Transcript Presentation Title, Arial Regular 29pt Sub title, Arial

Data for adaptation decision-making
Climate Adaptation National Research Flagship
Mark Stafford Smith, Science Director
Climate Adaptation Flagship
GEOSS/IPCC Workshop, Geneva, 1 Feb 2011
Data for adaptation decision-making
• Where I’m coming from
• 2010 Climate Adaptation Futures conference,
Australia
• Lessons from global desertification data needs?
CSIRO’s Climate Adaptation Flagship
Goal: Equip Australia with practical and
effective adaptation options to climate
change and variability and in doing so
create $3 billion per annum in net
benefits by 2030
Ca. 160 full time equivalent
researchers (~300 individuals),
established 2008
Engaged with policy, industry
and community decision-makers
in Australia and Asia-Pacific
Climate Adaptation Futures conference
• Gold Coast, Australia, Jun-Jul 2010
• ~1000 attendees, >30% overseas
• http://www.nccarf.edu.au/conference2010/
• 24 parallel sessions
Scenarios of the future for adaptation
Adapting agriculture to climate change
Indigenous vulnerabilities and adaptation
The economics and costs of adaptation
Coasts, deltas and small islands
Adaptation and the community
Climate Information for users
National and sub-national case studies of adaptn
The interface of adaptation and mitigation
Human security, social and equity issues
Risk communication and behavioural change
Water sector adaptation: innovations
Research meets business and industry
Impacts and adaptation in the tropics
Engineering and technology solutions for adaptn
Communication of information for adaptation
Ecosystems
Constructing and enabling local knowledge
National and international adaptation activities
Climate extremes and disaster management
New concepts in adaptation
Adaptation and development
Adapting to climate change in cities
Public health adaptn to variability and change
Climate Adaptation Futures: some lessons
• Some sessions dealt with data needs
• Climate services for early warning systems
• Often about communication of data in the right format
• Simplified datasets for many users, overload on local
government decision-makers, handling uncertainty
• “reframing information for risk management and away
from dependency on climate uncertainty” UKCIP
• “rapid assessment of climate uncertainty in evaluating [ag]
adaptation options”
• Emphasis on adaptation at multiple levels
• Also on measuring the effectiveness of adaptation
Mostly supporting decisions, not providing data
• In development, disasters, agriculture, water, cities,
health, engineering, business, human security, coasts
and small islands, policy, indigenous communities, etc
Some implications for adaptation datasets
• “Most of our current investment is [still] in defining the
problem, not in finding the solutions. A decisions and
outcomes focus is needed. ”
• A focus on use means dealing with:
• Multiple scales – geographic, governance, institutions,
industry sectors, etc
• Multiple levels in all of them – local, provincial, national,
supra-national, global
• Measuring the amount and effectiveness of action
• “Knowledge that is coproduced is more likely to be used”
Parallels in desertification
• 30 years of debate on desertification
• Plenty of political aspects though <<$$ than adaptation
• Persistent uncertainty even in how much (17-70% of
drylands desertified in late 1990s!), leading to policy
paralysis, loss of confidence in donors for action, etc
• Key gulf:
• Top-down universal indicators ~readily obtained
• But lacking credibility locally so not supporting action
• e.g. cover can increase or decrease with degradation
• Focused on biophysical or aggregated social indicators
• Bottom-up schemes engaging communities
• Locally credible indicators engendering local action
• But generally impossible to aggregate up
• measuring different things x different places (for good reason)
Purposes of monitoring data (my view!)
1. Determine where/how to invest resources (state)
• …from local decisions of a household or farmer to
investments of nations and the global community
2. Determine whether past such investments have
been successful (detect trend, signal from ‘noise’)
• …and change them if not
3. Understand cause and effect (causation)
• …to improve conceptual models driving investment,
the monitoring system or even the whole institutional
set-up around the monitoring (‘triple loop learning’).
• Science helps #1 and #2 but mainly through #3.
• Scientists tend to design monitoring systems for #3….
• Decision-makers actually need #1 and #2
Towards a ‘Global Drylands
Observing System’ - 1
• Need to sort out the clients for the data
– International, national, sub-national levels
– Different regions will care about different measures
• Some consistent meta-themes, other sensitive to locale
• On-ground measures legitimately differ by system
• Need to combine remote sensing (etc) and local
ground data for credible measures of change
– Tracking statistically significant change is much
harder than assessing state
• But needed to determine whether investments are
working, & to contribute to adaptive decision-making
Verstraete et al (2009) Frontiers in EcolEnvir 7: 421-8
Towards a ‘Global Drylands
Observing System’ - 2
• Need the right (multi-scaled) governance of the
system to be effective, owned, credible
– To sustain valuing of results and consequent
investment in collecting data
• All suggests a nested system:
– Nested clients, purposes (some generic data)
– Nested measures (mostly generic themes but
different indicators, able to be logically collated
upscale)
– Nested governance
– Nested, iterative development – can’t do it all at once!
