Anthropogenic Land Cover Change Experiments in the CCSM

1
Anthropogenic Land Cover Change Experiments in
the CCSM Participants NCAR
University of Kansas Gordon
Bonan Johannes Feddema Linda Mearns Trish
Jackson Keith Oleson Pei-Ling Lin Jerry
Meehl John Bauer Warren Washington Doug
Nychka Lawrence Buja
This research is supported by the Office of
Science (BER), U.S. Department of Energy,
Cooperative Agreement No. DE-FC02-97ER62402, by
the National Science Foundation grant numbers
ATM-0107404, and ATM-0413540, the NCAR Weather
and Climate Impact Assessment Science Initiative,
and the University of Kansas, Center for
Research.
2
Overview

• How are the experiments set up and developed?
• Equilibrium vs transient experiments
• Uncertainty about land cover and its impacts
• Multiple land cover forcings (e.g. agriculture vs
  grazing)
• Dealing with multiple climate forcings
• Land cover change alongside other forcing
• Statistical Significance in this framework
• 3. Separating out signals and feedbacks between
  forcings
• Complex and non-linear responses to the same
  forcing
• Optimizing experimental design

3

• How are the experiments set up and developed?
• Equilibrium vs transient experiments
• Uncertainty about land cover and its impacts
• Multiple land cover forcings (e.g. agriculture vs
  grazing)
• Dealing with multiple climate forcings
• Land cover change alongside other forcing
• Statistical Significance in this framework
• 3. Separating out signals and feedbacks between
  forcings
• Complex and non-linear responses to the same
  forcing
• Optimizing experimental design

4

• Equilibrium Experiments
• Hold all conditions equal and allow the model to
  run to equilibrium
• Compare a control and experiment where one or
  more boundary conditions are changed
• Typically compare 10-50 year time slices after
  equilibrium is reached
• Transient Experiments
• Starting from some equilibrium state the model
  runs through time as forcings change (e.g.
  increasing CO2 through time)
• Compare a control and experiment integrated over
  one or more time periods during the simulation
• Model usually does not reach equilibrium so
  equivalent time slices of 10-30 years are
  compared

5

• How are the experiments set up and developed?
• Equilibrium vs transient experiments
• Uncertainty about land cover and its impacts
• Multiple land cover forcings (e.g. agriculture
  vs. grazing)
• Dealing with multiple climate forcings
• Land cover change alongside other forcing
• Statistical Significance in this framework
• 3. Separating out signals and feedbacks between
  forcings
• Complex and non-linear responses to the same
  forcing
• Optimizing experimental design

6
PCM Uncertainty/Historical Equilibrium Land Cover
Simulations
7
PCM Uncertainty/Historical Equilibrium Land Cover
Simulations

• PRESENT DAY UNCERTAINTY
• Arctic albedo
• Amazon latent heat flux
• Australia albedo
• HISTORICAL CHANGE
• Climate difference from land cover classification
  is as large as the climate difference from land
  cover change
• Primarily shift to agriculture

Question How do we deal with input uncertainty?
8
Comparison of Agriculture land classes from 3
satellite products 10 degree tile over East Africa
IGBP
MODIS
GLC2000
Agreement
9
(No Transcript)
10
(No Transcript)
11

• How are the experiments set up and developed?
• Equilibrium vs transient experiments
• Uncertainty about land cover and its impacts
• Multiple land cover forcings (e.g. agriculture
  vs. grazing)
• Dealing with multiple climate forcings
• Land cover change alongside other forcing
• Statistical Significance in this framework
• 3. Separating out signals and feedbacks between
  forcings
• Complex and non-linear responses to the same
  forcing
• Optimizing experimental design

12
(No Transcript)
13
Question How do to isolate the impacts of
multiple forcings?
14

• How are the experiments set up and developed?
• Equilibrium vs transient experiments
• Uncertainty about land cover and its impacts
• Multiple land cover forcings (e.g. agriculture vs
  grazing)
• Dealing with multiple climate forcings
• Land cover change alongside other forcing
• Statistical Significance in this framework
• 3. Separating out signals and feedbacks between
  forcings
• Complex and non-linear responses to the same
  forcing
• Optimizing experimental design

