CSL Allocations 2011

The following projects had CSL allocations from April 2011 to June 2012. 

Community Earth System Model (CESM)

Project Lead: Jim Hurrell, NCAR
Allocation: 16,304,100 GAUs
Sponsors: DOE, NSF

NCAR has a proud and unique tradition of collaboration with scientists from universities, national laboratories, and other research organizations to develop, continuously improve and support the scientific use of a comprehensive Earth modeling system that is at the forefront of international efforts to understand and predict the behavior of Earth's climate. For many years this tradition has been realized through the development and application of the Community Climate System Model (CCSM), in large part through its access to and use of the Climate Simulation Laboratory (CSL) computing facilities. CCSM output, for instance, has been used in many hundreds of peer-reviewed studies to better understand the processes and mechanisms responsible for climate variability and climate change. In addition, simulations performed with CCSM have made a significant contribution to both national and international assessments of climate, including those of the Intergovernmental Panel on Climate Change (IPCC) and the U.S. Global Change Research Program (USGCRP). CCSM provides NSF and DOE, its primary sponsors and partners in the overall USGCRP, a core modeling system for multiple purposes, including studies of past and current climate, and projections of future climate change.

More recently, and through the extensive use of the previous CSL allocation, additional capabilities have been added to the CCSM in order to address a wider range of pressing scientific questions. These include, for instance, an interactive carbon cycle in the land component and an ecosystem-biogeochemical module in the ocean component. There is also an updated atmospheric chemistry component, a global dynamic vegetation component, and land use changes due to human activity in the land component. A new version of the atmospheric component model allows scientists to study both the direct and indirect effects of aerosols on climate. The model can be run using the Whole Atmosphere Community Climate Model (WACCM), in order to better understand the role of the upper atmosphere in climate variability and change. There is also an early version of a land-ice component that can be used to simulate changes to the Greenland ice sheet and its role in future climate change (although, to date, the development of the land ice model has leveraged DOE computing resources almost exclusively). Since the most widely used description of a model with these capabilities is an “Earth System Model”, the supported model is now called the “Community Earth System Model,” or CESM. The release of Version 1.0 of CESM occurred in June 2010, and a large number of simulations with it are being conducted, many of which will contribute to the next assessment of IPCC.

This proposal for CSL resources (over the period from April 2011 through June 2012) is thus directed toward the continued testing, development and application of CESM that is required in order to meet a wide variety of community needs and keep the project at the forefront of international Earth system modeling efforts. Importantly, the transition to CESM has expanded community involvement in its development and application, and there continues to be community governance of all its activities. Accordingly, the objectives and priorities outlined in this proposal emanate directly from the community of scientists who participate in the management of the CESM project – the 12 CESM working groups and the CESM Scientific Steering Committee (SSC).

Intraseasonal to Interannual Prediction and Predictability in a Changing Climate

Project Lead: Benjamin Cash, Center for Ocean-Land-Atmosphere Studies (COLA)
Allocation: 1,950,000 GAUs
Sponsors: NOAA, NSF, NASA

The predictability of intraseasonal to interannual (ISI) climate variability is possible due to the presence of ‘memory’ in the climate system at time scales longer than those of deterministic weather prediction. This memory resides primarily in the slowly varying (relative to the time scales of synoptic weather systems) ocean and land surface conditions. While the role of the ocean, particularly the El Niño-Southern Oscillation (ENSO) in the tropical Pacific, has long been recognized as a source of predictability on ISI time scales, it is only relatively recently that the impact of land surface conditions has received substantial attention. Similarly, numerous open questions remain regarding the influence of the ocean outside of ENSO, including the role of the other ocean basins and variability at higher spatial and temporal scales.

Building upon COLA’s long-standing expertise in the investigation of ISI variability and predictability, we propose to explore further the prediction skill and predictability limits of ISI phenomena, such as the El Niño-Southern Oscillation, the Tropical Intraseasonal Oscillations, and the monsoon systems of the world through a series of numerical experiments using new versions of national climate models; namely the National Centers for Environmental Prediction (NCEP) Climate Forecast System (CFSv2) and the National Center for Atmospheric Research (NCAR) Community Climate System Model (CCSM4.0). It should be noted that several of the proposed experiments make use of only CCSM4.0. The computational cost of CCSM4.0 is substantially less than CFSv2, and it would not be practical within the resources of the CSL program to perform all experiments with both models. In the event that the CCSM4.0 experiments yield interesting results for those integrations where a complementary CFSv2 integration was not performed, additional resources will be sought to repeat the experiments with CFSv2.

Earth System Modeling (EaSM) NCAR Collaborators

Project Lead: William Large, NCAR
Allocation: 1,500,000 GAUs
Sponsors: NSF, DOE, USDA

This EaSM activity enables interagency cooperation on one of the most pressing problems of the millennium—climate change--how it is likely to affect our world, and how we can proactively plan for its consequences. This process effectively creates a group of NCAR EaSM collaborators with some common purposes and issues. We are requesting CSL resources over the months leading up to NWSC only for the following 3 critical activities:

1) Importing: Code already developed by external partners is to be ported to existing climate modeling frameworks at NCAR and tested for compatibility, utility and efficiency. It is the second of these, utility, that mostly requires the computationally demanding long climate integrations most suitable for CSL resources. It is one thing to import code that is compatible and efficient as shown by short integrations, but it is a very different matter to discover if it is useful in the sense of leading to sensible and hopefully improved solutions. The biggest challenge is to avoid long term drifts, that given some of the very long (centennial) time scales of some climate processes, are detectable only after transient responses have stabilized.

