Improving Parallel Performance of a Finite-Difference AGCM on Modern High-Performance Computers

The paper introduces a moist, deterministic test case of intermediate complexity for Atmospheric General Circulation Models (AGCMs). We suggest pairing an AGCM dynamical core with simple physical parameterizations to test the evolution of a single, idealized, initially weak vortex into a tropical cyclone. The initial conditions are based on an initial vortex seed that is in gradient‐wind and hydrostatic balance. The suggested “simple‐physics” package consists of parameterizations of bulk aerodynamic surface fluxes for moisture, sensible heat and momentum, boundary layer diffusion, and large‐scale condensation. Such a configuration includes the important driving mechanisms for tropical cyclones, and leads to a rapid intensification of the initial vortex over a forecast period of ten days. The simple‐physics test paradigm is not limited to tropical cyclones, and can be universally applied to other flow fields. The physical parameterizations are described in detail to foster model intercomparisons. The characteristics of the intermediate‐complexity test case are demonstrated with the help of four hydrostatic dynamical cores that are part of the Community Atmosphere Model version 5 (CAM 5) developed at the National Center for Atmospheric Research (NCAR). In particular, these are the Finite‐Volume, Spectral Element, and spectral transform Eulerian and semi‐Lagrangian dynamical cores that are coupled to the simple‐physics suite. The simulations show that despite the simplicity of the physics forcings the models develop the tropical cyclone at horizontal grid spacings of about 55 km and finer. The simple‐physics simulations reveal essential differences in the storm's structure and strength due to the choice of the dynamical core. Similar differences are also seen in complex full‐physics aqua‐planet experiments with CAM 5 which serve as a motivator for this work. The results suggest that differences in complex full‐physics simulations can be, at least partly, replicated in simplified model setups. The simplified experiments might therefore provide easier access to an improved physical understanding of how the dynamical core and moist physical parameterizations interact. It is concluded that the simple‐physics test case has the potential to close the gap between dry dynamical core assessments and full‐physics aqua‐planet experiments, and can shed light on the role of the dynamical core in the presence of moisture processes.

Two numerical experiments of ensemble hindcasting for 17 years are performed here, initial conditions are NCEP-DOE Reanalysis 2 data (adopted in the real-time prediction) and that in the second are supplied by the system of three-dimensional variational assimilation (3D-VAR), and there are five-member ensembles in each experiment. The model is used here which is a grid-point Atmospheric General Circulation Model with 9 vertical levels and a resolution of 2° in latitudinal and 2.5° in longitudinal direction, developed by the Institute of Atmospheric Physics, the Chinese Academy of Sciences (IAP9L2x2.5-AGCM). The technique of Correlation Analysis is applied. Comparison results show that (1) the difference of temporal anomaly coefficient is very small in the tropics under two different initial conditions, whereas it is large in middle and high latitudes, especially in East Asia (China) region, (2) temporal anomaly coefficient in the experiment 2 is bigger than that in the experiment 1 in most of China region, (3) spatial anomaly correlation of many variables has also different degree improvement by use of the initial conditions after data assimilation(in the experiment 2). That is to say, the results demonstrate quantitatively that the role of initial conditions on short-term climate prediction in East Asia is more important than in the global, what's more, the initial conditions after data assimilation are better than that in the real-time prediction to some extend. So, the effect of initial conditions on atmospheric general circulation model is not neglected.

