Dynamic Model Atmospheres Research Articles

Simulation of Complex Systems Using the Observed Data Based on Recurrent Artificial Neural Networks

We propose a new approach to reconstructing complex, spatially distributed systems on the basis of the time series generated by such systems. It allows one to combine two basic steps of such a reconstruction, namely, the choice of a set of phase variables of the system using the observed time series and the development of the evolution operator acting in the chosen phase space with the help of an artificial neural network with special topology. This network, first, maps the initial high-dimensional data onto the lower-dimension space and, second, specifies the evolution operator in this space. The efficiency of this approach is demonstrated by an example of reconstructing the Lorenz system representing a high-dimensional model of atmospheric dynamics.

Optimizing the HOMME dynamical core for multicore platforms

The approach of the next-generation computing platforms offers a tremendous opportunity to advance the state-of-the-art in global atmospheric dynamical models. We detail our incremental approach to utilize this emerging technology by enhancing concurrency within the High-Order Method Modeling Environment (HOMME) atmospheric dynamical model developed at the National Center for Atmospheric Research (NCAR). The study focused on improvements to the performance of HOMME which is a Fortran 90 code with a hybrid (MPIOpenMP) programming model. The article describes the changes made to the use of message passing interface (MPI) and OpenMP as well as single-core optimizations to achieve significant improvements in concurrency and overall code performance. For our optimization studies, we utilize the “Cori” system with an Intel Xeon Phi Knights Landing processor deployed at the National Energy Research Supercomputing Center and the “`Cheyenne” system with an Intel Xeon Broadwell processor installed at the NCAR. The results from the studies, using “workhorse” configurations performed at NCAR, show that these changes have a transformative impact on the computational performance of HOMME. Our improvements have shown that we can effectively increase potential concurrency by efficiently threading the vertical dimension. Further, we have seen a factor of two overall improvement in the computational performance of the code resulting from the single-core optimizations. Most notably from the work is that our incremental approach allows for high-impact changes without disrupting existing scientific productivity in the HOMME community.

Radiation exposure of aircrews due to Space Radiation

Space radiation mainly consists of trapped particles inside the Earth’s magnetosphere, galactic and solar cosmic rays. Cosmic rays propagating in the interplanetary medium, reach the top of the Earth’s atmosphere and collide with the molecules of the atmospheric layers creating showers of secondary particles that can be recorded by ground-based neutron monitors or muon detectors. Due to these cascades, the radiation environment in various atmospheric altitudes is entirely different than the one experienced on the Earth’s surface. Space radiation is ionizing (trapped particles, galactic and solar cosmic rays) as well as non-ionizing (ultra-violet radiation). It is known that ionizing radiation is very dangerous for all biological systems, causing a variety of acute and chronic effects. The determination of the occupational exposure of aircrews to space radiation is of great importance. DYnamic Atmospheric Shower Tracking Interactive Model Application (DYASTIMA), as well as its extension DYASTIMA-R, is a standalone application, based on the toolkit Geant4, that simulates the propagation of cosmic radiation into the atmosphere. DYASTIMA provides all the necessary information about the study of the cascade in different atmospheric altitudes, while DYASTIMA-R, as a dosimetry application, calculates radiobiological quantities, such as dose rate and equivalent dose rate for the determination of the exposure, based on the output provided by DYASTIMA. These quantities are calculated for solar minimum and solar maximum conditions. DYASTIMA can be accessed through the portals ofthe European Space Agency (ESA) (http://swe.ssa.esa.int/spaceradiation) and the Athens Neutron Monitor Station (A.Ne.Mo.S) (http://cosray.phys.uoa.gr/index.php/dyastima).

