research proposal on - tropmet.res.in · research proposal on portable version of cfs model for the...

Research Proposal

on

Portable version of CFS model for the monsoon mission

to

National Monsoon Mission

Research Proposal Submitted to: IITM, Pune

Proposal Ref No: CDAC/B/HPCSG/SG/NMM/2011-12/002

Dated: 25.08.2011

Proposal by: Dr. S. Janakiraman, SPIM (Seasonal Prediction of Indian Monsoon) Group

National Monsoon Mission Proposal

Centre for Development of Advanced Computing (C-DAC) Pages 1-28

Pune, Mumbai, Bangalore, Delhi, Noida, Hyderabad, Chennai, Kolkata, Thiruvananthapuram, Mohali

Research Proposal

1.0 Title of the proposed project:

Portable version of CFS model for the monsoon mission.

2.0 Brief information about Principal Investigator (PI) and Co-PI(s) :

PI :

Name: Dr. S. Janakiraman

Date of birth: 18th

March 1973

Institution: Centre for Development of Advanced Computing

Qualification: Ph.D. (Numerical Weather Prediction)

Co - PI (1):

Name :

Date of birth :

Institution :

Qualification :

Co - PI (2):

Name :

Date of birth :

Institution :

Qualification :




3.0 Project Summary:

(a) Intellectual merits of the proposed work

Climate Forecast System is a large and complex scientific software used by the U.S.

National Weather Service for Seasonal Forecasting purposes. Presently, it has been

identified as a base tool towards the development of an improved forecast model for

Indian monsoon. National monsoon mission undertakes to build an improved model, the

collective efforts of the modeling experts available at the various academic and research

institutions across the country.

Porting of the code to Linux operating system will make it accessible to a wider

community of researchers. Making it happen in the absence of inadequate documentation

and a detailed user manual requires reverse engineering of the software, combined with

the knowledge of weather and climate modeling and other computational science

expertise.

(b) Broader impacts of the proposed work

The present version of the CFS model transferred to the Indian community was originally

developed on the IBM HPC systems of NCEP, USA. As such the model has many IBM

AIX specific architecture bindings. As the model has been developed and used by the

limited group of scientists at a research setting, there is not enough documentation

available for a typical user in the university environment with the commodity-off-the-

shelf clusters. Availability of the portable version of the model will break the barrier to its

wider usage. Portable version of the model is essential for wider participation in the

monsoon mission. It will help to build a nationwide modelling community.




Project Description:

1. Research Objectives

To create a portable version of the Climate Forecast System model

1.1 Intellectual merit of the proposed work

Porting of large scientific software such as CFS to a popular open source Linux

operating system will broaden its usage. The exercise of porting to another

platform makes the software robust. The software when exposed to a different

compiler system shows off the portions of the code that have specific platform

bindings. These platform-specific bindings of the software are among the most

vulnerable spots leading to bugs. Porting exercise transforms these vulnerable

spots and as a result software becomes robust. Open source software has the

advantage of being accessible by a wider community and as a result it is also

widely tested.

CDAC has expertise in building indigenous supercomputing systems and in

developing scientific applications on them. CDAC has successfully implemented

an end-to-end medium range weather forecasting suite from CRAY

supercomputing systems to PARAM series of parallel supercomputing systems.

The end-to-end forecasting suite of NCMRWF has decoder programs, complex

quality control, SSI based data assimilation system and Medium range forecast

model. VAX/VMS based decoder programs have been successfully ported to

PARAM parallel supercomputing systems. Complex Quality Control software,

Data assimilation system and Medium range forecast model are implemented

from CRAY-XMP supercomputer to PARAM parallel supercomputing systems.

The principal investigator has wide experience in porting, optimizing and

parallelizing the weather and climate models on PARAM supercomputing

architectures. The principal investigator had implemented the Community

Climate System Model (CCSM) on PARAM Padma parallel supercomputing

systems and has conducted 100-year climate integrations.5




The principal investigator had successfully guided his team in delivering the

simulations for DST sponsored ‗Seasonal Prediction of Indian Monsoon' project.

