September 2012 - Allocations-admins - mailman3.cels.anl.gov

[LCRC Accounts] Project Request: rod_bundle
by accounts＠lcrc.anl.gov 26 Sep '12

26 Sep '12

Hello, A new project on the LCRC cluster has been requested. Please forward the information on to the LCRC Allocation sub-committee. Applicant's name: Justin Walker Applicant's institution: ANL Applicant's division: MCS Project Name: rod_bundle Project title: Thermal hydraulics for tight rod bundle lattices Associated funding: Other Systems: MCS general compute servers, cosmea - unlimited Intrepid - 35525491.6 cpu hours Science: This project is an auxiliary project in support of computational fluid dynamics(CFD) calculations of the thermal hydraulics of nuclear reactor cores. Within the framework of the spectral element method, we are studying the effects of different large eddy simulation(LES) models and resolution levels on the fidelity of turbulent flow calculations. In particular, we would like to know what level of resolution is required to obtain data useable to the engineering and science communities. It is important to investigate this on smaller problems before moving to a full scale calculation on a machine like Intrepid, since it allows us to achieve better results while we push the limits of CFD. Project description: We plan to perform LES and direct numerical simulation(DNS) of turbulent flow past a single rod in a periodic domain, corresponding to a rod in an infinite lattice. This model problem closely resembles the conditions for a central rod in the large arrays found in nuclear reactor designs. From these calculations, we hope to obtain a measure of the resolution required to match published experimental results of important quantities such as the Reynolds stresses and wall shear stress. To perform these calculations, we will be using the computational fluid dynamics solver nek5000 which is based on the spectral element method. This code has already been validated and shown to scale very well by use on this machine as well as ALCF's Intrepid and MIRA. See http://nek5000.mcs.anl.gov/index.php/Scaling. We will be running jobs of no more than 8 nodes, since we've experienced problems with greater numbers of nodes on this machine. Job sizes will primary be about 1GB per core and run for about 60 hours each. Some will require longer, up to 120 hours. There will be up to four project members but the primary use of the allocation will be by just one of the members. Project URL: Requested allocation: 200000 Q1: 50000 Q2: 50000 Q3: 50000 Q4: 50000 Justification: The requester has used undetermined amount hours of their initial startup project. In addition to approving an initial amount, please specify a Category and Subcategory for this project. For a list of the current selection of approved categories, please see: https://wiki.lcrc.anl.gov/wiki/Processes/Categories Once the Allocation committee has approved the project, please go to the Project Management page to create it: https://accounts.lcrc.anl.gov/projects.php Thank You, The LCRC Accounts System

1 0

[LCRC Accounts] Yearly Allocation Request from POPAD
by accounts＠lcrc.anl.gov 25 Sep '12

25 Sep '12

Hello, A yearly allocation for the LCRC cluster has been requested with the following updated information: Submitter/PI: Jong Kim Project Name: POPAD Division: MCS Project title: Adjoint Development of POP Ocean Model for Climate Modeling Applications Associated funding: DOE CCPP Program: IMPACTS Project Other Systems: Science: The main purpose of the DOE IMPACTS project is to quantify the risks of abrupt climate change during the 21st century. The basic research effort is focused on assessing the near-term prospects for abrupt climate change on decadal time scales. The investigation is based on the NCAR-DOE Community Climate System Model (CCSM). It involves various types of new developments to examine several potentially significant sources of abrupt climate change. The role of ANL is to enhance the CCSM POP Ocean model with tangent linear and adjoint modeling (TLM/ADM) capabilities to allow a detail sensitivity analysis of the POP ocean model. Project description: In realizing the importance of the meridional overturning circulation (MOC) to the global climate variations, an essential element is to fully understand the MOC sensitivity to various forcing variables and modeling parameters. However, there are significant uncertainties in the model configurations such as mixing parameter values and initial and boundary conditions. More systematic computational tools are desirable for robust and efficient modeling experiments. Unfortunately, the computational costs to run large-scale sensitivity studies with conventional approaches such as ensemble approaches or perturbed numerical difference are prohibitively high. As an alternative to the conventional approaches, automatic differentiation (AD) has been applied to enhance the POP ocean model with the derivative computations of the TLM/ADM methods. In the previous year, major source codes of the AD-based POP TLM/ADM methods were successfully implemented and optimized f or short runs of a low resolution model. We plan to continue further applications of the new codes for the sensitivity studies of POP modeling variables with regard to global ocean circulations. In FY2013, the experiments for longer run simulations of the POP ocean models will be mainly performed with 1 degree resolution for the MPI-based parallel run test. It requires to run various types of the CCSM coupled and TLM/ADM simulations. Generally, the multi-decadal integrations of a coupled CCSM system for low resolution problems require a single computational time on the 8 cores of the fusion. The POP TLM/ADM runs take about 10 times of the computational cost compared to the forward POP integrations. Mainly written in Fortran 90, each of modeling component of the CCSM modeling system requires a C compiler and the MPI library to run with a link with Perl script. Project URL: Current FY Hours Used: undetermined amount New FY Requested allocation: 8000 Q1: 2000 Q2: 2000 Q3: 2000 Q4: 2000 Justification: Thank You, The LCRC Accounts System