Estimating model evidence using ensemble‐based data assimilation with localization – The model selection problem

In recent years, there has been increased interest in applying data assimilation (DA) methods, originally designed for state estimation, to the model selection problem. In this setting, previous studies introduced the contextual formulation of model evidence, or contextual model evidence (CME), and showed that CME can be efficiently computed using a hierarchy of ensemble‐based DA procedures. Although these studies analysed the DA methods most commonly used for operational atmospheric and oceanic prediction worldwide, they did not study these methods in conjunction with localization to a specific domain. Yet, any application of ensemble DA methods to realistic, very high‐dimensional geophysical models requires the implementation of some form of localization. The present study extends CME estimation to ensemble DA methods with domain localization. Domain‐localized CME (DL‐CME) developed in this article is tested for model selection with two models: (a) the Lorenz 40‐variable midlatitude atmospheric dynamics model (Lorenz‐95); and (b) the simplified global atmospheric SPEEDY model. CME is compared to the root‐mean‐square error (RMSE) as a metric for model selection. The experiments show that CME systematically outperforms RMSE in model selection skill, and that this skill improvement is further enhanced by applying localization to the CME estimate using DL‐CME. The potential use and range of applications of CME and DL‐CME as a model selection metric are also discussed.

Carbon star wind models at solar and sub-solar metallicities: a comparative study

Context. The heavy mass loss observed in evolved stars on the asymptotic giant branch (AGB) is usually attributed to dust-driven winds, but it is still an open question how much AGB stars contribute to the dust production in the interstellar medium, especially at lower metallicities. In the case of C-type AGB stars, where the wind is thought to be driven by radiation pressure on amorphous carbon grains, there should be significant dust production even in metal-poor environments. Carbon stars can manufacture the building blocks needed to form the wind-driving dust species themselves, irrespective of the chemical composition they have, by dredging up carbon from the stellar interior during thermal pulses. Aims. We investigate how the mass loss in carbon stars is affected by a low-metallicity environment, similar to the Large and Small Magellanic Clouds (LMC and SMC). Methods. The atmospheres and winds of C-type AGB stars are modeled with the 1D spherically symmetric radiation-hydrodynamical code Dynamic Atmosphere and Radiation-driven Wind models based on Implicit Numerics (DARWIN). The models include a time-dependent description for nucleation, growth, and evaporation of amorphous carbon grains directly out of the gas phase. To explore the metallicity-dependence of mass loss we calculate model grids at three different chemical abundances (solar, LMC, and SMC). Since carbon may be dredged up during the thermal pulses as AGB stars evolve, we keep the carbon abundance as a free parameter. The models in these three different grids all have a current mass of one solar mass; effective temperatures of 2600, 2800, 3000, or 3200 K; and stellar luminosities equal to logL*∕L⊙ = 3.70, 3.85, or 4.00. Results. The DARWIN models show that mass loss in carbon stars is facilitated by high luminosities, low effective temperatures, and a high carbon excess (C–O) at both solar and subsolar metallicities. Similar combinations of effective temperature, luminosity, and carbon excess produce outflows at both solar and subsolar metallicities. There are no large systematic differences in the mass-loss rates and wind velocities produced by these wind models with respect to metallicity, nor any systematic difference concerning the distribution of grain sizes or how much carbon is condensed into dust. DARWIN models at subsolar metallicity have approximately 15% lower mass-loss rates compared to DARWIN models at solar metallicity with the same stellar parameters and carbon excess. For both solar and subsolar environments typical grain sizes range between 0.1 and 0.5 μm, the degree of condensed carbon varies between 5 and 40%, and the gas-to-dust ratios between 500 and 10 000. Conclusions. C-type AGB stars can contribute to the dust production at subsolar metallicities (down to at least [Fe∕H] = −1) as long as they dredge up sufficient amounts of carbon from the stellar interior. Furthermore, stellar evolution models can use the mass-loss rates calculated from DARWIN models at solar metallicity when modeling the AGB phase at subsolar metallicities if carbon excess is used as the critical abundance parameter instead of the C/O ratio.

