Difference between pages "Presentations and Posters" and "Optimisation"
From UKCA
(Difference between pages)
FionaOConnor (talk | contribs) |
|||
| Line 1: | Line 1: | ||
| + | =Model Optimisation= |
||
| − | ==Presentations== |
||
| − | * {{pdf|PB_MIPAS_DUM.pdf|Confronting UKCA with MIPAS Observations}} (P. Braesicke) |
||
| − | * {{pdf|PB_UKCA_UMUM.pdf|The United Kingdom Chemistry and Aerosol Community Model (UKCA)}} (P. Braesicke) |
||
| − | * {{pdf|Telford_NCAS_0812.pdf|Effects of Climate Induced Changes in Isoprene Emissions After the Eruption of Mt Pinatubo}} (P. Telford) |
||
| − | * {{pdf|EGU_Vienna_200803XX.pdf|Reassessment of the Atmospheric Response to the Pinatubo eruption using a Nudged CCM}} (P. Telford) |
||
| − | * {{pdf|Seminar_Cambridge_20080225.pdf|Producing a "Nudged" version of the UKCA CCM}} (P. Telford) |
||
| − | * {{pdf|Telford_RMS_20070906.pdf|Development of a Nudged CCM}} (P. Telford) |
||
| − | * {{pdf|QUEST_Annual_Science_Meeting_DMS.pdf|Influence of oceanic DMS emissions on CCN concentrations over the Southern Ocean: A global model study}} (K. Carslaw) |
||
| − | * {{pdf|NCAS_aerosol_strategy.pdf|The UK Chemistry and Aerosol Project (UKCA)}} (K. Carslaw) |
||
| − | * {{pdf|NCAS_2008.pdf|Global aerosol and climate: Why do we care about microphysics?}} (K. Carslaw) |
||
| − | * {{pdf|AeroCom_Mann_NPF.pdf|Global aerosol microphysics modeling: Implications of new particle formation & growth for aerosol-climate simulations}} (G. Mann) |
||
| − | * {{pdf|Mann_QUEST_dust.pdf|Global size-resolved dust modeling with GLOMAP}} (G. Mann) |
||
| − | * {{pdf|Mann_Leeds.pdf|Enhanced UK capability in global aerosol modeling}} (G. Mann) |
||
| − | * {{pdf|NCAS_Dec08_Mann_UKCA.pdf|The UK Chemistry and Aerosol Model (UKCA): An NCAS-Met Office collaboration. Advancing the realism of aerosol-climate effects in UK climate model simulations}} (G. Mann) |
||
| − | * {{pdf|JamieRaeSeminar_14_10_08.pdf|Climate-chemistry-aerosol coupling for HadGEM2-ES}} (J. Rae) |
||
| − | * {{pdf|Ukca_rms_4sep2007.pdf|UKCA Tropospheric Chemistry: Evaluation and Sensitivity to Climate Change}} (F. O'Connor) |
||
| − | * {{pdf|Ukca_as_part_of_hadgem2es.pdf|Introduction to HadGEM2-ES}} (F. O'Connor) |
||
| − | * {{pdf|Aqum_savage.pdf|Application of the Met Office Unified Model for Air Quality Forecasting}} (Nick Savage) |
||
| + | The cost of the Stratosphere-Troposphere chemistry scheme in UKCA using the Newton-Raphson solver with the on-line photolysis scheme Fast-jX relative to the climate model HadGEM3-A is as follows: |
||
| + | |||
| − | ==Posters== |
||
| + | {| border="1" |
||
| + | ! Model || PEs ||OpenMP Threads ||Time Elapsed (sec) |
||
| + | |- |
||
| + | |HadGEM3-A || 8x16 || 1 || 3798 |
||
| + | |- |
||
| + | |HadGEM3-A + StratTrop(N-R) + Fast-jX || 8x16 || 1 || 15602 |
||
| + | |- |
||
| + | |HadGEM3-A || 8x16 || 2 || 2328 |
||
| + | |- |
||
| + | |HadGEM3-A + StratTrop(N-R) + Fast-jX || 8x16 || 2 || 11730 |
||
| + | |- |
||
| + | |} |
||
| − | * {{pdf|SPARC08_Age_PB.pdf|Zonal Asymmetries in Age-of-air and their Relevance for Transport into the Subtropical Lowermost Stratosphere}} (P. Braesicke) |
||
| − | * {{pdf|SPARC08_WWind_PB.pdf|The Role of Tropical Vertical Velocities in Determining Tracer Gradients in the Upper Troposphere and Lower Stratosphere}} (P. Braesicke) |
||
| − | * {{pdf|Telford_Annecy_20080901_poster.pdf|Impact of the Pinatubo eruption on Isoprene}} (P. Telford) |
||
| − | * {{pdf|Telford_Manchester_20080108_poster.pdf|Development and Assessment of a Nudged Version of the UKCA Model}} (P. Telford) |
||
| − | * {{pdf|NCAS_ASM_York_poster.pdf|Developments in Global Aerosol Modelling}} (K. Carslaw, G. Mann) |
||
| − | * {{pdf|GMann_IGAC_poster_Sep08.pdf|Realistic aerosol simulations with the UKCA composition-climate model}} (G. Mann) |
||
| − | * {{pdf|Poster_JamieRae_Nov08_Edited.pdf|Achieving an improved sulphur cycle in the HadGEM model through online calculation of oxidant concentrations}} (J. Rae) |
||
| − | * {{pdf|Hamburg_poster_foc.a4.pdf|Effect of Climate Change on Tropospheric Oxidants}} (F. O'Connor) |
||
| − | * {{pdf|Igac.pdf|Sensitivity of a coupled chemistry-climate model to climate model biases}} (F. O'Connor) |
