ICTP Regional Climate Model 1. Recent Developments 2. Applications

1
ICTP Regional Climate Model1. Recent
Developments2. Applications
Jeremy Pal Abdus Salam International Centre for
Theoretical Physics Trieste, Italy Contributors
Xunqiang Bi, Elfatih Eltahir, Raquel Francisco,
Xuejie Gao, Filippo Giorgi
2
Outline of Presentation

• Part I Latest Developments
• SUBEX Large-scale Cloud and Precipitation Scheme.
• Sub-grid Topography and Landuse Parameterization.
• Part II Model Applications
• The Effects of Soil Moisture on the Simulation of
  Surface Climate and Hydrology.
• Recent European Summer Climate Trends and Events
  Are Climate Change Projections Consistent with
  Observations?

3
Summary of RegCM3 Core

• Dynamics
• MM5 Hydrostatic (Grell et al 1994)
• Non-hydrostatic (in the works)
• Radiation
• CCM3 (Kiehl 1996)
• Large-Scale Clouds Precipitation
• SUBEX (Pal et al 2000)
• Cumulus convection
• Grell (1993) AS74 FC80 closure
• Anthes-Kuo (1977)
• Betts-Miller (1993)
• Emanuel (1991)
• Boundary Layer
• Holtslag (1990)

• Tracers/Aerosols
• Qian et al (2001) Solmon
• Land Surface
• BATS (Dickinson et al 1993)
• SUB-BATS (Giorgi et al 2003)
• CLM0 (Dai et al 2003)
• Ocean Fluxes
• Zeng et al (1998)
• BATS (Dickinson et al. 1986)
• Computations
• Parallel Code (Yeh Gao)
• User friendly
• Multiple platforms

(Giorgi et al 1993ab, Pal et al 2005ab)
4
Part I Latest Developments

• Tracer/Aerosol Scheme Qian et al (2001) Solmon
• Zeng et al (1998) Ocean Flux Model
• Convection Grell -- Fritsch Chappell (1980)
  Closure Betts-Miller (1993) Emanuel (1991)
• SUBEX Large-scale Cloud and Precipitation Scheme
  (Pal et al 2000)
• Sub-grid Topography and Landuse Parameterization
  (Giorgi, Francisco, Pal 2003)
• CLM0 (Dai et al 2003)
• Parallel Code Ye, Gao, Bi
• Input/Output improvements

5
Part I Latest Developments

• SUBEX Large-scale Cloud and Precipitation Scheme
• (Pal et al 2000)
• Sub-grid Topography and Landuse Parameterization
• Giorgi et al (2003) Elguindi, Pal, Nagarajan

6
Importance of Clouds

• Correctly simulating the surface solar radiation
  is crucial in simulating the hydrologic cycle.
• The response of climate to changes in CO2, soil
  moisture, vegetation, etc., is largely dependent
  on cloud feedbacks.
• E.g. The IPCC report indicates that the
  representation of cloud characteristics accounts
  for a large portion of the uncertainty in climate
  change predictions.
• In most climate models clouds and precipitation
  are represented in two forms
• Resolvable (large-scale)
• Unresolvable (convective)

7
Large-Scale Clouds and Precipitation

• The model had an on-off tendency with clouds.
• No middle ground
• Clouds only form when the gridcell is
  supersaturated.
• Clouds often exist at relative humidities below
  one.
• Fractional cloud cover (FC) is 80 when cloud
  water exists and 0 otherwise.
• FC varies between 0 and 100.
• Raindrop accretion evaporation are not
  accounted for.

8
SUBEX Methodology

• Accounts for sub-grid processes (Sundqvist et al
  1989)
• Gridcell is divided into cloud and cloud-free
  fraction
• Cloud fraction is based on gridcell averaged
  relative humidity
• In-cloud values as opposed to gridcell average
  values are used.
• Cloud liquid water threshold at which
  precipitation forms is based on Observations
  (Gulteppe 1997).
• Accretion of cloud droplets by precipitation
  (Beheng 1994).
• Precipitation evaporation (Sundqvist et al 1989).

Cloud Fraction, FC
Relative Humidity, f
9
SUBEX Incident Surface Solar (NASA-SRB)

• Old Model vs. Observations

• New Model vs. Observations

10
SUBEX Mean Surface Temperature (USHCN)

• New Model vs. Observations

• Old Model vs. Observations

11
SUBEX Precipitation (USHCN)

• Old Model vs. Observations

• New Model vs. Observations

12
SUBEXJune July 1993 Flood
USHCN Observations
RegCM Old Model
13
Convective Closure Assumptions June July 1993
Flood
USHCN Observations
RegCM AS74
RegCM FC80
14
SUBEX Summary of Results

• SUBEX provides a more accurate representation of
  the fields describing the energy and water
  budgets.
• The improvements are seen in mean conditions as
  well as the variability at the daily to
  interannual scales
• This suggests that SUBEX improves the model
  sensitivity which is critical to climate change
  and process studies.