Exploring the sensitivity of Northern Hemisphere atmospheric circulation to different surface temperature forcing using a statistical–dynamical atmospheric model

Abstract. Climate and weather conditions in the mid-latitudes are strongly driven by the large-scale atmosphere circulation. Observational data indicate that important components of the large-scale circulation have changed in recent decades, including the strength and the width of the Hadley cell, jets, storm tracks and planetary waves. Here, we use a new statistical–dynamical atmosphere model (SDAM) to test the individual sensitivities of the large-scale atmospheric circulation to changes in the zonal temperature gradient, meridional temperature gradient and global-mean temperature. We analyze the Northern Hemisphere Hadley circulation, jet streams, storm tracks and planetary waves by systematically altering the zonal temperature asymmetry, the meridional temperature gradient and the global-mean temperature. Our results show that the strength of the Hadley cell, storm tracks and jet streams depend, in terms of relative changes, almost linearly on both the global-mean temperature and the meridional temperature gradient, whereas the zonal temperature asymmetry has little or no influence. The magnitude of planetary waves is affected by all three temperature components, as expected from theoretical dynamical considerations. The width of the Hadley cell behaves nonlinearly with respect to all three temperature components in the SDAM. Moreover, some of these observed large-scale atmospheric changes are expected from dynamical equations and are therefore an important part of model validation.

ЛЕТНИЙ СЦЕНАРИЙ МОДЕЛИРОВАНИЯ РАССЕИВАНИЯ ПРИМЕСЕЙ ОТ ИСТОЧНИКОВ ВЫБРОСОВ В ЦЕНТРАЛЬНОЙ ЭКОЛОГИЧЕСКОЙ ЗОНЕ БАЙКАЛЬСКОЙ ПРИРОДНОЙ ТЕРРИТОРИИ

The wind in the lower layers of the atmosphere is a direct distributor of pollutants from small and medium energy facilities located near the waters of Lake Baikal. To assess the effect of various meteorological situations on the processes of impurity transport to the water area, a mesoscale model of atmospheric dynamics and impurity transport is used, which is being developed in ICM & MG SB RAS. The results of numerical modeling of local circulations in the Baikal region in the summer period are presented. The spatial and temporal variability of wind flow and the dynamics of the distribution of impurities in the model scenario are shown.

Global trends in carbon sinks and their relationships with CO2 and temperature

Elevated CO2 concentrations increase photosynthesis and, potentially, net ecosystem production (NEP), meaning a greater CO2 uptake. Climate, nutrients and ecosystem structure, however, influence the effect of increasing CO2. Here we analysed global NEP from MACC-II and Jena CarboScope atmospheric inversions and ten dynamic global vegetation models (TRENDY), using statistical models to attribute the trends in NEP to its potential drivers: CO2, climatic variables and land-use change. We found that an increased CO2 was consistently associated with an increased NEP (1995–2014). Conversely, increased temperatures were negatively associated with NEP. Using the two atmospheric inversions and TRENDY, the estimated global sensitivities for CO2 were 6.0 ± 0.1, 8.1 ± 0.3 and 3.1 ± 0.1 PgC per 100 ppm (~1 °C increase), and −0.5 ± 0.2, −0.9 ± 0.4 and −1.1 ± 0.1 PgC °C−1 for temperature. These results indicate a positive CO2 effect on terrestrial C sinks that is constrained by climate warming. Global net ecosystem production (NEP) from a number of atmospheric inversions and dynamic global vegetation models is analysed to attribute trends to potential drivers. CO2 is found to have a positive effect on NEP that is constrained by climate warming.

Understanding Anomalous Synoptic Eddy Vorticity Forcing in Pacific–North American Teleconnection Pattern Events