||
| − | * {{pdf|Ncgg5_poster_foc.a4.pdf|Methane Wetland Emissions: Influence of the El Niño Southern Oscillation}} (F. O'Connor) |
||
| − | * {{pdf|Ccmval_agu_hardiman.pdf|Preliminary investigation of ozone recovery and climate change in the 21st century using the UM-UKCA climate-compostion model}} (S. Hardiman) |
||
| − | * {{pdf|CO_EGU_2009_Edited.pdf|Regional air quality forecasting in the Met Office Unified Model}} (Carlos Ordonez) |
||
| − | * {{pdf|MW_RMS.pdf|Development of the Met-UM for Air Quality Forecasting}} (Mark Weeks) |
||
| − | * {{pdf|NLA_NCAS_20091110-11.pdf|Chemistry in the Unified Model: Climate-Chemistry Studies with UKCA}} (Luke Abraham) |
||
| − | * {{pdf|NCAS2009_N2O_PB.pdf|Tracer PDFs as touchstones for CCM validation: The N2O example}} (Peter Braesicke) |
||
| + | On a 8x16 PE configuration and 1 OpenMP thread on the Monsoon facility or the Met Office's Power6 IBM (hpc1e/1f), UKCA and Fast-jX together add 310% to the cost of HadGEM3-A. However, with 2 OpenMP threads and no OpenMP compiler directives in UKCA, the relative cost of UKCA is even higher, adding approximately 400% to the cost of HadGEM3-A. Adding aerosol chemistry and UKCA-MODE aerosols will make it even more costly. Therefore, there is a clear need for optimisation. |
||
| − | ==UKCA Launch Meeting Presentations== |
||
| + | As part of the HadGEM3-ES development project, currently being led by F. O'Connor, there are plans to do some optimisation work on UKCA. In particular, a more complete assessment of the model cost will be carried out and the potential speedup which may be gained from simple code re-writing, load balancing, and the use of OpenMP, dedicated I/O servers and maths libraries will be explored. Some scientific optimisation such as throwing out unwanted reactions (or species), tweaking chemistry to improve convergence, etc. |
||
| − | * John Pyle: {{pdf|01_jp_ukca.pdf|Introduction}} |
||
| + | should also be considered. The use of an alternative solver, such as a Rosenbrock solver, may also be investigated. |
||
| − | * Colin Johnson: {{pdf|02_cj_ukca.pdf|UKCA as a component of the UM}} |
||
| − | * Fiona O'Connor: {{pdf|03_foc_ukca.pdf|Tropospheric chemistry modelling}} |
||
| − | * Ken Carslaw: {{pdf|04_kc_ukca.pdf|Aerosol schemes for UKCA}} |
||
| − | * Peter Braesicke: {{pdf|05_pb_ukca.pdf|Stratosphere and whole atmosphere modelling}} |
||
| − | * Nick Savage: {{pdf|06_ns_ukca.pdf|Application of UKCA in an air quality forecast model}} |
||
| − | * Peter Braesicke (on behalf of Luke Abraham): {{pdf|07_pb_for_la_ukca.pdf|Using UKCA}} |
||
Revision as of 10:06, 19 April 2012
Model Optimisation
The cost of the Stratosphere-Troposphere chemistry scheme in UKCA using the Newton-Raphson solver with the on-line photolysis scheme Fast-jX relative to the climate model HadGEM3-A is as follows:
| Model | PEs | OpenMP Threads | Time Elapsed (sec) |
|---|---|---|---|
| HadGEM3-A | 8x16 | 1 | 3798 |
| HadGEM3-A + StratTrop(N-R) + Fast-jX | 8x16 | 1 | 15602 |
| HadGEM3-A | 8x16 | 2 | 2328 |
| HadGEM3-A + StratTrop(N-R) + Fast-jX | 8x16 | 2 | 11730 |
On a 8x16 PE configuration and 1 OpenMP thread on the Monsoon facility or the Met Office's Power6 IBM (hpc1e/1f), UKCA and Fast-jX together add 310% to the cost of HadGEM3-A. However, with 2 OpenMP threads and no OpenMP compiler directives in UKCA, the relative cost of UKCA is even higher, adding approximately 400% to the cost of HadGEM3-A. Adding aerosol chemistry and UKCA-MODE aerosols will make it even more costly. Therefore, there is a clear need for optimisation.
As part of the HadGEM3-ES development project, currently being led by F. O'Connor, there are plans to do some optimisation work on UKCA. In particular, a more complete assessment of the model cost will be carried out and the potential speedup which may be gained from simple code re-writing, load balancing, and the use of OpenMP, dedicated I/O servers and maths libraries will be explored. Some scientific optimisation such as throwing out unwanted reactions (or species), tweaking chemistry to improve convergence, etc.
should also be considered. The use of an alternative solver, such as a Rosenbrock solver, may also be investigated.
