, we will be presenting the vision and solutions developed by the second proof of concept project for the water sector. We thought it would best to describe out project in one big presentation, and take questions at the end.
You’ll remember that 4 focus groups were established. Meetings were held in all 3 countries. Common workshop materials were used and standard templates to gather and record information.
We asked you:..
All countries have good access to climate change data info and to observed climate and river flow data, but would like:
Skilful seasonal hydrological forecasts, though most suggest these are unlikely to offer enough skill to change their need to plan for the worst.
In upland areas users find the current spatial resolution of climate change information inadequate. Other users want to understand hourly or sub-hourly changes, for example in rainfall intensity.
For water supply in the UK and for all uses in the Júcar basin, understanding how drought indices may change is important
This is not readily available from current generations of climate models.
Less sophisticated groups (eg. Norway LA group) makes little use of climate info at present
forcings
I will improved it. I can also show an indicator for CP
Main points for the demo:
- High resolution indicators for climate change over this century
- Design has focussed on user requirements for highly performant system to let them really look at and understand the information on display
- Maps show average of the indicator from the modelled ensemble (currently base on the draft model output we've produced)
- Describe list of indicators across precip, flow, soil moisture,etc.
- Show how users can include / exclude climate and hydrological models to understand their affect on ensemble average
- Show performance for varying time slice on display and for zooming
- Show how dynamic legend can allow variation in less contrasting areas to be accentuated, and also to create maps to pick up on specific features, e.g. change </> 10%. Is this a good idea?
- Highlight links to help information and in future user guidance
- Show graphing tool, focussing in on a specific area and indicator (we'll have to find one that doesn't look too weird)
- Demonstrate download functionality, and explain how this comes from an API for the indicators
Main points for the demo:
- High resolution indicators for climate change over this century
- Design has focussed on user requirements for highly performant system to let them really look at and understand the information on display
- Maps show average of the indicator from the modelled ensemble (currently base on the draft model output we've produced)
- Describe list of indicators across precip, flow, soil moisture,etc.
- Show how users can include / exclude climate and hydrological models to understand their affect on ensemble average
- Show performance for varying time slice on display and for zooming
- Show how dynamic legend can allow variation in less contrasting areas to be accentuated, and also to create maps to pick up on specific features, e.g. change </> 10%. Is this a good idea?
- Highlight links to help information and in future user guidance
- Show graphing tool, focussing in on a specific area and indicator (we'll have to find one that doesn't look too weird)
- Demonstrate download functionality, and explain how this comes from an API for the indicators
The vision of EDgE is to break down barriers for users of a range of different backgrounds to access and understand hydrological predictions. This is delivered through a fully integrative project identifying who are the users and what is their need, developing state-of-the-art hydrological solutions, and building a fast and easy to use web interface application.
At the core of EDgE is the iterative co-design and testing of products and services with the users.