10th Annual CMAS Conference
24-26th October, 2011
Regional Climate Downscaling
Study in Eastern United States
Yang Gao, Joshua S. Fu, John B. Drake and Yun-Fat Lam
University of Tennessee, Knoxville
Oct 26, 2011
Overview of the Study
Community Earth System Model
Regional Climate Model
CESM 1.0
WRF 3.2.1
D2
D3
D1
D1: 36 km by 36km CONUS
Global gridded domain
D2: 12 km by 12km domain
D3: 4km by 4km Eastern US domain
Global Climate Model CESM
Community Earth System Model - CESM 1.0
Community Atmosphere Model
(CAM)
Community Land
Model
(CLM)
Community Sea Ice
Model
(CSIM)
Spatial Resolution: 0.9 x 1.25 degree (~100 x 140 km)
Ocean component
(POP)
Temporal Resolution: 3 hourly/Daily/monthly average data
Global Climate Simulation Scenarios
RCP 8.5
RCP 4.5
RCP:20052100
• Representative Concentration Pathways (RCP) 4.5
Low-medium emissions increasing scenario
• Representative Concentration Pathways (RCP) 8.5
Fossil fuel intensive scenario
Source: http://stratus.astr.ucl.ac.be/textbook/pdf/Chapter_6.pdf
Evaluation of CESM in Present Climate
Daily average temperature at 2 meter from 2001-2010
Bias
Correlation
 NCEP/NCAR Reanalysis daily average data (2001-2010) was used to
evaluate CESM output.
The overall biases are within -1.5 to 1.5 degree for majority of the areas
and the correlation coefficient are mostly 0.8-0.9 except the tropical
areas.
Temperature Changes in Future Climate
RCP 4.5
RCP 8.5
2050
2090
The global mean projected temperature increases by 1.2
ºC under RCP 4.5 and 1.7 ºC under RCP 8.5 by 2050s.
By the decade of 2090s, the temperature increase reaches
1.6 ºC under RCP 4.5 and 4.2 ºC under RCP 8.5.
Motivation of heat wave studies
Deaths in Chicago heat wave in 1995
http://severe-wx.pbworks.com/w/page/15957981/Droughts-and-Heat-Waves
How to Define Heat Waves?
 Continuous night minimum temperatures
 Two maximum daily temperature thresholds
Probability Distribution
T1 and T2: the 97.5th and 81st percentiles
Source: CCSP, 2008
A heat wave period is:
Heat Wave Duration and Frequency
Heat Wave Duration
Present climate
2001-2010
RCP 4.5
2050-2059
RCP 8.5
2050-2059
Heat Wave Frequency
Daily Max T (K)
Why are ocean heat wave durations higher?
Land: 2001-2010
Land: 2050-2059
Ocean: 2001-2010
Ocean: 2050-2059
The diurnal variations of ocean air temperature are much smaller than land,
which leads to higher increase of heat wave duration
Seasonal extreme temperature durations
2050-2059
2001-2010
Global
Global Land
Global Ocean
Dramatic increases in durations show up almost in every region in
all the seasons
Linkage from Global to Regional Model
2001-2004/2005-2100
Global Model
CESM
chem link
met link
WRF
Regional
Climate Model
Heat waves
SMOKE
Emission
Inventory
CMAQ
Regional
Chemistry Model
PM2.5, O3,
Dry/Wet depositions
Regional Simulation Domains
Regional Model WRF Simulation domains
D2
D3
D1
Simulation period: Base: 2001-2004
Future: RCP8.5 (2057-2059)
Evaluation of Horizontal Patterns
Temperature (2 meter)
CESM
WPS
(K)
• Spatial patterns are similar between CESM and WPS
• Other variables (both surface and vertical layers) show similar
patterns
Daily mean temperature from 2001-2004
Temperature at 2 meter (unit: K)
D2
D3
MADIS
WRF
CESM
• Similar diurnal trends were performed between CESM and WRF
• Both CESM and WRF captures the major diurnal trends of MADIS
(Meteorological Assimilation Data Ingest System)
Spatial patterns of precipitation
UDel_AirT_Precip
WRF-D2
CESM
WRF-D3
Improvement in WRF precipitations in Southeast areas
compared with CESM, but overestimation in Northwest
Average temperature change from 2001-2004 to 2057-2059
CESM
WRF-D2
WRF-D3
• Overall, temperature increases by 1.5 to 3.5 K.
• Temperature increase in WRF is slightly lower than CESM in
Southeast US
Annual heat wave durations
CESM
2001-2004
2057-2059
WRF
2001-2004
2057-2059
• both CESM and WRF show significant increase of heat wave
duration days
• Compared with CESM, WRF shows slightly low increase in
Northeast, but more spatial variations due to high resolution
Summary
• Dynamical climate downscaling technique has been
applied from CESM to WRF, and both CESM and WRF
captures the major diurnal trends of MADIS
• From both global and regional climate simulations,
the heat wave duration in future shows significant
increasing trends
• In addition to climate downscaling, chemistry
downscaling is show high heat wave duration increase
in future climate conditions
Acknowledgement
This research was supported in part by the National
Science Foundation through TeraGrid resources provided
by National Institute for Computational Sciences (NICS)
under grant number [TG-ATM110009].
This research also used resources of the Oak Ridge
Leadership Computing Facility at the Oak Ridge National
Laboratory, which is supported by the Office of Science of
the U.S. Department of Energy under Contract No. DEAC05-00OR22725.
This work was partially sponsored by the Centers for
Disease Control and Prevention (CDC) under a research
project cooperative agreement (5 U01 EH000405).
Thanks for your attention!
Questions?