the democratization of climate change science the edgcm project columbia university and nasa goddard...
Post on 22-Dec-2015
216 views
TRANSCRIPT
The Democratization of Climate Change Science
The EdGCM ProjectColumbia University and NASA Goddard Inst. for Space Studies
MATC and We Energies 3rd Annual Green Colleges WorkshopApril 20, 2007 - Mequon, Wisconsin
Democratization of Advanced Computing
• Putting complex models into the hands of researchers and educators who otherwise wouldn’t have advanced capabilities means the tools must be:
– Highly intuitive– Free of jargon, except where its intent is instructive– Usable on existing computing resources– Inexpensive– Responsive support– Applicable to real world problems (as opposed to “toy” problems)– Tied to current events and societal needs
Based on a concept by K. DroegemeierUnidata Summer Workshop 2006Keynote
Why the need for climate modeling in the classroom?
Climate models are the primary research tools used for assessing future climate change.
Public policy decisions rely heavily on climate model simulations.
CAUSE EFFECT
Recent Climate Change:Observed Forcings and Results
Uncertain Processes
Future Climate Change:Observed Forcings
Uncertain Processes and Results
Past Climate Change:Observed Results
Uncertain Processes and Forcings
The “Keeling Curve”: CO2 rise during the 20th Century
Image Credit: NASA Earth Observatory
Year
Car
bon
Dio
xide
(pa
rts
per
mil
lion
)
280
380
300
320
340
360
1960 1970 1980 1990 2000
Mauna Loa Observatory
2005 - Warmest Year on Record
http://data.giss.nasa.gov/gistemp/2005/
Hansen, 2004
DAI
EdGCM Project Objectives
1) Allow schools to run a global climate model on a desktop computer, encouraging students to participate in the full scientific process including: experiment design, running simulations, analyzing data and reporting on results.
EdGCM Project Objectives
1) Allow schools to run a global climate model on a desktop computer, encouraging students to participate in the full scientific process including: experiment design, running simulations, analyzing data and reporting on results.
2) Facilitate collaborations between the education community and research institutions and among secondary schools and universities. In this way, students will become familiar with the role that teamwork plays in scientific research.
EdGCM Project Objectives
1) Allow schools to run a global climate model on a desktop computer, encouraging students to participate in the full scientific process including: experiment design, running simulations, analyzing data and reporting on results.
2) Facilitate collaborations between the education community and research institutions and among secondary schools and universities. In this way, students will become familiar with the role that teamwork plays in scientific research.
3) To demystify some of the complexities of how scientists are making forecasts about future climate change.
While I am prepared to accept that there may be global warming, nothing I have seen shows a causal connection. Tim, UK
One of the most persistent arguments made by those who do not believe that climate change is caused by human activity is that there is "no consensus" amongst climatologists about this. Tim Dennell, Sheffield, United Kingdom
Five hundred years ago the scientific consensus was that the world was flat. Thirty years ago, we were about to enter another ice age. Twenty years from now we will all have a good laugh about "global warming". Thumper31
BBC website blog - a typical day in July (2006)
Addressing Public SkepticismBy Reducing the Mystery of the Techniques
Global Climate Models
(Henderson-Sellers, 1985) (Hansen et al., 1983)
What is the “GCM” in EdGCM?
Computer simulations of the Earth based onfundamental physics and empirical parameterizations
Schematic View of GCM Parameterizations
SnowballEarth_Sim2.R Model II 8/24/2000
Owner: Dr. Mark Chandler, [email protected]: Paleoclimate GroupThis experiment simulates a time period approximately 600 million years ago.
Object modules:MainC9DiagC9RadC9FFTC9UTILC9
Data input files:7=G8X10_600Ma9=NOV1910.rsf_snowball15=O8X10_600Ma17=25=Modern_OceanTransports19=CD8X10_600Ma23=V8X10_600Ma26=Z8X101_600Ma21=RTAU.G25L1522=RPLK2529=Snowball_Earth_Regions
Label and Namelist:Snowball_sim2 (Snowball Earth Experiment: 600 million years ago)
&INPUTZ TAUI=10176.,IYEAR=1900, KOCEAN=1, SRCOR=.95485638151, S0X=1.,CO2=.31746031746031, USET=0.,TAUE=35040., USESLP=-12., ISTART=3,KCOPY=2,NDPRNT=-1,TAUE=10177.,TAUP=95616., &END
C** INITIALIZE SOME ARRAYS AT THE BEGINNING OF SPECIFIED DAYS fName = './prt/'//JMNTH0(1:3)//CYEAR//'.prt'//LABEL1(
IF(JDAY.NE.32) GO TO 294 JEQ=1+JM/2 DO 292 J=JEQ,JM DO 292 I=1,IM 292 TSFREZ(I,J,1)=JDAY JEQM1=JEQ-1 DO 293 J=1,JEQM1 DO 293 I=1,IM 293 TSFREZ(I,J,2)=JDAY GO TO 296 294 IF(JDAY.NE.213) GO TO 296 JEQM1=JM/2 DO 295 J=1,JEQM1 DO 295 I=1,IM 295 TSFREZ(I,J,1)=JDAY C**** INITIALIZE SOME ARRAYS AT THE BEGINNING OF EACH DAY 296 DO 297 J=1,JM DO 297 I=1,IM TDIURN(I,J,1)=1000. TDIURN(I,J,2)=-1000. TDIURN(I,J,3)=1000. TDIURN(I,J,4)=-1000. TDIURN(I,J,5)=0. TDIURN(I,J,6)=-1000.