1 0

[LCRC Accounts] Yearly Allocation Request from 7_SEAS_WRF
by accounts＠lcrc.anl.gov 24 Sep '12

24 Sep '12

Hello, A yearly allocation for the LCRC cluster has been requested with the following updated information: Submitter/PI: Rao Kotamarthi Project Name: 7_SEAS_WRF Division: DIS Project title: Forecasting for the 7-SEAS Study Associated funding: NASA, DOE Other Systems: None Science: SEAC4RS will address the influence of Asian emissions on atmospheric chemistry, clouds, climate and air quality, with specific focus on studying the redistribution of atmospheric composition under Asian monsoon circulation and convection conditions. The primary goals of the experiment are to provide key observations to help validate satellite observations under these conditions, and to advance our understanding of the role of deep convection in redistributing pollutants emitted in the tropics and the subsequent influence of these pollutants on the chemistry of the upper troposphere, and the influence of feedbacks between the pollutants and local meteorology through direct, semi-direct and indirect radiative effects. The successful execution of the mission and the interpretation of the observations to meet the science objectives require contributions from Chemical Transport Models (CTMs). CTMs play important roles in the design, execution, and analysis of large-scale atmospheric chemistry field studies. They are used in forecast-mode to enhance flight planning by enabling the representation of important three-dimensional atmospheric chemical structures (such as dust storm plumes, polluted air masses associated with large cities, widespread biomass burning events, and large scale convective redistribution events) and how they evolve over time. CTM forecasts play the additional important role of providing a 4-dimensional contextual representation of the experiment. CTMs also facilitate the integration of the different measurements and measurement platforms (e.g., aircraft, ground stations and satellite observations). Finally, CTMs are used to help evaluate and improve emission estimates. Project description: We have developed an operational regional-scale forecasting and analysis system to assist in atmospheric field experiments. The system consists of 3 major components: 1) detailed mesoscale meteorological model with an off-line CTM and a wide variety of air mass and emission tracers; 2) detailed 3-dimensional photochemical and radiative transfer calculations using a CTM; and 3) enhanced emission products that intimately link emitted amounts and activities to the transport and chemistry analysis. This system was successfully applied in the design and execution of the NASA TRACE-P, INTEX A & B and ARCTAS experiments, the NSF ACE-Asia, MILAGRO, Pacific Dust Experiment (PACDEX) and VOCALS REx experiments, and the NOAA ITCT-2K2 and ICARTT intensive field experiments. In these previous applications we have demonstrated that these models can be used effectively in the execution and analysis of field experiments. Forecast products, emissions and publications from these previous missions can be found at http://www.cgrer.uiowa.edu/ACESS/acess_index.htm. We plan to generate SEAC4RS with once per day WRF-Chem forecasts of hourly meteorology and chemical transport with fully coupled feedbacks. The WRF-Chem forecasts will be of particular use in this experiment as the model is well-suited for simulating regional-scale redistribution by convection (see Section 2.3 for examples) and evaluating feedbacks between pollution and meteorology. The on-line coupling allows better temporal resolution and coupling between the processes, and the higher spatial resolution that we plan to use will directly resolve cloud processes. These features in WRF-Chem will be used to generate a range of new mission support products tailored to specific objectives and needs of the SEAC4RS experiment, such as convective indices, convective outflow regions, and indices of locations and time periods where new particle activation and aerosol radiative forcing are strongest. Project URL: http://espo.nasa.gov/missions/seac4rs Current FY Hours Used: undetermined amount New FY Requested allocation: 450000 Q1: 150000 Q2: 150000 Q3: 100000 Q4: 50000 Justification: Thank You, The LCRC Accounts System