Abstract Utilizing a decomposition of anomalous eddy vorticity forcing (EVF) proposed by Song in 2016 and a modified Geophysical Fluid Dynamics Laboratory (GFDL) dynamical core atmospheric model, this study provides a different understanding of physical mechanisms that are responsible for the formation of the anomalous synoptic EVF (SEVF) associated with Pacific–North American teleconnection pattern (PNA) events. A series of short-term control experiments (CEs) and initial-value modified experiments (IVMEs) is conducted. In each case of CEs, there are no obvious PNA-like circulation anomalies. IVMEs are exactly the same as CEs except that appropriate small perturbations are introduced into the initial-value fields of CEs. The modified initial-value fields led to a gradual development of the PNA-like flow anomalies in IVMEs. Based on these numerical results, deformations of the synoptic eddy due to the emergence of the PNA pattern can be easily acquired by subtracting the synoptic eddy in CEs from the synoptic eddy in IVMEs . The anomalous SEVF associated with the PNA events in the model can be decomposed into ensembles of two linear and interaction terms (EVF1 and EVF2) and a nonlinear self-interaction term (EVF3). It is demonstrated that the physical essence of the anomalous SEVF associated with the PNA events is a competition result between EVF1 plus EVF2 and EVF3. Results also indicate that the different signs of SEVF associated with the positive and negative PNA events are not necessarily related to the different tilts of the synoptic eddy.

Front speed simulation with a hyperbolic model of atmospheric dynamics

In this paper the front speed is simulated with a non-hydrostatic finite-element model of atmospheric dynamics. The model is based on the compressible Navier-Stokes equations in two dimensions. Artificial compressibility is introduced into the model in order to make the governing equations hyperbolic. Two tests of the model on different spatial scales are considered: the propagation of a mesoscale atmospheric gravity current (cold front) over flat terrain and the motion of a small scale coastal flow over irregular terrain. In the first test the calculated values of the simulation are compared with an empirical formula first introduced by T. von Karman and later developed for atmospheric fronts by many authors. In the second test, the model is applied to simulating the dynamic pressure effects caused by changes in roughness and interaction of the flow with the non-homogeneous terrain. For this purpose, we consider flow over a low hill at a coastal site. Both model simulation results are compared with available observations and simulations performed by other authors. In general, good agreement between the results of the test calculations and the theory has been obtained.

AGB star atmospheres modeling as feedback to stellar evolutionary and galaxy models

AbstractThe chemical enrichment of the Universe is considerably affected by the contribution of cool evolved stars. We studied the O-rich star R Peg and the C-rich star V Oph, using respectively the VLTI/GRAVITY and VLTI/MIDI instruments. We interpret the data using grids of 1-D and 3-D dynamic model atmospheres.

Lumpy stars and bumpy winds

AbstractThe wind-driving process of AGB stars is thought to be a two-step process: first matter is levitated by shock waves, and then accelerated outwards by radiation pressure on newly condensed dust grains. When modelling such a wind, spherical symmetry is usually assumed. This is in stark contrast with recent observations, which shows significant non-spherical structures. Giant convection cells cover the surface of the star, and matter is being ejected into the atmosphere where it condenses into lumpy dust clouds. We try to quantify the differences between what is simulated in the 3D star-in-a-box models (CO5BOLD code) and the 1D dynamical atmosphere and wind models (DARWIN code). The impact of having a non-spherical star on the wind properties is also investigated. We find that the inherent non-spherical behaviour of AGB stars might induce a dust-driven weak wind already early on the AGB, and including that the star is anisotropic when simulating the wind leads to large time variations in the density of the outflow. Such variations might be observable as small-scale structures in the circumstellar envelope.

Oxygen-rich Long Period Variables in the X-Shooter Spectral Library

AbstractThe X-Shooter Spectral Library (XSL) contains more than 800 spectra of stars across the color-magnitude diagram, that extend from near-UV to near-IR wavelengths (320-2450 nm). We summarize properties of the spectra of O-rich Long Period Variables in the XSL, such as phase-related features, and we confront the data with synthetic spectra based on static and dynamical stellar atmosphere models. We discuss successes and remaining discrepancies, keeping in mind the applications to population synthesis modeling that XSL is designed for.

