Project Title: Assessing future flood risk from the emerging threat of Madden-Julian
Oscillation amplification using a convection-permitting climate model
PI Name: Kooperman, Gabe
Allocation End:
Project Team: "Liu, Si"
Objectives:
Climate model codes generally span many millions of lines such that exposing kernels
throughout to new forms of parallelism on modern architectures can be a major effort
only manageable by a large institutional modeling center. However, for the unique case
of super-parameterized climate models, as used in this project, the readiness problem is
simpler. This is because at the heart of the super-parameterized climate code exists a
relatively small module SAM (the System for Atmospheric Modeling) for cloudresolving calculations.
SAM ultimately comprises 99% of the work in target overall model calculations. This
reduces the scope of development work to boost performance by orders of magnitude and
makes it approachable even by small XSEDE staff / PhD student teams. Some profiling
of the SAM code has been done. But we are looking forward to porting the program into
it to the next-generation KNL systems.
In this project, we plan on closing the gap between national flood forecasting and local
emergency response, creating real-time flood information services, and engaging
academic community through National Water Center. A flow prediction using land
surface runoff prediction pilot has been implemented in Texas. The current goal is to
expand the work to the US including different prediction models (i.e. WRF-Hydro). The
success of this collaboration will allow the success of this venture since the NFIE project
will use the model results from Stampede to support the activities that have been
proposed for the coming year.
Summary:
In this quarter, Si has worked with other ECSS team members and the SAM developers a
lot to get the code working parallelly on Stampede and Stampede-KNL systems.
Efforts:
Si Liu: over 25% in the last quarter (Jan-Mar, 2017)
Work plan Updates:
- Si has ported SAM on the Stampede and Stampede-KNL systems. There are a lot of
issues of this code. Si has spent a lot of time compiling and debugging the code itself to
run them with multiple cores on Stampede and Stampede KNL.
-Si has got the updated CESM and SPCESM code and made a few test runs on
Stampede and Stampede KNL systems.
Additional Questions for Consideration:
* Is this project ready to highlight in a symposium talk?
Not yet
* Is there any advanced topic documentation I need to prepare?
Not yet
* If the PI was satisfied, have I asked for a quote regarding the value of the collaboration?
N/A
* Would others benefit from a tutorial (asynchronous or other) on the techniques I used in this
project?
N/A
* Should I present about this work at a domain conference (paper, poster)?
Not yet
* Is the work worthy of a paper, either in collaboration with the PI or individually?
Not yet
* Are there new user requirements that have resulted from this work? Would a change in
XSEDE¹s offerings make this project much easier? If so, feed info UREP
N/A
* Is it a good time for Nancy and Ralph to check in with the PI?
Not yet