1 0

New LCRC report submitted for NuclearStructure
by support＠lcrc.anl.gov 24 Sep '12

24 Sep '12

Full project report at https://accounts.lcrc.anl.gov/reporting/project/2012/?project=NuclearStruct…

1 0

[LCRC Accounts] Yearly Allocation Request from metal_clusters
by accounts＠lcrc.anl.gov 12 Sep '12

12 Sep '12

Hello, A yearly allocation for the LCRC cluster has been requested with the following updated information: Submitter/PI: Glen Ferguson Project Name: metal_clusters Division: MSD Project title: Catalytic Activity of Subnanometer Metal Clusters Associated funding: Competitive LDRD Other Systems: CNM Carbon Cluster 200,000 core-hours, ALCF Intrepid ~1,000,000 core-hours, EMSL Chinook ~200,000 core-hours Science: The production of classical and next generation fuels using highly active, selective and environmentally safe catalysts is a major challenge that, if met, would provide one approach to meeting future energy needs. This challenge, along with others, is a strong driving force for developing catalysts that are more effective. Toward this goal, there is a desire for the rational design of catalysts with a high degree of selectivity and activity. An important, recently published, advancement is the use of monofunctional surface supported subnanometer metal clusters that catalyze several important reactions of small hydrocarbons, i.e., the oxidative dehydrogen of propane or separately the epoxidation of propene. These advancements open the door for new class of catalysts, constructed by rational design, with high activity and selectivity at relatively low temperatures. Two important directions to explore that may improve and extend the usefulness of these catalysts are th e design of bifunctional catalysts and the effect of the surface support on the catalyst's reactivity. Our work in these areas will focus on the use of both density functional and ab initio calculations to predict the reactivity of bimetallic subnanometer alloys and how the surface support affects their catalytic activity and selectivity. One problem is the prediction of catalysts to perform the stepwise dehydrogenation and partial oxidation of propane to propylene oxide. Currently, there is no bifunctional catalyst that forms partially oxidized propene starting from propane. The possibility of a bifunctional metal alloy cluster catalyst gives a unique opportunity to accomplish this goal in a stepwise fashion exploiting the reactivity of the two catalytic sites of the alloy. The alumina (Al2O3) surface was shown to have a central role in the reactivity of supported Ag3 clusters with propene to form propylene oxide. Specifically, the surface was a key player in the dissociation of oxygen, which is the first step in the partial oxidation of propene. This surface reactivity indic ates the possibility the surface support is important in other reactions. We will therefore investigate other surface supports to understand their effects on the catalytic activity of subnanometer bimetallic alloys. These calculations will be closely coupled with experimental work to produce and test the predicted catalysts and surface supports. Project description: New Results and Goals: We have investigated the use of Co4O2 on alumina as a model system for the production of cyclohexene from cyclohexane using periodic density functional calculations. During the course of the investigation it was determined that the oxidation state of the cluster on the surface is Co4O4 and not Co4O2. This required a change in focus to the cluster with the new oxidation state. The previous mechanism of oxidative dehydrogenation of cyclohexane to cyclohexene was reexamined and found to be unsupported for this oxidation state. The higher oxidized clusters are not as reactive as the Co4O2 cluster but consequently are not as easily poisoned by the reaction products. Using this information we have calculated a new pathway for cyclohexane dehydrogenation. The reaction can be represent by the following scheme. Cyclohexene formation Co4O4 + H2O + C6H12 → Co4O3 + C6H10 + 2H2O Oxygen defect passivation Co4O3 + ½O2 → Co4O4 This pathway is distinct from others calculated in that the surface is required as a reactant. The inclusion of water in the reaction mechanism is supported by the strongly exothermic nature of water reacting with oxide surfaces. This work is currently being prepared for submission. While a highly active dehydrogenation catalyst is a major discovery these results can be extended using catalyst design methods. While these methods were generally used only for catalytic surfaces our group has recently extended them for use with subnanometer cluster, with the results presented in an invited review in the journal Topics in Catalysis. Our method requires calculating binding energies