Forecasting High Impact Weather:
the 2014 Glasgow Games & HIWeather
Brian Golding, Met Office
Co-Chair WWRP HIWeather project
© Crown copyright Met Office
Outline
•
Context
• Application of the UK regional ensemble to surface water
flood forecasting
• Surface water flood management during the Glasgow
Commonwealth Games of 2014
•
Background to the WMO HIWeather project
•
HIWeather Objectives
•
Opportunities for working with HIWeather
© Crown copyright Met Office
Context
• The UK regional convection-permitting NWP model now has a 1.5km grid
length. It can represent the scales of weather that are relevant to most hazards.
• The UK regional ensemble prediction system now has a 2.2km grid length. It
can quantify the probability of hazardous weather up to 36 hours ahead.
• The UK global ensemble prediction system now has a 33km grid length. It can
quantify the probability of weather systems that may produce hazardous weather
up to a week ahead.
•
Other leading weather services have similar capabilities.
• Governments & agencies around the world have implemented improved
disaster management procedures including preparedness, response and
recovery.
• Despite these advances, weather-related disasters continue to kill, displace
populations & damage property / infrastructure, while less severe weather events
place an increasing strain on society, especially in countries with fragile
economies and infrastructure.
© Crown copyright Met Office
MOGREPS
Global DA
Global deterministic forecast 17km
Global ETKF
Perturbations
MOGREPS-G ensemble forecast 33km
Boundary values
Initial state
Interpolation
MOGREPS-UK ensemble
forecast 2.2km
© Crown copyright Met Office
Probability of more than 40mm rain in 3 hrs
from ensemble forecast for 8th July 2011
3hr accumulations >40mm
cause local flooding
Plot % of members exceeding
this at each grid point
There is no pattern because
there are too few members
Observation
Key
X - >40mm/3hr
x - >30mm/3hr
© Crown copyright Met Office
x - >20mm/3hr
Box marks
25km from
Edinburgh
Probability of more than 40mm in 3 hrs
using 25km neighbourhood smoothing
Observation
Key
X - >40mm/3hr
x - >30mm/3hr
x - >20mm/3hr
© Crown copyright Met Office
London Olympics Trial: 5th July 2012
© Crown copyright Met Office
2012 Trial Overview
•24% of days had forecast
prob of intense rain >60%
& 33% of days had hourly
accumulations >16mm.
•When intense rain was
forecast with prob >60%, it
occurred on 84% of days.
•When intense rain was
forecast with prob <60%, it
occurred on 17% of
occasions.
•This overview says
nothing about timing or
location
© Crown copyright Met Office
Application to the July 2014
Commonwealth Games in Glasgow
2km Gridded Ensemble Rainfall:
6hr nowcast, 3-hourly
24hr forecast, 6-hourly
1km Gridded Ensemble Runoff
1km gridded ensemble
threshold exceedance
Probability of exceeding
impact thresholds,
somewhere in 1km squares,
sometime in defined periods
Impact of flooding on property,
roads, railways, utilities
2m Flood map for runoff threshold
© Crown copyright Met Office
Application to Glasgow:
10th August overview
Estimated probabilities of overall
surface water flood impact,
sometime in the 24 hours
© Crown copyright SEPA
HIWeather
research for resilience
WMO
OMM
HIWeather
A 10-year programme of the World Meteorological Organisation
The potential of advanced weather-related hazard
forecasting has been demonstrated. There is a huge
opportunity to protect lives and benefit communities, if
we can realise this potential across the world.
HIWeather will
WMO
OMM
“Promote Co-Operative International Research
to achieve a Dramatic Increase
in Resilience to High Impact Weather, worldwide,
through Improving Forecasts for timescales of minutes to two weeks
and Enhancing their Communication & Utility
in Social, Economic & Environmental Applications”
Scope defined by a set of hazards...
Urban Flood:
Reducing mortality, morbidity, damage and disruption from flood
inundation by intense rain, out-of-bank river flow, coastal wave &
surge overtopping and from consequent urban landslides.
© www.thamai.net
Disruptive Winter Weather:
Reducing mortality, morbidity, damage and disruption from snow,
ice and fog to transport, power & communications infrastructure.
Wildfire:
© Tommy Hindley/ Professional Sport
Reducing mortality, morbidity, damage and disruption from
wildfires & their smoke.
MALAO@pictzz.blogspot.com
Urban Heat Waves & Air Pollution:
Reducing mortality, morbidity and disruption from extreme heat
& pollution in the megacities of the developing and newly
developed world.
© www.energydigital.com
Extreme Local Wind:
WMO
OMM
Reducing mortality, morbidity, damage and disruption from wind &
wind blown debris in tropical & extra-tropical cyclones, downslope
windstorms & convective storms, including tornadoes.