15
Part I Latest Developments

• SUBEX Large-scale Cloud and Precipitation Scheme
• (Pal et al 2000)
• Sub-grid Topography and Landuse Parameterization
• (Giorgi, Francisco, Pal 2003)

16
Subgrid Topography and Landuse Scheme

• Land surfaces are characterized by pronounced
  spatial heterogeneity (100s of meters).
• Topography and landuse exert a strong forcing on
  atmospheric circulations and land-atmosphere
  exchanges.
• Current climate models cannot capture the full
  range of scales, thus intermediate techniques can
  be used.

17
Biosphere-Atmosphere Transfer Scheme BATS
(Dickinson et al 1993)

• One canopy layer
• Stomatal conductance (Jarvis-type) model
• One snow layer
• 3 soil layers
• Soil T Force-restore
• Soil moisture Diffusive/gravitational

18
SUB-BATS General Methodology

• Define a regular fine scale sub-grid for each
  coarse scale model grid-box.
• Landuse, topography, and soil texture are
  characterized on the fine grid.
• Disaggregate climatic fields from the coarse grid
  to the fine grid (e.g. temperature, water vapor).
• Based on the elevation differences between coarse
  grid and fine grid.
• For example, Temperature
• Perform BATS surface physics computations on the
  fine grid.
• Reaggregate the surface fields from the fine grid
  to the coarse grid.

19
NumericalExperiments

• Simulation period
• 1 Oct 1994 to 1 Sept 1995
• Surface computations performed on subgrid.
• CTL
• 60-km no subgrid cells
• 10,000 gridpoints
• EXP10
• 10-km 36 subgrid cells
• 350,000 gridpoints
• A high resolution run at 10-km would take 200
  times longer.
• 36 times more points
• 6 times lower timestep
• 1.3 times longer here.

20
SUB-BATS Results Temperature
OBS (CRU)
CTL
WINTER (DJF)
OBS (CRU)
CTL
SUMMER (JJA)
21
SUB-BATS Results Precipitation
OBS (CRU)
CTL
EXP10
WINTER (DJF)
OBS (CRU)
CTL
EXP10
SUMMER (JJA)
22
SUB-BATS Results Snow
STATION OBS
CTL
EXP10
WINTER (DJF)
STATION OBS
CTL
EXP10
SPRING (MAM)
23
In the works

• Implement parameterization of subgrid scale
  effects on the formation of precipitation (both
  large-scale and convective).
• Apply disaggregation techniques for other
  variables (e.g. precipitation, radiation)

24
Precipitation over East Asia
CRU Observations
RegCM3
September 1994 thru August 1995
25
Part II Model Applications

• Recent European Summer Climate Trends and Events
  Are Climate Change Projections Consistent with
  Observations?
• (Pal, Giorgi, Bi GRL 2004)
• The Effects of Soil Moisture on the Simulation of
  Surface Climate and Hydrology
• (Pal Eltahir 2000, 2002, 2003)

26
Part II Model Applications

• Recent European Summer Climate Trends and Events
  Are Climate Change Projections Consistent with
  Observations?
• (Pal, Giorgi, Bi GRL 2004)
• The Effects of Soil Moisture on the Simulation of
  Surface Climate and Hydrology
• (Pal Eltahir 2000, 2002, 2003)

27
Recent European Extreme Summers

• The western European summer drought of 2003 is
  considered one of the severest on record.
• 20,000 heat related casualties in Western Europe.
• Worst harvest since World War II.
• In contrast, during 2002, many European countries
  experienced one of their wettest summers on
  record.
• Weather systems brought widespread heavy rainfall
  to central Europe, causing severe flooding along
  all the major rivers.
• The Elbe River reached its highest level in over
  500 years of record
• Both of these contrasting events resulted in
  severe damages and losses.
• This study addresses whether these seemingly
  opposites in extremes are consistent the current
  climate change projections.

28
Model Configuration

• ICTP RegCM3
• 50 km
• 121 x 100 x 14
• HadAMH SST, GHG Sulfate
• Aerosol effects (direct indirect)
• Simulations
• Reference run
• 1961-1990
• A2 B2 Scenario runs
• 2071-2100