and estimating activation barriers for various clusters develop volcano type relationships for the same reaction on new subnanometer clusters. The primary codes used in these calculations are VASP and Gaussian 09. Both of which are currently available on Fusion. While it is often acceptable to omit the surface significantly reducing the computational cost, as is indicated above, this is not possible for our systems, therefore, the resources needed for each binding energy is ~ 1000 core hours with over 200 binding energies needed. To generate the necessary data to estimate barrier a nother 100,000 hours are necessary. One of the essential tasks of using catalyst design methods for subnanometer clusters is finding the minimum energy subnanometer clusters that correspond to the experimentally observed clusters. Previously we have attempted to use the genetic algorithm of Johnston and coworkers. However, after implementation this algorithm was found to be insufficient to locate the global minima of high spin clusters used in this work. We have instead moved to the new method, called the coalescence kick method of Boldyrev and coworkers (Utah State University). In this method the clusters are randomly generated then formed into fragments and coalesced followed by optimization. The optimization is performed using a standard code but the first steps are performed using novel code that forms and coalesces the clusters. Preliminary tests using the lithium peroxide trimer and Co4O4 as test molecules indicate that the code is currently not suitable for use in our applications but indicate which imp rovements are necessary to make the method useful. For the case of lithium peroxide trimer the minima found via coalescence kick differed from the lowest minimum found by brute force by the interchange of two atoms. This immediately suggested that adding functionality to the code to perform atom interchanges would fix this problem. While the coalescence kick program is not computationally demanding the structures produced are optimized to the nearest geometric local minimum using the Gaussian 09 suite of programs. To test the additional functionality to the code would require an estimated 50,000 core hours. Another problem with this approach, revealed by attempts to find the global minimum for Co4 and Co4O4 clusters, is that random starting configurations of metal oxide and metal clusters cannot be efficiently optimized using Gaussian 09. To alleviate this problem the code will be modified to use the VASP program. Preliminary calculations performed by modifying the output of the coalescence kick program to run using VASP optimized efficiently showing the global minimum produced by brute force calculations. To validate the addition of using VASP as the optimizer for the clusters will require 100,000 core hours. Project URL: Current FY Hours Used: undetermined amount New FY Requested allocation: 400000 Q1: 100000 Q2: 100000 Q3: 100000 Q4: 100000 Justification: Thank You, The LCRC Accounts System

1 0

[LCRC Accounts] Project Request: Hydro-model
by accounts＠lcrc.anl.gov 11 Sep '12

11 Sep '12

Hello, A new project on the LCRC cluster has been requested. Please forward the information on to the LCRC Allocation sub-committee. Applicant's name: Eugene Yan Applicant's institution: ANL Applicant's division: EVS Project Name: Hydro-model Project title: Hydrologic modeling with GRACE Associated funding: NASA Other Systems: MCS UNIX machines Science: Improve assessment of water resources and their response to climate change: case study for Africa continent. Project description: Adjust parameters for hydrologic processes in the land surface model (CLM with CESM) with remote sensing data and recent GRACE gravity measurements. Community Land Model (CLM4) offline and CLM4 coupled with CESM will be used to simulate hydrologic processes for Africa. We will test CLM4 with our measured data from GRACE. CLM4 model has been used on fusion by other Argonne colleagues. The new version is quite efficient. Initially, we will have 1-2 members to test. later, we will have 2-3 members to run simulations for different scenarios and various calibrations for number of parameters. Project URL: Requested allocation: 321000 Q1: 1000 Q2: 20000 Q3: 100000 Q4: 200000 Justification: The requester has used undetermined amount hours of their initial startup project. In addition to approving an initial amount, please specify a Category and Subcategory for this project. For a list of the current selection of approved categories, please see: https://wiki.lcrc.anl.gov/wiki/Processes/Categories Once the Allocation committee has approved the project, please go to the Project Management page to create it: https://accounts.lcrc.anl.gov/projects.php Thank You, The LCRC Accounts System