15
IPCC and human impacts
16
(No Transcript)
17
PCM Future SRES A2 Transient Simulations
The A2 ScenarioThe A2 storyline and scenario
family describes a very heterogeneous world. The
underlying theme is self-reliance and
preservation of local identities. Fertility
patterns across regions converge very slowly,
which results in continuously increasing global
population. Economic development is primarily
regionally oriented and per capita economic
growth and technological change are more
fragmented and slower than in other storylines.
1970
By 2100, expansion of agricultural land in North
America, South America, Africa, and Southeast
Asia Question What is the land use forcing
relative to other natural and anthropogenic
forcings?
2100
18
(No Transcript)
19
Future IPCC SRES Scenarios for PCM
20
(No Transcript)
21
PCM Future SRES A2 Transient Simulations
Projected change by 2100 Annual Average
Temperature
GHG only
GHG LC
LC contribution (GHGLC) GHG only
Note Shift in Divergent Scale
22
B1
A2
2050
2100
Question How to best identify land cover
impacts in a multi-forcing run?
23
PCM Future SRES A2 Transient Simulations
Relative impact of land cover forcing compared to
GHG effects
On average LULC contributes 11 of 2100 forcing
compared to GHG-only forcing. However, this is
highly regional and offsetting with respect to
global average temperature
Question What is a good measure to compare
different forcings? (radiative forcing)
Given that we have spatial and temporal
results that can be offsetting.
24
B1
A2
2050
2100
Question How to best isolate direct impacts
from teleconnections?
25

• How are the experiments set up and developed?
• Equilibrium vs transient experiments
• Uncertainty about land cover and its impacts
• Multiple land cover forcings (e.g. agriculture vs
  grazing)
• Dealing with multiple climate forcings
• Land cover change alongside other forcing
• Statistical Significance in this framework
• 3. Separating out signals and feedbacks between
  forcings
• Complex and non-linear responses to the same
  forcing
• Optimizing experimental design

26
PCM Historical Comparison
Annual
Change in temperature Shading
standard t test 0.95 confidence level Contour
bootstrap 0.95 confidence level
DJF
Bootstrap confidence test shows strong summer
hemisphere signal in sub-tropics Many of the
areas are over land cover change locations
JJA
Question How to best /most efficiently evaluate
confidence?
27

• How are the experiments set up and developed?
• Equilibrium vs transient experiments
• Uncertainty about land cover and its impacts
• Multiple land cover forcings (e.g. agriculture vs
  grazing)
• Dealing with multiple climate forcings
• Land cover change alongside other forcing
• Statistical Significance in this framework
• 3. Separating out signals and feedbacks between
  forcings
• Complex and non-linear responses to the same
  forcing
• Optimizing experimental design

28
PCM Present Day Comparison Image - LSM
Seasonal Change in Albedo
Strong winter/spring albedo change in the
Northern Hemisphere translates to spring/summer
net radiation change due to solar seasonality
Seasonal Change in Net Radiation
Question How to best detect seasonally varying
responses?
29
PCM Historical Comparison
Albedo
Incident radiation
Albedo changes, but cloud cover also plays a
major role Local feedbacks or changes in
circulation?
DJF JJA Cloud cover change
Question How to identify feedbacks, and can we
have confidence in these signals?
30
Future Scenario All grid cells that have been
converted from tropical rain forest to
agricultural change The Amazon response is very
different from SE Asia response in part because
of the large scale circulation conditions
Question How to best detect spatial variability
in specific responses?
31
Variability in Simulated Heat Island caused by
Climate and Rural Environment

• Atmospheric forcing from CAM (offline model)
• Default city with H/W0.5,,3.0
• Rural environment from CLM Surface Data

The urban model has very distinctly different
responses depending on weather conditions and on
surrounding vegetation types
Question How to organize output to best analyze
the variability in responses?
32

• How are the experiments set up and developed?
• Equilibrium vs transient experiments
• Uncertainty about land cover and its impacts
• Multiple land cover forcings (e.g. agriculture vs
  grazing)
• Dealing with multiple climate forcings
• Land cover change alongside other forcing
• Statistical Significance in this framework
• 3. Separating out signals and feedbacks between
  forcings
• Complex and non-linear responses to the same
  forcing
• Optimizing experimental design

33

• Currently simulations are run independently for
  all possible forcings then in combination.
• Question Knowing there are non linear feedbacks,
  is there a way to reduce the number of runs with
  combinations of experiments to
• Extract the individual climate impacts of each
  forcing
• Understand the non linear interactions between
  the forcings