Under the SPIM project, Atmospheric General Circulation Models used by the

Indian research groups namely UKMO PUM v4.5 model (IITM-Pune), COLA

AGCM (IITM-Pune), NCMRWF T80 model (NCMRWF), SFM model

(IISc/IMD), ECMWF GCM (IIT-Delhi) and Varsha model (NAL) were

successfully ported on to PARAM Padma supercomputer; Seasonal hindcasts

were performed for the summer monsoon seasons of 1981 – 2004. The post-

processing of model outputs and visualization for the project were successfully

conducted at CDAC – Bangalore 2.

The principal investigator and his team have implemented CFS v1 model on

PARAM Padma parallel computing platform.3 Hindcast experiments were

conducted for 2009 monsoon season to identify the systematic model errors in

simulating the Indian summer monsoon.4

1.2 Scope of the proposed work

The proposed project aims to apply the open source software paradigm for the

purpose of evolving a better seasonal prediction model for Indian monsoon under

the mission. The strengths of the open source software paradigm is widely

discussed in the computing literature; notable among them is the article titled

'The Cathedral and The Bazaar' written by Eric S. Raymond.1 Porting the CFS

software to the widely popular Linux open source platform is an important step in

making the model open source. CFS has been legacy software available only to

the select researchers and collaborators of NCEP. The modeling experts of NCEP

and their collaborators have so far driven the model development process. This

kind of development process is known as ‗the Cathedral‘ model of software

evolution. Inspired by the success of the open source software approach to model

development, NCEP has been making the source code available to the researchers

through the World Wide Web. But the software available in the present form has

many platform bindings and can be used only by the weather and climate




modelers with several years of computational expertise. 'The Bazaar model' of

software development, successfully carried in the open source software

community ensures stable software and wider usage. This project proposes to

employ ‗Bazaar‘ paradigm of software development to the CFS model by making

it portable on the Linux operating system.

The project proposes to undertake the porting of CFS model component by

component, and test the porting of the component as individual units. Testing of

the integrated system will be done after completion of unit test of all the

components. CFS model installed on the IBM AIX platform will serve as a

reference for all the checks. The Linux port of the model will be compared with

the reference at each stage. After port verification of the model, optimization of

the program will be carried out. Also, the user scripts will be modified to suit the

needs of the Indian monsoon research community. IBM AIX nodes and Intel P4

Linux clusters are available for the porting work. Intel Compute cluster has Intel

Fortran compiler and Intel MPI libraries.

1.3 Broader Impact of proposed work

The portable version of CFS will enable wider participation from the research

community. The broader participation of researchers will enhance the

understanding of the modeled processes of Monsoon dynamics. Due to the

existing manpower crunch, no single institution or a team will be able to gain

adequate expertise for an exhaustive and complete understanding of the CFS

model. Many researchers need to focus on the diverse aspects of the model to gain

in-depth knowledge. The in-depth knowledge gained by different researchers need

to be consolidated to make improvements to the model. The science of modeling

& the aspects of the climate system are vast and diverse. To increase the

understanding of modeling and the science of Indian monsoon, as many

researchers as possible need to participate in the mission. Portable version of the

model will be a vital factor to enroll and enlist the expertise of the researchers

working in academia and research institutions across the country. As more and




more researchers join the mission to tackle the challenge of seasonal prediction in

a systematic manner, CFS will evolve as a community resource of the nation.

Thus, the scientific knowledge of the researchers can be consolidated to achieve

significant improvement to the modeling system. It is envisaged that the portable

version of CFS will serve as the seed to germinate such a community of modelers.

Many of the university researchers have access to Beowulf Linux clusters. With

the availability of portable version of CFS model, these researchers from

universities and the affiliated colleges will be able to participate in the monsoon

mission. Tapping the expertise in the universities and colleges is important for the

success of the vast scientific challenge of improving the monsoon prediction

with CFS.

Availability of CFS model in the public domain will help more researchers to use

and experiment with the model. Many student projects can be created with the

portable version of the model that can substantially contribute to the trained

manpower in the subject area. For this purpose, the portable model will be made

available to the user community through the World Wide Web.