• Mediterranean Focused

29
SCENARIOS
CO2 Emissions (Gt C)
CO2 Concentrations (ppm)
A2
A2
B2
B2
30
Surface Air Temperature (1961-1990)
Observations, DJF
RegCM3, DJF
Observations, JJA
RegCM3, JJA
31
Precipitation (1961-1990)
Observations, DJF
RegCM3, DJF
Observations, JJA
RegCM3, JJA
32
Changes in Summer500 hPa Geopotential Heights
NCEP Reanalysis (1976-2000) minus (1951-1975)
(? meters)
33
Changes in Summer Temperature
B2-Reference (2071-2100) minus (1961-1990)
CRU Observations (1976-2000) minus (1951-1975)
(C)
(C)
34
Changes in Summer Precipitation
B2-Reference (2071-2100) minus (1961-1990)
CRU Observations (1976-2000) minus (1951-1975)
( change)
( change)
35
Changes in Summer TemperatureB2-Reference
Interannual Variability (2071-2100) minus
(1961-1990)
Mean Surface (2071-2100) minus (1961-1990)
(C)
(C)
36
Changes in Summer ExtremesB2-Reference
Max 5-Day Precipitation (2071-2100) minus
(1961-1990)
Dry Spell Length (2071-2100) minus (1961-1990)
( change)
(? Days)
37
Precipitation Distribution(Hypothetical)
38
Summer Climate ChangeSummary of Results

• Projected changes in mean summer precipitation
  and large-scale dynamics over Europe are broadly
  consistent with the observed changes in the last
  25 years.
• Even though the projections suggest a tendency
  for drier summers in Western Europe, the
  incidence of severe precipitation tends to
  increase.
• Perhaps the intensified hydrological cycle from
  the longer hotter summers
• increases the likelihood of drought.
• Increases the atmosphere's moisture capacity and
  therefore the likelihood of flooding.

39
Part II Model Applications

• Recent European Summer Climate Trends and Events
  Are Climate Change Projections Consistent with
  Observations?
• (Pal, Giorgi, Bi In preparation)
• The Effects of Soil Moisture on the Simulation of
  Surface Climate and Hydrology
• (Pal Eltahir 2000, 2002, 2003)

40
Rainfall Anomalies (mm/d)
Rainfall Anomalies (mm/d)
June July 1993
May June 1988
1988 North American Drought Driest warmest
since 1936 10,000 deaths 30 billion in
Agricultural Damage
1993 Midwest Flood Record high rainfall Thousands
homeless 48 deaths 15-20 billion in Damage
41
Numerical Experiments

• Month Long Simulations
• July 1986, 1987, 1988, 1989, 1990 1993
• Soil Saturation Initializations
• Observed (CTL)
• Climatology (CLM)
• Fixed Patch
• 25MW
• 25GC
• 50SW

• Domain
• 129x80x14 at 55.6-km

42
25MW Fixed Patch Experiment Initial Root Zone
Soil Moisture
Midwest 25MW
Fixed Soil Moisture (25)
25
Interactive Soil Moisture (CTL)
43
Boundary Layer Height
Net Radiation
25MW-CTL
25MW-CTL
Moist Static Energy

• Decrease in the energy per unit depth of boundary
  layer via radiative effects
• Should decrease the likelihood and magnitude of
  rainfall of the region of the anomaly

25MW-CTL
44

• Decrease in convection via local feedbacks
• Anomalous high pressure
• Anomalous anticyclonic flow
• Increased descent and a northward stormtrack
  shift
• Changes in rainfall distribution

500mb Zonal Winds
500mb Winds Heights
25MW-CTL
25MW-CTL
45
75SW Fixed Patch Experiment Initial Root Zone
Soil Moisture
Southwest 75SW
Interactive Soil Moisture (CTL)
75
Fixed Soil Moisture (75)
46
75SW Experiments
500mb Zonal Winds
Rainfall (U.S. only)
75SW-CTL
75SW-CTL
47
Effect of Local Feedbackson the Large-Scale
DRY SOIL ANOMALY
WET SOIL ANOMALY
Less Rainfall
More Rainfall
Northward Shift in Storm Track
High Pressure Anomaly
Low Pressure Anomaly
Northward Shift in Storm Track
Less Rainfall
More Rainfall
48
1993 Flood Precipitation
49
Soil Moisture-Rainfall FeedbackSummary of
Results

• The feedbacks of soil moisture to the local
  climate can induce positive feedbacks to the
  large-scale circulation patterns.
• Local soil moisture anomalies can potentially
  lead to drought- and flood-like conditions not
  only in the local region, but also in remote
  regions.
• An accurate representation of the distribution of
  soil moisture is crucial to accurately represent
  observed rainfall.
• The spatial variability of soil moisture in North
  America appears to be an important in predicting
  rainfall.

50
Final Summary

• The latest developments to the RegCM3 keep it on
  par with other state-of-the-art regional climate
  models.
• The RegCM3 is an effective tool that can be and
  is being used by scientists from developing
  nations for a variety of relevant climate related
  studies.
• Easily portable to different computer platforms.
• Advanced level research can be performed using
  relatively inexpensive computer equipment (e.g.
  PRUDENCE).

51
??