Bastin et al (2009.) ACRIS. The Rangeland Journal 31: 111-125
Verstraete et al (2009) Frontiers in EcolEnvir 7: 421-8
Verstraete et al (2009): Fig. 3
Nested monitoring of human
and environmental slow
variables, chosen so local data
systematically contributes to
broader scale data, with remote
sensing providing context at
broader scales
Dryland countries
experiencing major
regional syndromes
at global scale
Dryland provinces
within countries
with different
trajectories
Dryland areas
where uses
prioritise different
ecosystem services
within the country’s
general trajectory
Developing dryland country
with increasing population
Developed dryland
country with land
abandonment
Conservation /
amenity province
Mining
province
Mixed subsistence
agriculture province
Water supply
catchment with
grazing
Area dominated by
smallholder grazing
and cropping
Data aimed at primary
dryland syndromes, e.g.
population, poverty, market
orientation, access to finance,
health, food reserves in
developing country, age of
managers, NRM investment,
pests and weeds, indigenous
minority access in developed
country
Indicator tailored to
provinces’ trajectories, eg.
population density, food and
water per head, net agricultural
productivity per unit area in
agricultural region, endangered
species, weed invasions, fire
regimes, tourist income in
amenity region
Data on locally important
ecosystem services, e.g.
pasture productivity, soil
nitrogen, household poverty in
agricultural area, involvement of
women; water quality, pasture
cover in catchment. Measures
suited to different ecosystems
Scaling up through major themes
e.g. forage; management responses; governance
capacity; household economics
Global NPP
datasets
State/trend of vegetation
for primary production
Palatable
perennial forage
Palatable
shrub cover
Crop weed
invasions
Perennial
grass cover
Household level adaptive
capacity for drought
National
$$
Household food
capital
# grazing
animals
Remittances
from outside
Stored grain
stock
Is it happening at UNCCD?
• Many attempts to do global assessments
• Mostly fail to account for different local causation and
hence can’t tell what management/investments are
needed or responsible for change
• But provide vital context if interpreted correctly
• Various more locally-sensitive systems
• Aim to support action on the ground/within the nation
• e.g. LADA (Land Degradation Assessment in Drylands)
http://www.fao.org/nr/lada/
• Papers emerging with a new architecture
• Not universally accepted (or even understood) yet
• But important lessons for adaptation
Bastin et al (2009.) ACRIS. The Rangeland Journal 31: 111-125
Verstraete et al (2009) Frontiers in Ecol.Envir 7: 421-8
Verstraete et al (2011) Land Deg. Dev. 21 in press
Australian Collaborative Rangelands
Information System - ACRIS
• Why a national system for reporting change?
• There are reasons for having nationally comparable information!
• Investment planning and evaluation
• SoE, international reporting
• Sustaining a sustainable image
• Who for?
• National, state-level, regional stakeholders (different needs)
• What’s the challenge?
• Spatial and temporal variability, and sparse resources
• Detecting change, then attributing it
1955
CSIRO. Insert presentation title
1992
Rangelands 2008 – Taking the Pulse
What we learned: some headlines - 4
Theme
Summary
Climate variability
• Seasonal quality: above-average in the north and north-west; variable in central
Australia; above average then dry in most of WA & SA shrublands; below
average then drought in eastern grasslands & mulga lands
Total grazing
pressure
• Mapped, including 20-40% by roos in S & E; feral densities still poorly tracked
• In some pastorally important bioregions,
domestic
stock remain
high despite
Landscape
function
provides
a
declining seasons
Landscape function
measure of the landscape’s
• WA, SA, NSW & NT: generallycapacity
positive signs
seasonal
quality. and
togiven
capture
rainfall
• Queensland: 6 of 11 bioregions - decreased landscape function
nutrients, the essential resources
for plant growth.
functional – non leaky
CSIRO. Insert presentation title
dysfunctional –leaky landscape
Hierarchical system of diverse sources of data
$$ from states and
federal govts
Sources of national
data (CSIRO,
Queensland, Federal
Govt) – rainfall, land
use, remote sensing,
dust, socio-economic
stats, etc,
Steering Committee
Management Unit
to collate, interpret
and synthesise,
and return to users
Users:
Federal and
State
Governments,
regional ‘NRM
Bodies’
Sources of on-ground monitoring data
+ some regional socio-economic & R/S data
+ regional interpretative ‘local’ knowledge
Western
Australia
CSIRO. Insert presentation title
Northern
Territory
South
Australia
Queensland
New South
Wales
Implications: adaptation info system architecture
• Think multiple levels in multiple scales
• Hierarchically nested structure with data themes
• Data themes relevant to decision-makers
• Focus on indicators that can be acted upon (decisions)
• Support with indicators of causation (science; =PSR+)
• Flexible indicators locally, within agreed themes
• Design processes to engender ownership at each level
• Design and resource mechanisms for upscaling
• At least at national and global levels
• Meta-analysis for consistency at that level in the theme
• Bring in global datasets for context, extrapolation, truthing
• Expect iterative system development
• Can’t be done in a day, but worth it eventually!
Climate Adaptation Flagship
Director: Andrew Ash
[+61] 07 3214 2234 / [email protected]
Science Director: Mark Stafford Smith
[+61] 0408 852 082 / [email protected]
Climate Adaptation Flagship