2. Technical details of the proposed work

2.1 Scientific Context

National Mission on Monsoon undertakes to put in place a dynamical modeling

framework for the accurate seasonal prediction of Indian monsoon. Despite

several advances in the understanding of the physics of the monsoon variability,

advancements in the numerical modeling, availability of more observational data

and the availability of higher computational capabilities, accurate prediction of

Indian summer monsoon remains a serious challenge. As the scientific challenge

is vast due to its interdisciplinary nature, mission undertakes to bring about a

community of researchers to tackle the problem. The communities of researchers

working on the diverse aspects of the monsoon system are expected to use CFS

model as a base resource to bring about an improved model. So, it is imperative

that the base model is accessible to a wider spectrum of the research community.




Accessibility of the software here means making it independent of computer

platform so that it can even run on the commodity compute clusters.

2.2 CFS modeling system

CFS is a collection of many software programs that are the global models of the

atmosphere and ocean systems, land-surface model, sea-ice model and the

coupler. The copy of the software program running on the IBM AIX computer

systems of NCEP is provided to the Indian research community as it is. As such

it assumes prior familiarity with the software. The existing User document

available will be just sufficient for the person with few years of experience with

running the model and using their output. Most of the researchers, who want to

participate in the monsoon mission, would want to spend less of their time in

dealing with the software issues and spend more of their time on improving the

scientific aspects of the model. So there is an immediate need to make the CFS

model platform independent and easier to use by the meteorological community.

The absence of detailed software documentation poses some challenges in porting

the model to the Linux operating system.

CFS model is large suite of FORTRAN programs comprising the global models

of atmosphere, ocean, land-surface and sea-ice. Porting to Linux operating system

requires careful attention to the details.

The directory structure of Climate Forecast System v.2 (CFS v.2) is given in

Fig.1.

bin com exec exp fix IC jobs lib parms scripts sorc ush util vsdb

Figure 1: The directory structure of the CFS software system.

The details of the directories are given below.

a. bin - submit scripts

b. com - model output directory

cfsv2




c. exec - model executable files

d. exp - experiment directory

e. Fix - fixed input files for the model

f. IC - initial conditions

g. Jobs - driver scripts for running the model

h. Lib - user built libraries for the model

i. parms - parameter configuration files and namelist files

j. scripts - main scripts to run the model and the post-

processor

k. sorc - the source directories for model parts

l. ush - scripts to run model parts

m. util - utility scripts and fix files

n. vsdb - verification package

2.3 Porting and optimization

The CFS v2 software provided by NCEP will be installed as is on the IBM AIX

system available at CDAC. This base version of the system will be port verified.

This AIX version of CFS v2 will be used as a reference software setup for the port

verification of the Linux version.

After the completion of the porting of the model, software performance profiling

will be carried out to identify the portions of the code that consume large amounts

of time. These portions of the code will be taken up for performance optimization

through compiler. Job scripts will be modified so as to make them easier for the

general scientific user.

3. Statement of Work (methodology to be adopted)

The original version of the CFS v2 software will be installed on IBM AIX

platform available at CDAC. Validated version of this model along with its test




datasets will be used as a reference. Each component of the CFS v2 model will be

ported on the Linux cluster available at CDAC. Each of the ported components

will undergo unit testing. After the completion of the unit testing of every

component, they will be integrated together. The integrated version will be

validated through the reforecast data available for CFS.

Given below are the break-up of the work elements under the project and the

estimates of required time for them.

Atmosphere component of CFS

Porting - 2 months

Validation – 15 days

Ocean component of CFS

Porting – 2 months


Coupler

Porting – 1 month


Post processors

Porting – 15 days

Validation – 1 week

Utilities and Tools (user-built libraries, utility scripts, run scripts etc.)

Porting – 2 months

Validation – 1 month

Complete testing of the coupled system

Driver scripts development and testing – 1 month

Verification packages porting & testing – 1 month




3.1 Schedule (Year wise)

Year Expected Outcome Deliverables

Year - 1 Portable version of CFS

v2. Linux port of CFS v2,

User manual,

Validation datasets.