The purpose of this book is to provide an introduction to numerical modelling of the ocean and the atmosphere. It originates from courses given at Stockholm University and is intended to serve as a textbook for students in meteorology and oceanography with a background in mathematics and physics. Focus is on numerical schemes for the most commonly used equations in oceanography and meteorology as well as on the stability, precision and other properties of these schemes. Simple equations capturing the properties of the primitive equations employed in models of the ocean and atmosphere will be used. These model equations are solved numerically on a grid by discretisation, the derivatives of the differential equations being replaced by finite-difference approximations. The focus will be on the basic numerical methods used for oceanographic and atmospheric modelling. These models are based on the Navier-Stokes equations (including the Coriolis effect) and a tracer equation for heat in both the atmosphere and ocean and tracer equations for humidity and salt in the atmosphere and ocean, respectively. A coupled atmospheric and oceanic general circulation model represents the core part of an Earth System climate model. The book starts by presenting the most common types of partial differential equations and finite difference schemes used in meteorology and oceanography. Subsequently the limitations of these numerical schemes as regards stability, accuracy, presence of computational modes and accuracy the computationally determined phase speed are discussed. The shallow-water equations are discretised for different spatial grids and friction and diffusion terms are introduced. Hereafter implicit and semi-implicit schemes are discussed as well as the semi-Lagrangian technique. Coordinates for atmospheric as well as oceanic models are presented as well as a highly simplified 3D model. A brief description is given of how some atmospheric general circulation models use spectral methods as "horizontal coordinates". Finally, some "pen-and-paper" theoretical exercises and a number of GFD computer exercises are given.

A brief description of a nine-level grid-point global atmospheric general circulation model is presented with the emphasis on the physics parameterizations. This model was developed by the modeling group from Institute of Atmospheric Physics as one task of the CO2-Climate cooperation project between Chinese Academy of Sciences and United States Department of Energy. The task was initiated by Qing-Cun Zeng (IAP) and Robert D. Cess (SUNY). The operational design, computer coding and climate simulation tuning of the model were mainly carried out by Xue-Hong Zhang (for dynamics) and the author (for physics) in SUNY at Stony Brook. The final version was frozen in September 1990. Preliminary diagnoses showed that the model reproduces principal features of the observed climatology.

The aim of this study is to investigate if the representation of Northern Hemisphere blocking is sensitive to resolution in current-generation atmospheric global circulation models (AGCMs). An evaluation is conducted of how well atmospheric blocking is represented in four AGCMs whose horizontal resolution is increased from a grid spacing of more than 100 km to about 25 km. It is shown that Euro-Atlantic blocking is simulated overall more credibly at higher resolution (i.e., in better agreement with a 50-yr reference blocking climatology created from the reanalyses ERA-40 and ERA-Interim). The improvement seen with resolution depends on the season and to some extent on the model considered. Euro-Atlantic blocking is simulated more realistically at higher resolution in winter, spring, and autumn, and robustly so across the model ensemble. The improvement in spring is larger than that in winter and autumn. Summer blocking is found to be better simulated at higher resolution by one model only, with little change seen in the other three models. The representation of Pacific blocking is not found to systematically depend on resolution. Despite the improvements seen with resolution, the 25-km models still exhibit large biases in Euro-Atlantic blocking. For example, three of the four 25-km models underestimate winter northern European blocking frequency by about one-third. The resolution sensitivity and biases in the simulated blocking are shown to be in part associated with the mean-state biases in the models’ midlatitude circulation.

Because precipitation and net radiation in an atmospheric general circulation model (AGCM) are typically biased relative to observations, the simulated evaporative regime of a region may be biased, with consequent negative effects on the AGCM’s ability to translate an initialized soil moisture anomaly into an improved seasonal prediction. These potential problems are investigated through extensive offline analyses with the Mosaic land surface model (LSM). The LSM was first forced globally with a 15-yr observation-based dataset. The simulation was then repeated after imposing a representative set of GCM climate biases onto the forcings—the observational forcings were scaled so that their mean seasonal cycles matched those simulated by the NASA Seasonal-to-Interannual Prediction Project (NSIPP-1; NASA Global Modeling and Assimilation Office) AGCM over the same period. The AGCM’s climate biases do indeed lead to significant biases in evaporative regime in certain regions, with the expected impacts on soil moisture memory time scales. Furthermore, the offline simulations suggest that the biased forcing in the AGCM should contribute to overestimated feedback in certain parts of North America—parts already identified in previous studies as having excessive feedback. The present study thus supports the notion that the reduction of climate biases in the AGCM will lead to more appropriate translations of soil moisture initialization into seasonal prediction skill.