2)  Science: A limited number of science calculations are to be performed to ensure all the projects begin strong, so they can meet their ultimate deliverables within their 3 – 5 year periods of performance. It is crucial that the NCAR collaborators meet all their obligations to their projects. The science cases listed in Section 4, include the long climate integrations suitable for CSL.

3)  NWSC:In order to take full advantage of the new NWSC facility as soon as it becomes available, a number of preparatory activities should be accomplished beforehand. It is important that a number of external collaborators gain firsthand experience with NCAR-CSL based computing. This experience will lead to a much better proposal for NWSC computing, backed by a strong section of results from any allocations from this first request.

Influence of aviation aerosols and contrails on cirrus clouds and anthropogenic forcing

Project Lead: Joyce Penner, University of Michigan
Allocation: 750,000 GAUs
Sponsors: FAA

The objective of this proposal is to examine the forcing by aerosols emitted from aviation relative to other anthropogenic aerosol emissions, and to evaluate the adequacy of the aerosol treatment and cloud treatments within the models used here by comparison with data. This is a team effort funded through the peer-reviewed FAA Aviation Climate Change Research Initiative (ACCRI) program and involves science team members who will develop new observations of contrails and cirrus, as well as other modeling teams who will provide results from simulations that are similar to those proposed here (for intercomparison of different models). The primary research questions being addressed in this proposal are:

How do different modeling treatments of the emissions of aerosols from aircraft lead to different spatial distributions of aerosols, their effects on clouds and climate forcing? What is the overall forcing associated with present day and future aircraft emissions in comparison with other projected emissions? How do off-line CTM estimates of aircraft climate forcing differ from estimates from a coupled CTM with a atmospheric GCM and a full climate model? What is the strength of the coupling between clouds and aerosols from aircraft and climate change and is it possible to discern a climate impact from aircraft emissions? Answering these questions will require a large number of runs with different model set-ups. The code we use has a number of added features to both its aerosols and its aerosol/cirrus cloud interactions that are not available in the standard NCAR CAM models (CAM3, 4, or 5), and, thus, the simulations proposed here will greatly benefit from dedicated supercomputing time.

Ensemble Data Assimilation for Climate Model Development

Project Lead: Jeffrey Anderson, NCAR
Allocation: 1,500,000 GAUs
Sponsors: NSF, NASA

The proposed experiments require multiple decades of CAM and POP integrations with the added computational expense of 80-member ensemble data assimilation, and therefore require dedicated supercomputing resources. The experimental results will be used by researchers at NCAR, CCSM/AMWG and NOAA/ESRL and universities to improve the climate models used in the IPCC and National Assessment. Failure to continue investing in a DA capability for climate modeling systems places U.S. climate modelers at risk of falling behind organizations like the UK Met. Office in the development of next generation climate models. The proposed work includes ensemble data assimilation for coupled climate models. Collaboration between data assimilation scientists, atmospheric modelers, ocean modelers, climate scientists, mesoscale modelers, software engineers, and atmospheric chemists is required. The results are directly relevant to national assessments and IPCC including improving models and providing initial conditions for IPCC mandated predictions.

Development and Application of Seasonal Climate Prediction

Project Lead: David DeWitt, Columbia University
Allocation: 1,232,972 GAUs
Sponsors: NOAA

The goal of this research is to improve the prediction of seasonal climate variations, such as rainfall and near­surface air temperature, for application in climate risk management problems. Seasonal climate predictability stems from the components of the climate system with slow time scales, particularly the upper ocean and its interaction with the tropical atmosphere. Due to the atmosphere’s sensitivity to initial condition “chaos”, seasonal forecasts of precipitation and near­surface air temperature must be probabilistic if they are to be useful in applied areas such as water resource management or the forecasting of outbreaks of diseases such as malaria or dengue. The proposed experiments therefore will use relatively large Monte Carlo (or ensemble) simulations to develop probabilistic seasonal forecast systems.

US Freshwater Resources in the Coming Decades, An Integrated Climate-Hydrologic Modeling Study

Project Lead: Ying Fan Reinfelder, Rutgers University
Allocation: 750,000 GAUs
Sponsor: EPA

The modeling tool we use for this project is the Weather Research and Forecasting model (WRF) coupled to a land surface model developed by the PI (simply WRF-Hydro hereafter). The land surface model includes a groundwater module and river network module for surface runoff and groundwater input to rivers and coastal ocean. Thus, the coupled system encompasses all the reservoirs included in a standard WRF and a common land surface model, and a groundwater reservoir and coastal forcing that are not in the standard WRF. For this project, we will conduct 16 ensemble members, multidecadal simulations (2011-2050). The study domain is the N. America (including the Central America for the purpose of groundwater flow calculation) with 25km spatial resolution for the atmosphere on a Lambert Conformal Projection, totaling 2,189,500 grid points (1450x1510). The resolution is chosen to balance the high computational demand and required details. The land surface has a higher resolution of 5 km to account for heterogeneity of the surface conditions. Such high resolution for the land surface is necessary for calculating groundwater lateral flow. On a preliminary test, it will take ~6 months for an ensemble member run to complete on a 16 processor cluster, i.e., 8 years for all the ensemble runs to finish (assuming no unfinished and repeated runs)! If granted the CSL computational time, we will be able to complete the simulations well before 2012 (also the next cycle of CSL allocation). This will guarantee we have enough time for data analysis, preparing publications and reports, and transferring data for community share.