3.2 Team Composition and expertise

Investigator Qualification Expertise

PI M.Sc. (Maths), 1995, IIT-Madras.

M.Sc. (Engg.)

Atmosphere modeling, 1999, IISc

– Bangalore.

Ph.D. (Numerical methods for

atmosphere modelling), 2010,

IISc – Bangalore.

15+ years of research

expertise in scientific

computing related to

weather and climate

modelling. High

Performance computing,

Porting and debugging of

legacy codes, software

optimization and

parallelization, Numerical

methods for atmosphere

modelling.

3.3 Connections to Operational forecast and Human resource development

Availability of CFS model in the public domain will help more researchers

to use and perform experiments with the model. Many student projects can

be created with the portable version of the model that can substantially

contribute to the trained manpower in the subject area.

4. Related works

4.1 National status

To the best of our knowledge, there is no specific work carried out in India to

make CFS model portable.

There are significant efforts expended to make the NCMRWF atmospheric

general circulation model on Linux platform by National Aerospace Laboratories




(NAL). Varsha GCM of NAL is an example of an atmosphere model evolved

through porting and extensive code refactoring. Otherwise, there are no known

efforts as far as coupled ocean-atmosphere models are concerned.

4.2 International status

National Centre for Atmospheric Research (NCAR), Colorado, Boulder, U.S.A. is

a pioneering effort that introduced community wide open source development

paradigm for Atmosphere modeling. Community Atmosphere Models (CAM) of

NCAR successfully evolved to become Community Climate System Models

(CCSM). CCSM is a coupled atmosphere-ocean-land – sea-ice model of our

climate system. CCSM has now further evolved in to Community Earth System

Model (CESM) (http://www.cesm.ucar.edu).

There are many other modelling centers such as GFDL, MIT that also followed

the open source software paradigm for atmosphere/ocean modelling.

Ported Unified Model (PUM) is another excellent example for the case of porting

the operational model of the U.K. Meteorological Office. National Centre for

Atmospheric Sciences (NCAS), U.K has ported the Unified model of the U.K.

meteorological office for research usage. A consortium of atmosphere and ocean

model researchers from the U.K. universities and research laboratories contribute

to the improvement of the model and further understanding of science.

The Australian Community Climate and Earth-System Simulator (ACCESS) is a

coupled climate and earth system simulator to be developed as a joint initiative of

the Bureau of Meteorology and CSIRO in cooperation with the university

community in Australia (www.accessimulator.org.au).

5. Results from prior MoES support (if any)

Not Applicable

6. Facilities available at the workspace

(e.g., existing computer facilities)

http://www.accessimulator.org.au/




Resources at CDAC Tera scale Facility

Following are the resources currently available at CTSF. These are continuously

upgraded to address the requirements as they emerge.

Hardware Resources

GARUDA Grid -Bangalore Compute nodes: Aggregate Peak Computing Power

is 4 TeraFlops

GG-BLR Cluster

Head node / File server

Processor 2 X Quad core Xeon @ 3.16 Ghz

Memory 24 GB

Internal Storage 2 / 4 * 146 G SAS HDD

Operating System Rocks 5.0 on RHEL 5.1 x86_64

Compute Node [40 numbers ]

Processor 2 X Quad core Xeon @ 3.16 Ghz

Memory 16 GB

Internal Storage 2 X 250 G SATA HDD

Operating System Rocks 5.0 on RHEL 5.1 x86_64

Networks

Primary Infiniband @ 20 Gbps Full Duplex

Backup Gigabit Ethernet @ 1 Gbps Full Duplex

Management 10/100 Mbps Fast Ethernet




External Storage

Storage Array 10 TeraBytes SAS

24 TeraBytes SATA

PARAM PADMA Cluster

Compute Nodes

1. Power4 Servers

Configuration: 54 Nos. of 4 Way SMP nodes and 1 No. of 32 Way SMP

node

No of Processors: 216 (Power4 @1GHz)& 32 (Power 4 @1.1GHz)