<p>Global or regional atmospheric circulation models often work with horizontal resolutions too large to be able to represent clouds, who have to be parameterized. The way clouds are parameterized and the way they overlap can have a significant impact on their radiative properties. A first objective of this work is to quantify the effect of a vertical description of cloud properties finer than that of current atmospheric models on the calculation of the radiative fluxes at the top of the atmosphere. A second objective is to propose a representation of these subgrid effects that is consistent with the representation of the cloud overlap between layers. For low-level clouds and using LES results as reference, we show the ability of the exponential-random overlap algorithm to represent the vertical distribution of the cloud fraction over a wide range of vertical scales that includes both subgrid scales and overlap between layers, with a constant value of the overlap parameter. Starting from a coarse vertical grid representative of that of atmospheric models, this algorithm is then used to construct the vertical profile of the cloud fraction with a much finer vertical resolution. This reconstruction allows us to test different simplifying hypotheses. We confirm that the frequently used maximum-random overlap leads to a significant error by underestimating the low-level clouds cover with a relative error of about 50%. We suggest some possible representations of subgrid effects and recommend to consider the vertical distribution of the cloud fraction seen from above, which depends on the volume cloud fraction but also on the cloud overlap and the subgrid vertical heterogeneity, when developing or evaluating cloud properties.</p>

A land surface model (LSM) has been included in the ECMWF Hamburg version 4 (ECHAM4) atmospheric general circulation model (AGCM). The LSM is an early version of the Organizing Carbon and Hydrology in Dynamic Ecosystems (ORCHIDEE) and it replaces the simple land surface scheme previously included in ECHAM4. The purpose of this paper is to document how a more exhaustive consideration of the land surface–vegetation processes affects the simulated boreal summer surface climate. To investigate the impacts on the simulated climate, different sets of Atmospheric Model Intercomparison Project (AMIP)-type simulations have been performed with ECHAM4 alone and with the AGCM coupled with ORCHIDEE. Furthermore, to assess the effects of the increase in horizontal resolution the coupling of ECHAM4 with the LSM has been implemented at different horizontal resolutions. The analysis reveals that the LSM has large effects on the simulated boreal summer surface climate of the atmospheric model. Considerable impacts are found in the surface energy balance due to changes in the surface latent heat fluxes over tropical and midlatitude areas covered with vegetation. Rainfall and atmospheric circulation are substantially affected by these changes. In particular, increased precipitation is found over evergreen and summergreen vegetated areas. Because of the socioeconomical relevance, particular attention has been devoted to the Indian summer monsoon (ISM) region. The results of this study indicate that precipitation over the Indian subcontinent is better simulated with the coupled ECHAM4–ORCHIDEE model compared to the atmospheric model alone.

Capabilities of the Grid-point Atmospheric Model of Institute of Atmospheric Physics/National Key Laboratory of Numerical Modeling for Atmospheric Sciences and Geophysical Fluid Dynamics (IAP/LASG; GAMIL) and the Spectral Atmospheric General Circulation Model of IAP/LASG (SAMIL) in simulating the Spring Persistent Rainfall (SPR) in East Asia are evaluated by the Global Precipitation Climatology Project (GPCP). Similar to that observed in the Coupled Model Intercomparison Project Phases 3 and 5 (CMIP3/CMIP5) ensembles, rainfall amount in both GAMIL and SAMIL are slightly overestimated over the SPR domain and are underestimated over the SPR center. For the spatial structure, the northeast–southwest tilted rain belt could not be effectively reproduced by GAMIL, and the main rain belt shifts to the north of its normal position in SAMIL. According to the Taylor diagram, SAMIL generally shows better performances in simulating the SPR pattern. Similar to other atmospheric general circulation models (AGCMs) employed in CMIP3/CMIP5, the thermal contrast between East Asia and the South China Sea is overestimated in SAMIL and GAMIL. The overestimated meridional thermal contrast intensifies the southerly water vapor supply and favors the northward shifting of main rain belt.