© Adrian Pearman/Caters
Rain
Snow
Temperature
Snow Melt
Run-off
…their impacts,
Drains
Pressure
Storm
surge
River
Wind
Ocean
Waves
Breach/ Overtopping
Landslide
Dirty water in the “wrong” place
Depth
Velocity
Dirt/Trash
Pollutants
Contamination
Buildings
damaged
Contents
destroyed
Transport
blocked
Occupants
displaced
Injury
WMO
OMM
WMO
OMM
Water/energy
interrupted
Business
interrupted
Food/fuel
shortage
Distress
Death
Drowning
Bankruptcy
Illness
Recovery
cost
Redundancy
…mitigation actions,
WMO
OMM
…the required information,
Pmsl
H500
???
WMO
OMM
20% prob of intense
rain within 30km
between 1-2pm
tomorrow
…define our five research pillars
WMO
OMM
• Understanding the processes & predictability of weather systems
that generate hazards
• Multi-scale forecasting of hazards using coupled numerical weather,
ocean, land surface, ice & air quality modelling, nowcasting, data
assimilation & post-processing systems.
• Forecasting the human impacts, exposure, vulnerability & risk of
hazards to people, buildings, businesses, infrastructure and the
environment using a variety of tools.
• Communicating hazard forecasts & warnings so as to reach
vulnerable communities and achieve responses from risk managers
and the public that increase resilience.
• Evaluating hazardous weather, impact & risk forecasts, alerts &
warnings and the resulting responses with user-relevant metrics.
• supported by eight cross-cutting activities
The HIWeather
Challenges
WMO
OMM
Predictability &
Processes
Initiation & evolution of hazardrelated weather systems
• Scale interactions & implications for
predictability at km-scale
• Error growth for hazardous weather
• Quasi-stationary weather systems
• Dynamic role of diabatic heating
• Processes specific to individual hazards
WMO
OMM
Multi-Scale
Forecasting
•
•
•
•
Multi-scale prediction of weather
hazards in coupled modelling
systems
Observing km-scale weather, hazards, impacts
Hazard nowcasting methods
Coupled km-scale data assimilation methods
Improved convective initiation, microphysics &
land surface interactions in km-scale models
• Coupled km-scale ensemble systems
• Products tailored to user needs, inc uncertainty
WMO
OMM
Vulnerability
& Risk
•
•
•
•
•
•
WMO
OMM
Hazard impacts on individuals,
communities & businesses, their
vulnerability & risk
Growing impacts research capacity
Synthesising previous fragmented work
Using social media
Identifying & characterising vulnerability
Representing dynamic vulnerability
Understanding counter-intuitive responses
Communication
•
•
•
•
More effective responses to
forecasts through better
communication of hazard risk
warnings
Effectiveness of communication methods
Effective use of social media
Reasons for lack of trust - how to build trust
Good practice in communicating forecasts &
warnings
• Growing research capacity
WMO
OMM
User-Oriented
Evaluation
•
•
•
•
Measure skill and value of
forecasts & warnings at all stages
of production to focus research
in weak areas & support users in
developing responses
Verification of hazards, impacts & responses
Information loss through the production chain
Information needed for user response strategies
Economic value of forecasting & warning
services
• Growing research capacity
WMO
OMM
RDPs & FDPs to test key generic
Field Campaigns &
hypotheses in specific local
Demonstrations
environments
WMO
OMM
• Build local capability & capacity
• Integrate research
• NAWDEX/DOWNSTREAM: diabatic growth of
waveguide disturbances & downstream impacts
• LVB-HyNEWS: warning of dangerous nocturnal
convection over Lake Victoria
• ALERT.AR/RELAMPAGO: urban flood impacts in
the La Plata basin of South America
• Other experiments & testbeds
Planned Collaboration Activities
•
•
•
•
2015 Review of wind hazard
2015 Workshop on fire hazard modelling
2015 Report on good practice in verifying hazards
2015 Combined workshops on:
– Monitoring & forecasting weather related impacts & responses,
including use of social media;
– factors influencing communication & application of weather information
in decision making;
– How to measure what makes a successful forecast
WMO
OMM
• 2015 Initiate catalogue of forecast/warning successes & failures
• 2015 Agree archive formats & metadata for HIWeather RDPs
• 2015 Planning for T-NAWDEX, LVB-HyNEWS & RELAMPAGO,
focusing on local participation & response
• 2016 workshop on km-scale Nowcasting & Data Assimilation
• 2017 workshop on km-scale Ensemble hazard prediction
taking part…
WMO
OMM
• Orientation of planned research programmes to
HIWeather priorities
• Sharing of expertise through international
reviews & intercomparisons
• Inviting international participation in national
projects
• Participating in international projects with
relevant objectives
• A coordinated “China HIWeather programme”?
HIWeather
Predictability &
Processes
WMO
OMM
CommuVulnerMultiability & nication
scale
Risk
Forecasts
Benefits in Operational Forecasting
Design of observing strategies
Uncertainty
Field campaigns & demonstrations
Knowledge Transfer
Verification
Impact Forecasting
Data Management & Archiving
Evaluation