Aggregate Memory: 0.5 TB (@ 8GB per node and 64 GB for large SMP node)

Internal Storage: 4.5 TB (@ 72GB per node and 576 GB for large SMP

node)

Operating System: AIX 5L / Linux

Aggregate Peak Computing Power: 1005 GFs (~1 TF)

2. Power5 Servers

Configuration: 18 Nos of 8 Way SMP nodes (P550, 1.5 GHz)

1 No. of 16 Way SMP node (P570, 1.9GHz) and

1 No of 2 Way SMP node (P505, 1.65GHz)

No of Processors: 162 (Power [email protected])

Aggregate Memory: 356GB (@ 16GB per node and 64 GB for large SMP node)

Internal Storage: 3 TB (@ 144GB per node and 576 GB for large SMP node)

Operating System: AIX 5.3L

Aggregate Peak Computing Power: 998.8 GFs (~1 TF)

File Server

Configuration: 6 Nos. of 4 Way SMP

No of Processors: 24 (UltraSparc-III@900MHz)

mailto:[email protected]




Aggregate Memory: 96 GigaBytes

Internal Storage: 0.4 TeraBytes

File System: QFS

Parallel File System: CPFS

Operating System: Solaris 10

Network

Primary: PARAMNet-II @ 2.5 Gbps Full Duplex

Backup: Gigabit Ethernet @ 1 Gbps Full Duplex

Management: 10/100 Mbps Fast Ethernet

External Storage

Storage Array: 10 TeraBytes with Stk6140 & T3 Storage disk arrays

Tape Library: 12 TeraBytes - (5 LTO drives)

Storage Area Network consisting of 2GB FC 16 port switches and a 1 Gbps

Brocade 2800 FC switch connected through a fiber channel.

Two Compaq Evo W6 graphic workstations running Slackware Linux 9.0

and Windows 2000 and number of user terminals on the LAN to connect to

the PARAM Padma Cluster.

Laser printer.

Software Resources

Operating System

The Operating System of PARAM Padma is AIX 5.3L which supports

multithreading and multiprocessing. This System V-style UNIX provides a

familiar development environment to the users. Optionally Linux operating

environment is also provided.

Compilers and Related tools

The following compilers and related tools are available on PARAM Padma.




- XL Fortran compiler and Run Time Environment

- F90 Compiler

- F77 Compiler

- C for AIX V 6.0

- Visualage C++ Compiler

- CDF90 Compiler – C-DAC‘s Parallel FORTRAN Compiler

- C-F77to90 – C-DAC‘s converter for Fortran77 to Fortran90 programs

- Complete gcc & GNU libtools

Parallel Processing Libraries

- Message Passing Interface (MPI) – IBM PE (Parallel Environment)

- Public domain MPI - MPICH 1.2.4 for XL & gcc, g77

- C-MPI – C-DAC‘s optimized implementation of MPI for PARAM Padma

- KSHIPRA – C-DAC‘s Communication substrate for PARAMNet II

Parallel Numerical Libraries

- IBM ESSL & PESSL

- SCALAPACK

- LINPACK

- PLAPACK

- MASS

- PetSC

Program Development Tools

- DIViA – C-DAC‘s parallel program-debugging environment




- Metric Advisor – C-DAC tools for enabling developers to build run

quickly.

- dbx, xprofiler

System Administration and Job Management Tools

- Cluster Monitoring Tools

- System Accounting Tool

- Dedicated Slot Booking Tool

- Load leveler 3.1

- PARMON – C-DAC‘s comprehensive cluster monitoring system

- Veritas Backup Tool (Net Backup 4.5)

7. Budget requirements (with justifications)

(a) Emoluments for research personnel, technical and administrative support

7.1 Budget requirement for Key personnel:

One Senior Scientist of E5 Grade for 12 Months= `9.6 Lakhs

7.2 Budget requirement for other personnel (e.g. Research Assistants):

One Scientist of E1 grade + One Scientist of E2 Grade + Two project

Engineers for 12 Months = ` 20.52 Lakhs

7.3 Budget for Technical and administrative support

One Technical Writer for Documentation = ` 1.8 Lakhs

(b) Budget requirement for Travel

7.4 Budget for Domestic Travel:

` 2 Lakhs

7.5 Budget for Foreign Travel:

Not Applicable




(c) Budget for Other Direct Costs:

7.6 Budget for Material & Supplies

` 2 Lakhs

7.7 Budget for Computer Services

` 2 Lakhs

7.8 Budget for Equipments

` 5 Lakhs

(d ) Budget for Indirect Costs

7.9 Budgets for Facilities & Administrative Costs

` 6.44 Lakhs

Total Budget requirement: `49.36 Lakhs

Budget Outlay

Slno Budget Head Expenditure Total Rs in Lakhs

1 Capital Equipment & Software

5.00 5.00

2 Consumable Stores

2.00 2.00

3 Manpower

31.92 31.92

4 Travel & Training

2.00 2.00

5 Contingencies including TA/DA for Project Review Meetings

2.00 2.00

6 Overheads, if any

6.438 6.44

Grand Total 49.36




8. Bio-data (CV) of the Investigators:

8.1 PI Biography (Person A)

Name: Dr. S. Janakiraman

Date of birth: 18/03/1973

Institution: Centre for Development of Advanced Computing

(C-DAC)

Address (Residence):

No. 412, 4B Cross, BDA Layout, East of NGEF, Off R.M. Nagar police station,

Dooravani Nagar P.O., Bangalore – 560 016.

Tel. No.: +91-9342205014.

Address (Office):

Centre For Development Of Advanced Computing (C-DAC)

(A Scientific Society of the Ministry of Communications and Information

Technology, Govt. Of India)

'C-DAC Knowledge Park' No.1, Old Madras Road, Opp.HAL Aero-engine

Division, Byappanahalli, Bangalore-560 038

Tel Nos: +91-80-66116416

Fax: +91-80-25247724

Educational Qualification:

School/College/University Degree Year Main

subjects

Division/Class

Indian Institute of

Technology, Madras

M.Sc

1995

Mathematics

Indian Institute of Science,

Bangalore

M.Sc (Engg)

1999

Atmosphere

Modelling

Indian Institute of Science PHD 2010 Numerical

methods in

atmosphere

modelling

Awards / Honors / Fellowship etc.:

McGuire medal for undergraduate English – 1992.

GATE – All India Rank of 100 in Mathematics – 1996.




IISc research fellowship (1996 – 1998).

Citation for exemplary performing member of the student's council, IISc

(1999).

Appointments (Professional experience/employment record):

Organization Designation / Position Duration (Year / date)

Ocean Engineering

Centre, Indian Institute of

Technology, Madras.

Project Associate

July 1995 – June 1996

CDAC, Pune

Computational

Atmospheric Sciences

team

Member Technical

Staff/ Project Leader.

Oct. 1998 – Nov. 2003.

CDAC, Bangalore Member Technical

Staff / Project Leader

Dec. 2003 – Oct. 2007.

CDAC, Bangalore Member Technical

staff/Team Coordinator

Nov. 2007 – present.

List of important and relevant research publications:

S. Janakiraman, Mohit Ved, Ramesh Naidu Laveti, Priyanka Yadav and

Sulochana Gadgil, Prediction of the Indian Summer Monsoon rainfall by a state-

of-the-art coupled ocean-atmosphere model. Current Science, Vol. 100, No. 3,

2011, pp 354 – 362.

S. Janakiraman, Ravi S Nanjunidah and A.S. Vasudeva Murthy : A novel variable

resolution global spectral method on the sphere , Journal of Computational

Physics (under revision).

S. Janakiraman, Ravi S Nanjundiah and P.N. Vinayachandran: Simulations of the

Indian summer monsoon with a coupled ocean-atmosphere model on PARAM

Padma. Current Science, Vol. 89, No. 9, pp 1555 – 1562.




J.V. Ratnam, Manik Bali, S. Janakiraman and Akshara Kaginalkar: Performance

of Parallel Spectral Statistical Interpolation code on PARAM 10000, HPC-Asia

2002, Proceedings of the Sixth International Conference on High-Performance

Computing in the Asia-Pacific region. IEEE Computer Society, 2002.