PEARTH=FDATA(I,J,2)*(1.-FDATA(I,J,3)) IF(PEARTH.GT.0.) GO TO 297 TSFREZ(I,J,1)=365. TSFREZ(I,J,2)=365. 297 CONTINUE C**** INTEGRATE DYNAMIC TERMS 300 MODD5D=MOD(NSTEP,NDA5D) IF(MODD5D.EQ.0) CALL DIAG5A (2,0)
Unix Scripts FORTRAN Code
SnowballEarth_Sim2.R Model II 8/24/2000
Owner: Dr. Mark Chandler, [email protected]: Paleoclimate GroupThis experiment simulates a time period approximately 600 million years ago.
Object modules:MainC9DiagC9RadC9FFTC9UTILC9
Data input files:7=G8X10_600Ma9=NOV1910.rsf_snowball15=O8X10_600Ma17=25=Modern_OceanTransports19=CD8X10_600Ma23=V8X10_600Ma26=Z8X101_600Ma21=RTAU.G25L1522=RPLK2529=Snowball_Earth_Regions
Label and Namelist:Snowball_sim2 (Snowball Earth Experiment: 600 million years ago)
&INPUTZ TAUI=10176.,IYEAR=1900, KOCEAN=1, SRCOR=.95485638151, S0X=1.,CO2=.31746031746031, USET=0.,TAUE=35040., USESLP=-12., ISTART=3,KCOPY=2,NDPRNT=-1,TAUE=10177.,TAUP=95616., &END
C** INITIALIZE SOME ARRAYS AT THE BEGINNING OF SPECIFIED DAYS fName = './prt/'//JMNTH0(1:3)//CYEAR//'.prt'//LABEL1(
IF(JDAY.NE.32) GO TO 294 JEQ=1+JM/2 DO 292 J=JEQ,JM DO 292 I=1,IM 292 TSFREZ(I,J,1)=JDAY JEQM1=JEQ-1 DO 293 J=1,JEQM1 DO 293 I=1,IM 293 TSFREZ(I,J,2)=JDAY GO TO 296 294 IF(JDAY.NE.213) GO TO 296 JEQM1=JM/2 DO 295 J=1,JEQM1 DO 295 I=1,IM 295 TSFREZ(I,J,1)=JDAY C**** INITIALIZE SOME ARRAYS AT THE BEGINNING OF EACH DAY 296 DO 297 J=1,JM DO 297 I=1,IM TDIURN(I,J,1)=1000. TDIURN(I,J,2)=-1000. TDIURN(I,J,3)=1000. TDIURN(I,J,4)=-1000. TDIURN(I,J,5)=0. TDIURN(I,J,6)=-1000.
PEARTH=FDATA(I,J,2)*(1.-FDATA(I,J,3)) IF(PEARTH.GT.0.) GO TO 297 TSFREZ(I,J,1)=365. TSFREZ(I,J,2)=365. 297 CONTINUE C**** INTEGRATE DYNAMIC TERMS 300 MODD5D=MOD(NSTEP,NDA5D) IF(MODD5D.EQ.0) CALL DIAG5A (2,0)
Unix Scripts FORTRAN Code
Run by advanced scientific programmers on supercomputing clusters.
• A Global Climate Model• Relational Database (organize massive I/O)• Graphical User Interface• Post-Processing Software• Scientific Visualization Tools• Software for constructing scientific manuscripts
and publishing them to the web• Administrative Tools (disk management,
security, links to assignments, etc.)
What is EdGCM?
Educational Global Climate Model
Scarsdale Teachers Institute, March 3 and 4, 2006
eJournals Image Libraries
Web-based Publishing
The EdGCM Website • Download EdGCM Software and Materials• Electronic Forums• Community and Student Showcases• Development News
http://edgcm.columbia.edu
Alternate Title Slide ----->
The Democratization of Climate Change ScienceThe EdGCM Project
Mark ChandlerNASA/GISS at Columbia University