The mean evolution of the West African monsoon (WAM) circulation and its interannual variability have been studied using an ensemble of 21 simulations (common period 1961–1994) performed with four different atmospheric general circulation models (AGCMs) (European Center/Hamburg (ECHAM) 3, ECHAM 4, Action de Recherche Petite Echelle Grande Echelle (ARPEGE), and Goddard Institute for Space Studies (GISS)) and forced by the same observed sea surface temperature (SST) data set. The results have been compared with European Centre for Medium‐Range Weather Forecasts reanalyses (ERA‐40). The climatological means of WAM winds for the AGCMs are similar to the ERA‐40 ones. However, the AGCMs tend to underestimate the southern wind component at low levels around 10°N compared to the ERA‐40. The simulated Tropical Easterly Jet (TEJ) is usually shifted northward and also too weak for ECHAM 3 and ECHAM 4 compared to ERA‐40. The interannual variability of an atmospheric WAM index (WAMI) is quite successfully reproduced (the correlations between the mean ensemble of each AGCM and ERA‐40 time series over 1961–1994 range between 0.51 and 0.64). In particular, the four AGCMs reproduce quite well the mean teleconnection structure with El Niño–Southern Oscillation, i.e., a strong (weak) monsoon during La Niña (El Niño) events, even if the largest absolute correlations between WAMI and SST in the eastern and central equatorial Pacific are weaker than in ERA‐40. On a yearly basis, WAMI is more predictable and skillful during the cold ENSO years than during the warm ENSO ones. The unskillful warm ENSO events are associated with a significant cooling over the equatorial Atlantic and Western Pacific Ocean and a significant warming in the tropical Indian Ocean.

The potential use of the National Centers for Environmental Prediction (NCEP) Regional Spectral Model (RSM) for downscaling seasonal tropical cyclone (TC) activity was analyzed here. The NCEP RSM with horizontal resolution of 50 km, was used to downscale the ECHAM4.5 Atmospheric General Circulation Model (AGCM) simulations forced with observed sea surface temperature (SST) over the western North Pacific. An ensemble of ten runs for June–November 1994 and 1998 was studied. The representation of the TCs is much improved compared to the low‐resolution forcing AGCM, but the TCs are not as intense as observed ones, as the RSM horizontal resolution is not sufficiently high. The large‐scale fields of the RSM are examined and compared to both the AGCM and the European Centre for Medium‐Range Weather Forecasts (ECMWF) reanalysis. The large‐scale fields of RSM characteristics are in general similar to those of the reanalysis. Various properties of the TCs in the RSM are also examined such as first positions, tracks, accumulated cyclone energy (ACE) and duration. While the RSM does not reproduce the higher number of TCs in 1994 than in 1998, other measures of TC activity (ACE, number of cyclone days) in the RSM are higher in 1994 than in 1998. Copyright © 2006 Royal Meteorological Society.

The role of atmospheric general circulation model (AGCM) horizontal resolution in representing the global energy budget and hydrological cycle is assessed, with the aim of improving the understanding of model uncertainties in simulating the hydrological cycle. We use two AGCMs from the UK Met Office Hadley Centre: HadGEM1-A at resolutions ranging from 270 to 60 km, and HadGEM3-A ranging from 135 to 25 km. The models exhibit a stable hydrological cycle, although too intense compared to reanalyses and observations. This over-intensity is explained by excess surface shortwave radiation, a common error in general circulation models (GCMs). This result is insensitive to resolution. However, as resolution is increased, precipitation decreases over the ocean and increases over the land. This is associated with an increase in atmospheric moisture transport from ocean to land, which changes the partitioning of moisture fluxes that contribute to precipitation over land from less local to more non-local moisture sources. The results start to converge at 60-km resolution, which underlines the excessive reliance of the mean hydrological cycle on physical parametrization (local unresolved processes) versus model dynamics (large-scale resolved processes) in coarser HadGEM1 and HadGEM3 GCMs. This finding may be valid for other GCMs, showing the necessity to analyze other chains of GCMs that may become available in the future with such a range of horizontal resolutions. Our finding supports the hypothesis that heterogeneity in model parametrization is one of the underlying causes of model disagreement in the Coupled Model Intercomparison Project (CMIP) exercises.