S. Janakiraman, J.V. Ratnam and Akshara Kaginalkar : Study of Machine Round-

Off response on Weather Forecasting Simulations Using High Performance

Computing Systems, HPC-Asia 2000, Proceedings of the Fourth International

Conference on High-Performance Computing in the Asia-Pacific region, IEEE

Computer Society, Vol.2, 2000, pp 1070 – 1074.

S.C. Purohit, Akshara Kaginalkar, J.V. Ratnam, S. Janakiraman and Manik Bali :

Implementation of complete weather forecasting suite on PARAM 10000 ,

Developments in Teracomputing : Proceedings of the Ninth ECMWF Workshop

on the Use of High Performance Computing in Meteorology, 2000, World

Scientific, pp 104—109.

Other publications:

S. Janakiraman, Ravi S Nanjundiah and A.S. Vasudeva Murthy: A variable

resolution global spectral method on the sphere with finer resolution over tropics,

Contributed talk at the 2007 Workshop on Solution of Partial Differential

Equations on the Sphere, U.K. Met Office, Exeter, Sep. 24—27, 2007.

S. Janakiraman, Ravi S Nanjundiah and A.S. Vasudeva Murthy: High resolution

Tropical Belt Transformation for the variable resolution spectral models,

Contributed talk at the 2004 Workshop on the Solution of Partial Differential

Equations on the Sphere, Frontier Research System for Global Change,

Yokohama, Japan, July 20—24, 2004.

S. Janakiraman and Ravi S Nanjundiah: Impact of Water Mass Forcing on

simulation of Indian summer monsoon, HPC in Science and Engineering (HPC-

SE 2003), Moscow, June 16—23, 2003.




Participation in Conference/Seminar/Workshop/ Summer Schools

(Relevant/Recent)

1. ―Scoping workshop on Monsoon Mission‖ held at IITM, Pune. April 11th

–

April 15th,

, 2011.

2. ―Workshop on Asian Climate Change: Trends and Policy‖ held at IISc.,

Bangalore. July 20th

– July 22nd

, 2011.

3. ―Workshop on Monsoon Variability‖ held at IISc., Bangalore. August 17th

–

August 19th

, 2011.

Synergistic Activities:

Activity 1

Hosting of NCEP CFS v1 retrospective forecast datasets

(http://www.cdacb.in/html/spim_data.aspx)

The retrospective forecasts generated for 1982 - 2009 from the first version of the

climate forecast system (CFS), the fully coupled ocean-land-atmosphere

dynamical seasonal prediction system, which became operational at National

Centres for Environmental Prediction (NCEP), USA in August 2004, were made

available at the NCEP website. This data set has proved to be extremely valuable

with a large number of studies by several scientists the world over. Now that

NCEP has switched over to the next version of the model, the retrospective

forecasts for that model (CFS version 2) are available from NCEP. Recognizing

that the retrospective forecasts of CFS version 1, which is one of the best models

in the world, will continue to be an important source for researchers, CDAC

offered to host this data set on the earth sciences virtual community portal on the

GARUDA Grid (which is part of the National Knowledge Network programme of

the Govt. of India, whose aim is to interlink research institutions in India with

high bandwidth connectivity). NCEP accepted this offer and have facilitated the

transfer of this dataset.

Refer to <http://www.cdacb.in/html/spim_data.aspx>

http://www.cdacb.in/html/spim_data.aspx

http://www.cdacb.in/html/spim_data.aspx




Activity 2

Atmospheric model kernel development with variable resolution global spectral

method

9. List of supplementary documents:

(e.g., authorization letter from the Head of the organization, endorsements etc.)

Refer

Annexure 1 - C-DAC – A Brief Profile

Annexure 2 - Endorsement from the Head of the Organisation

Annexure 3 - Certificate from Investigators




10. References Cited:

1. Eric S. Raymond. 2001. The Cathedral and the Bazaar: Musings on Linux

and Open Source by an Accidental Revolutionary. O'Reilly & Associates,

Inc., Sebastopol, CA, USA.