1 0

New LCRC report submitted for InversedCatalysts
by support＠lcrc.anl.gov 04 Sep '12

04 Sep '12

Full project report at https://accounts.lcrc.anl.gov/reporting/project/2012/?project=InversedCatal…

1 0

[LCRC Accounts] Project Request: MDcounterions
by accounts＠lcrc.anl.gov 04 Sep '12

04 Sep '12

Hello, A new project on the LCRC cluster has been requested. Please forward the information on to the LCRC Allocation sub-committee. Applicant's name: Lynda Soderholm Applicant's institution: ANL Applicant's division: CSE Project Name: MDcounterions Project title: Competing solvent and counter ion coordination and its effect on metal hydration properties: a polarizable molecular dynamics simulation Associated funding: DOE, OBES, Chemical Sciences Other Systems: We currently have access to the French GENCI CINES computer center, with 200,000 core-hours for calendar year 2012. Science: A detailed description of a metal-ion’s coordination environment in aqueous solution, specifically its coordinating ligands, their numbers and their distances, is a necessary prerequisite to the development of a predictive understanding of the system’s physical and chemical behavior. At the core of this chemistry is the understanding of how a specific ionic configuration around the metal takes place, and how the other solution components, including other dissolved cations, counter ions, and the solvent, influence the structure and dynamics of the targeted species. Historically this knowledge has been hampered both by a lack of an experimental probe able to provide the necessary metrical data and by the absence of computer codes with enough atomistic detail to provide the accuracies necessary to distinguish the small differences significantly impacting the chemistry. This situation has recently changed from the experimental perspective with the av ailability of high-energy (> 60 keV) X-rays that can be used in total scattering (HEXS) experiments to probe atom-atom correlations in solutions out to distances of 1 nm or more. This capability to routinely collect such high quality data is currently unique to the APS. Lead by ANL’s Heavy Elements group, analyses of solution HEXS data suggest that there may be an energetically important competition between the solvent and the counter ions in the metal-ion vicinity that is currently ignored by high-level calculations attempting to reproduce solution behaviors. Depending on the nature of the counter ion, it may either enter the first hydration sphere, with or without replacing first-sphere solvent molecules, or remain in the second coordination sphere. The latter case is particularly interesting. Experiments suggest that the primary metal-anion interaction may include a cluster formed with its own structured hydration shell. This would add significant, but perhaps necessary, complications to the calculations because such an interaction would indirectly perturb the first-hydration sphere interactions, eventually inducing a change in the structural and temporal properties of that hydration sphere. This behavior has neither been previously rec ognized experimentally nor studied theoretically. It is currently unclear what factors determine the competition between solvent molecules and counter ions, and which interactions drive the molecular-level structure and dynamics. The calculations are necessarily too large to be applied in the comprehensive manner required to address this problem. A unifying molecular picture of ion hydration and the directing role of anions and their clusters represents a potentially break-through improvement to current knowledge. We request computer time to complement our ongoing experimental efforts to provide a fundamental basis upon which to build predictive capabilities. Project description: In an attempt to explain, on a molecular level, the heuristic ordering of a solution in which lanthanide/actinide cations, solvent molecules and anions are present, we will employ molecular dynamics modeling to simulate HEXS data. We will use the POLARIS(MD) classical molecular dynamics code developed by our collaborating theorist in Paris (M. Masella, CEA Saclay), in which we have developed the parameters of a many-body polarizable force-field model. This force-field is more sophisticated than common polarizable force-fields as it accounts for all the subtle many-body interactions taking place in a solute-solvent-anion system, including hydrogen bonding and covalent interactions, the latter of which is in the form of a charge-transfer term. The force-field parameters are