<p>This study investigates the prediction skill of the sub-seasonal prediction model that can depend on the choice of dynamical cores: the finite volume (FV) dynamical core on a latitude-longitude grid system versus spectral element (SE) dynamical core on a cubed-sphere grid system. Recent researches showed that the SE dynamical core on a uniform grid system increases parallel scalability and removes the need for polar filters mitigating uncertainty in climate prediction, particularly for the Arctic region. However, it remains unclear whether the choice of dynamical cores can actually yield significant skill changes or not. To tackle this issue, we implemented a sub-seasonal prediction model based on the Community Atmospheric Model version 5 (CAM5) by incorporating the above two dynamical cores with virtually the same physics schemes. Sub-seasonal prediction skills of the SE dynamical core and FV dynamical core are verified with ERA-interim reanalysis during the early winter (November – December) and the late winter (January – February) from 2001/2002 to 2017/2018. The prediction skills of the two different dynamical cores were significantly different regardless of the virtually same physics schemes. In the ocean, the predictability of the SE dynamical core is similar to the FV dynamical core, mostly because of our simulation configuration imposing the same boundary and initial conditions at the surface. Notable differences in the one-month predictability between the two cores are found for the wintertime Arctic and mid-latitudes, particularly over North America and Eurasia continents. With the one-month lead, SE dynamical core exhibited higher predictability over North America in late winter, whereas the FV dynamical core showed relatively higher predictability in East Asia and Eurasia in early winter. One of the reasons for these differences may be the different manifestations of Arctic-midlatitudes linkage in the two dynamical cores; the SE dynamical core captures warmer Arctic and colder mid-latitudes relatively well than the FV dynamical core. Therefore, we conclude that the careful choice of dynamical cores of sub-seasonal prediction models is needed.</p>

Nitrate as a rapidly increasing aerosol species in recent years affects the present climate and potentially has large implications on the future climate. In this study, the long-term direct radiative forcing (DRF) of nitrate aerosol is investigated using State Key Laboratory of Numerical Modeling for Atmospheric Sciences and Geophysical Fluid Dynamics (LASG) atmospheric general circulation model (AGCM) and the aerosol dataset simulated by a chemical transport model with focus on East Asia. The DRF due to other aerosols, especially sulfate, is also evaluated for comparisons. Although the chemical transport model underestimates the magnitudes of nitrate and sulfate aerosols when compared with Chinese site observations, some insights into the significances of nitrate climate effects still emerge. The present-day global annual mean all-sky DRF of nitrate is calculated to be −0.025 W m−2 relative to the preindustrial era, which is much weaker than −0.37 W m−2 for sulfate. However, nitrate DRF may become increasingly important in the future especially over East Asia, given the expectation that decreasing trend in global sulfate continues while the projected nitrate maintains at the present level for a mid-range forcing scenario and even be a factor of two larger by the end of the 21st century for high emission scenarios. For example, the anthropogenic nitrate DRF of −2.0 W m−2 over eastern China could persist until the 2050s, and nitrate is projected to account for over 60 % of total anthropogenic aerosol DRF over East Asia by 2100. In addition, we illustrate that the regional nitrate DRF and its seasonal variation are sensitive to meteorological parameters, in particular the relative humidity and cloud amount. It thus remains a need for climate models to include more realistically nitrate aerosol in projecting future climate changes.

This work presents snapshot simulations of the late 20th and late 21st century Antarctic climate under the RCP8.5 scenario carried out with an empirically bias‐corrected global atmospheric general circulation model (AGCM), forced with bias‐corrected sea‐surface temperatures and sea ice and run with about 100‐km resolution over Antarctica. The bias correction substantially improves the simulated mean late 20th century climate. The simulated atmospheric circulation of the bias‐corrected model compares very favorably to the best available AMIP (Atmospheric Model Intercomparison Project)‐type climate models. The simulated interannual circulation variability is improved by the bias correction. Depending on the metric, a slight improvement or degradation is found in the simulated variability on synoptic timescales. The simulated climate change over the 21st century is broadly similar in the corrected and uncorrected versions of the atmospheric model, and atmospheric circulation patterns are not geographically “pinned” by the applied bias correction. These results suggest that the method presented here can be used for bias‐corrected climate projections. Finally, the authors discuss different possible choices in terms of the place of bias corrections and other intermediate steps in the modeling chain leading from global coupled climate simulations to impact assessment.