2. Sulochana Gadgil and J. Srinivasan. 2011. Seasonal Prediction of Indian

Monsoon. Current Science, Vol 100, No. 3, pp 182 – 194.

3. S. Janakiraman, Mohit Ved and Ramesh Naidu Laveti. 2011. Seasonal

prediction capability of NCEP CFS – a state-of-the-art coupled ocean—land--

atmosphere model. CDAC Technical report – CDACB/KP/SPIM/2011/01.

4. S. Janakiraman, Mohit Ved, Ramesh Naidu Laveti, Priyanka Yadav and

Sulochana Gadgil , Prediction of the Indian Summer Monsoon rainfall by a

state-of-the-art coupled ocean-atmosphere model. Current Science, Vol. 100

No. 3, 2011, pp 354 – 362.

5. S. Janakiraman, Ravi S Nanjundiah and P.N. Vinayachandran: Simulations of

the Indian summer monsoon with a coupled ocean-atmosphere model on

PARAM Padma. Current Science , Vol. 89, No. 9, pp 1555 – 1562.




Annexure 1: C-DAC – A Brief Profile

The Centre for Development of Advanced Computing (C-DAC) is an autonomous

Scientific Society under the Ministry of Communications & Information

Technology, Government of India. Set up over two decades ago, as India‘s

national initiative for design, development and delivery of high performance

computing (Supercomputer) systems and solutions based on parallel processing

technology, C-DAC has over the years diversified its activities, transferring the

expertise it has acquired and technologies it has developed in the high-end

computing to develop and deploy advanced Information Technology (IT) based

solutions in the key sectors of economy. Through this approach, it has maintained

a balance between developing strategic technologies needed in the country in the

high performance computing area for achieving self-reliance, and addressing

mission critical problems in the science and engineering fields on one hand, and

using expertise developed to commercialize its technologies and products to meet

the requirements in the key sectors on the other.

C-DAC has been operating in a mission mode in order to develop the

technologies in the specified time targeted manner. It has accordingly developed

and commissioned PARAM Padma, a tera-scale supercomputing system at its

Knowledge Park, Bangalore. C-DAC has also earlier set up a National PARAM

Supercomputing Facility (NPSF) & recently PARAM YUVA at Pune to allow

access to researchers to solve their compute intensive problems in various areas of

Science and Engineering.

C-DAC‘s other activity, the language technology mission, is to create a

framework for support to the various living languages with diverse scripts on

standard computers. C-DAC has innovated its trail blazing Graphics and

Intelligence based Script Technology (GIST) to achieve this goal. This

technology is now extended to include multimedia and multilingual computing

solutions covering a wide range of applications such as publishing and printing,

word processing, office application suites with language interfaces for popular




third party softwares on various operating platforms, electronic mail, Optical

Character recognition, machine translation, language learning, video and

television and multimedia content in Indian Languages. These have been

successfully commercialized.

Under the mandate given to C-DAC to build up and mobilize skilled manpower to

address the growing demand for trained manpower in the extremely fast moving

sector of Information Technology, C-DAC established its Advanced Computing

Training School (ACTS). This school currently offers a variety of course options

in areas of Software technologies, Enterprise System Management (ESM),

Bioinformatics, Geomatics, VLSI designs, Embedded System Design, Digital

Multimedia & Animation etc, through a dozen different courses currently run at

its own centers and about 100 plus authorized centers around the country, and

outside.

Building on its foundation in almost all major areas of Information Technology,

C-DAC offers advanced computing products, solutions and services for several

sectors including Education, Research, Power, Steel, Telecom, Health Care,

Agriculture and Finance. Specific areas of focus currently are e-Governance, e-

Commerce, e-learning and advanced solutions based on Genetic Algorithms,

Cyber Security, Geomatics, Artificial Intelligence, Natural Language Processing,

Real Time Systems and software, Digital / broad band networks etc.

research proposal on - tropmet.res.in · research proposal on portable version of cfs model for the...

Documents