Climate Change, models and uncertainty

1
Climate Change, models and uncertainty
2
Confusing terms?

• Climate Change
• IPCC Climate change may be due to natural
  internal processes or persistent anthropogenic
  changes
• UNFCCC A change in climate which is attributed
  directly or indirectly to human activity and
  which is in addition to natural climate
  variability

3
Weather or Climate?

• Weather The local state of the atmosphere on a
  day-to-day basis. Wind, rain, cold, hot
• Climate Long-term average of temperature,
  precipitation, extremes and other variables.
  Different scales and different periods.
• Climate sets the background conditions for
  weather and will affect the frequency of certain
  weather events.

4
Climate Change or just Climate Variability?

• Anthropogenic climate change is a significant and
  persistent change in the average conditions or
  extremes of a region.
• But the climate system contains natural cycles
  and variability seasonal, inter-annual (e.g.
  ENSO, Indian Dipole), multi-decadal (e.g. North
  Atlantic Oscillation) or longer!

5

• El Niño Global effects
• Every 3-7 years
• Warm and dry in Asia
• Drought

6
Variability happens

• A particularly hot/wet year in a particular place
  isnt necessarily evidence of climate change.
• A particularly cold winter isnt evidence that
  climate change isnt happening
• What is, however, is changes to the average over
  time, and all of these observations taken
  together and across the world.
• Natural climate processes cannot explain the
  global changes that we have observed in the last
  50 years.

7
Current trends and climate projections

• Changes experienced now may be due to climate
  change but may also be part of natural cycles
• Just because it is getting wetter in a particular
  location, doesnt necessarily mean this will
  continue
• If the observed trend matches the climate
  projection, we can be more confident of the
  changes to expect.

8
Context before projection!

• It is (really!) important to understand the
  current climate context before jumping into
  projections
• What are the historical conditions?
• Are there any trends?
• Why are things the way they are?
• Without the context it is difficult to place the
  projections. . .

9
Climate Projections and uncertainty

• We do not know what the future holds. . .but this
  doesnt mean we cant do anything

10
Background

• A model is a representation of the real world, it
  is not an exact copy
• Projections vs Predictions.
• We do not know which model is best fit to
  historical climate is not necessarily an
  indicator of quality of projection
• Climate is a complex system there are a range
  of plausible future states

11
Types of Model pros and cons

• Global models (smallest resolution 200km)
• Regional Models (25-50km)
• Statistically Downscaled models (station level)

12
Global models

• Coarse resolution
• Good at capturing circulation changes
• Good for ideas of global change
• Skill decreases at smaller scales, which is where
  we need information for adaptation!
• Topography and feedbacks are problems

13
Where is this?
14
Regional Models

• Better resolution (up to 25-50km2)
• Good at a national/water basin level.
• Can model feedbacks.
• Provide information for whole region
• Computer intense so difficult to run lots of
  models hard to look at a range of change.
• Topography a problem

15

• Statistical Downscaling from large-scale to the
  local.

16
Statistically downscaled models

• Provide projections at the level of an adaptation
  activity.
• Good at topography and islands
• Computer light, so can look at many different
  projections and assess the range.
• Cant provide data where there is not a good
  historical record (at least 10 years continuous)
• Cant capture feedback processes.

17
Station level projections

• Over 300 in Asia
• More expected through the platform

18
Confidence in projections

• Uncertainty emissions, models, processes, (e.g.
  El Niño), land use. Variability will continue.
• Projections should be treated as indicative, not
  predictions.
• More confident on certain variables than others
• The more models and data sources agree, the more
  confident we can be. Including trends!

19
What we cant say

• A specific year will be bad
• We are certain of our projections
• Exact timing of changes
• Much detail about variability
• Much about the next 25 years

20
So. . .

• Multiple sources of information give us a more
  rounded view
• We cant simply discard information from a model
  giving us a result we dont like
• An average can hide valuable information the
  extreme projection could be the right one
• In the real world decisions made under imperfect
  conditions all the time the key is what
  information we can use to help the decision.

21
Thank-you!

• Ben Smith, Stockholm Environment Institute
  bensmith.sei_at_gmail.com
• www.weadapt.org www.wikiadapt.org