adjusted to reproduce highly accurate quantum chemical data on representative hydrated clusters. Working together with our collaborative experimentalists, we have chosen to focus on f-ion bromide solutions. In the lanthanide series, we will initially focus on La(III), Eu(III), Er(III), and Yb(III), while our actinide element sets consist of thorium(IV) and curium(III). Much of the experimental data is already available. Parameters for Ln/An-water, water-bromide, and anion-bromide interactions have been developed, and are available for running nanoseconds molecular dynamics simulations with periodic boundary conditions for various solvated models: 1/ for a given Ln/An element, we wish to investigate the effect of the counter ions by comparing the results of simulations conducted without and with counter ions present in the simulation box. In the latter, the counter ion will be placed in the “second coordination shell” as experiment predict, to start from a structural configuration close to what is observed in solution. 2/ for a given Ln/An-bromide system, we will explore the changes of free energy as a function of the Ln/An-bromide distance (potential of mean-force estimated from the results 10 trajectories at fixed Ln/An-bromide distances), to see if and where it reaches a minimum that would correspond to the formation of contact ion pairs. Our specific focus will be a competition for metal ion coordination between a bare water molecule and a water molecule bound in an anion complex. The MD simulations will be used to generate atomistic pair-correlations that can be directly compared with experiment. Code to transform the MD simulations into HEXS pdf spectra has been written by the ANL group and is currently available. The fundamentally new aspect of this work is that these comparsions can be made, with significant accuracies, out to distances of about 0.8 nm in real space. Such comparisons will play a critical role in assessing the force-fields used in the simulations. Estimated computer usage: Requested computer time. Altogether we wish to study 6 metal ions, with various numbers of counter-ions, summing up to about 50 trajectories. The PMF calculations require 10 trajectories per metal-bromide couple, and various numbers of counter ions, adding up to about 200 trajectories. Altogether the requested time sums up to 600,000 core hours. To test the accuracy of the quantum chemical reference data (MP2-level), we would like to perform a couple of benchmark calculations using the latest coupled-cluster implementations in the NWChem code. Given the size of the system, we expect to run these calculations across several nodes (probably 4). Each calculation may request up 24-48 hours walltime. To be conduct, we request 80,000 core hours. In total the project sums up to 680,000 core hours. We do not plan to run non-parallel jobs. Project URL: Requested allocation: 680 Q1: 200 Q2: 160 Q3: 160 Q4: 160 Justification: Each trajectory described above should be run on a 5 ns time scales to reach the equilibrium of the system. Typically for a simulation box size of a cation/anion system solvated by 1000 water molecules, the calculation requests 60 hours (walltime) on 32 processors, that is about 2000 core hours. The POLARIS(MD) code is currently parallelized with MPI scheme. In its current version it well up to about 32 processors, but its scalability drops down to 50 % when reaching about 100 CPUs. This is why we ideally run these calculations on 32 processors on the FUSION platform. However, with the help of Intel scientists and the French Exascale Research Laboratory (ERL), the code is currently being rewritten in parts to adapt to forthcoming exascale multicore computers. A new version will be ready by the end of September. The requester has used undetermined amount hours of their initial startup project. In addition to approving an initial amount, please specify a Category and Subcategory for this project. For a list of the current selection of approved categories, please see: https://wiki.lcrc.anl.gov/wiki/Processes/Categories Once the Allocation committee has approved the project, please go to the Project Management page to create it: https://accounts.lcrc.anl.gov/projects.php Thank You, The LCRC Accounts System

1 0

New LCRC report submitted for Parallel_Fusion_PDEs
by support＠lcrc.anl.gov 01 Sep '12

01 Sep '12

Full project report at https://accounts.lcrc.anl.gov/reporting/project/2012/?project=Parallel_Fusi…

1 0

[LCRC Accounts] Yearly Allocation Request from foam
by accounts＠lcrc.anl.gov 31 Aug '12

31 Aug '12

Hello, A yearly allocation for the LCRC cluster has been requested with the following updated information: Submitter/PI: Robert Jacob Project Name: foam Division: MCS Project title: Global Climate Modeling with the Fast Ocean Atmosphere Model Associated funding: Department of Energy, Office of Science National Science Foundation Other Systems: external: NERSC, NCAR Science: Climate models treat the atmosphere and ocean as interacting fluids. They contain numerical methods to solve the basic partial differential equations which are in turn derived from the Navier-Stokes equations for a fluid on a rotating sphere. They also contain detailed parameterizations to compute the internal and external forcing terms that come from such diverse phenomena as the passage of radiation through the atmosphere, the release of latent heat by phase changes of water and the effects of friction and unresolved turbulent scales. In general, climate model development is focused on increasing the fidelity of the simulation compared to the observations. On long time scales, the climate is controlled mostly by the ocean because it is a slower, more dense fluid then the atmosphere and has a higher heat capacity. For some problems in climate science, such as understanding the different climates of the geologic past and the internal variability of the climate, it may not be necessary to have the most accurate representation of the atmosphere in the climate model being used. Project description: Our major objective is to test the application of the Ensemble Kalman Filter to FOAM for tuning model parameters in a coupled mode. In spite of decades of effort, coupled ocean-atmosphere general circulation models (OAGCM) like FOAM still suffer from significant model biases, notably in the tropics. A significant part of these biases is caused by uncertainties in the model parameters. So far, however, parameter tuning in OAGCMs remains rudimentary. We will use our time to develop a strategy of a systematic parameter optimization in an OAGCM. In Fy12, we have successfully developed adaptive coupled ensemble Kalman Filter system (AcEnKF) for a simultaneous estimation of model parameter and state. We have optimized a parameter – solar penetration depth by assimilating monthly mean observation of sea surface temperature (SST) and sea surface salinity (SSS) in the perfect model framework. Next, we will further investigate AcEnKF in FOAM. Based on our previous work, the parameter estimation is sensitive to ensemble size and some updating technics. And it is also sensitive to the response of model to the parameters. We will further investigate the AcEnKF in FOAM this year: -We will investigate multi-parameter estimations in FOAM. Some atmosphere and coupler parameters will be estimated in our perfect model experiments. -We will further investigate the effect of observation quality and ensemble size to the parameter estimation. For this request, we'll assume fifty members will be used for each experiment. We expect about 150 different sets of experiments for parameter estimation. Each integration should take 30 years such that the total model years are: Parameter estimation experiments: 50x30x150 =225,000 years. Good parameter estimation requires good data assimilation. We also need to refine our pure data assimilation skills that will provide support for us to estimate parameter using real observation in the future. - We plan to use our newly built FOAM data assimilation system to do some sensitivity tests by using the real ARGO observational network to further check if we can reconstruct the ocean state, especially, the sub-surface ocean. - Given the initial condition we get from the first stage, we will do some forecast experiments to check the predictability of longer-term variability in FOAM model. - The second stage we will imply the real observation to the system to produce a data assimilation and decadal forecast product. We expect about 60 different sets of experiments for data assimilation. By average, each experiment takes 30yr integration with an ensemble size of 50. Data assimilation experiments: 50x30x60 =90,000 years We have measured a cost of 2.133 core-hours/simulated year leading to a request of 640000 core-hours. This gives us the total hours for model experiments as Total cpu for experiments: (225000+90,000)yearsx2.133 = 692K core-hours With some minor time for testing, we request a total of 700k core-hours for FY13. Project URL: http://www.mcs.anl.gov/foam Current FY Hours Used: undetermined amount New FY Requested allocation: 700000 Q1: 150000 Q2: 150000 Q3: 200000 Q4: 200000 Justification: We have measured a cost of 2.133 core-hours/simulated year. The total experiments are independent and can be run simultaneously. FOAM is very scalable between 4 and 64 processors. We choose to run an individual ensemble member on 16 or 32 because the additional parallelism comes from the ensemble members. We could effectively use the entire machine if it was available. Thank You, The LCRC Accounts System

1 0