Simulation of Daily Rainfall from Concurrent Meteorological Parameters over Core Monsoon Region of India: A Novel Approach

In the present work, daily rainfall is simulated over the core monsoon region of India by using a feedforward multilayer perceptron (MLP) model. Daily rainfall is found to be optimally dependent on four concurrent meteorological parameters, namely, geopotential height, specific humidity, zonal, and meridional wind at 1000 mb, 925 mb, 850 mb, and 700 mb pressure levels during 00, 06, 12, and 18 Greenwich Mean Time (GMT). The architecture of the optimized feedforward MLP model consists of 64 nodes in the input layer, 10 nodes in the hidden layer, and 1 node in the output layer. The results from the model are compared with the 3B42 (version 7) rainfall product. In terms of root mean square error (rmse) and correlation coefficient (cc), the model is performing better compared to the satellite-derived 3B42 rainfall product, whereas in terms of bias, the performance of the 3B42 product is better compared to the model. The weight matrices of the feedforward MLP model are estimated at a particular location (22.5°N, 82.5°E). These weight matrices are able to simulate daily rainfall at neighbourhood locations also with reasonably good accuracy with cc in the range of 0.41 to 0.55. The performance of the model improves in case of an aerial average of daily rainfall with significantly enhanced cc (0.72). The model is able to capture monthly and intraseasonal variation of rainfall with reasonably good accuracy, with cc of 0.88 and 0.68, respectively. The simulation model has a limitation that it is not able to simulate extreme high rainfall events (>60 mm/day). Overall, the developed model is performing reasonably well. This approach has a potential to be used as a rain parameterization scheme in the dynamical atmospheric and coupled models to simulate daily rainfall. Nevertheless, the present approach can also be used for multistep prediction of rainfall.

Toward Consistent Modeling of Atmospheric Chemistry and Dynamics in Exoplanets: Validation and Generalization of the Chemical Relaxation Method

Abstract Motivated by the work of Cooper & Showman, we revisit the chemical relaxation method, which seeks to enhance the computational efficiency of chemical kinetics calculations by replacing the chemical network with a handful of independent source/sink terms. Chemical relaxation solves the evolution of the system and can treat disequilibrium chemistry, as the source/sink terms are driven toward chemical equilibrium on a prescribed chemical timescale, but it has surprisingly never been validated. First, we generalize the treatment by forgoing the use of a single chemical timescale, instead developing a pathway analysis tool that allows us to identify the rate-limiting reaction as a function of temperature and pressure. For the interconversion between methane and carbon monoxide, and between ammonia and molecular nitrogen, we identify the key rate-limiting reactions for conditions relevant to currently characterizable exo-atmospheres (500–3000 K, 0.1 mbar to 1 kbar). Second, we extend chemical relaxation to include carbon dioxide and water. Third, we examine the role of metallicity and the carbon-to-oxygen ratio in chemical relaxation. Fourth, we apply our pathway analysis tool to diagnose the differences between our chemical network and that of Moses and Venot. Finally, we validate the chemical relaxation method against full chemical kinetics calculations in one dimension. For WASP-18b-, HD 189733b-, and GJ 1214-b-like atmospheres, we show that chemical relaxation is mostly accurate to within an order of magnitude, a factor of 2, and ∼10%, respectively. The level of accuracy attained allows for the chemical relaxation method to be included in three-dimensional general circulation models.

Spatiotemporal distribution and dynamic modeling of atmospheric gaseous polycyclic aromatic hydrocarbons in a rapidly urbanizing city: Nanjing, China.

Multiple studies have evaluated the concentration and lung cancer risk of polycyclic aromatic hydrocarbons (PAHs). However, the monitoring and dynamic modeling of PAHs with a high resolution were relatively insufficient. We investigated the spatiotemporal distribution of gaseous PAH concentrations using passive air samplers with high sampling density in an industrial city of Nanjing, China (January and October 2015) and found that the gaseous PAH concentrations in western Nanjing were higher than those in eastern Nanjing, mainly because of emission source distribution and wind action. There were notable seasonal changes in PAH concentrations: winter > autumn > spring > summer. We developed an atmospheric PAH dynamic model with a high resolution of 1km2 based on the advection-diffusion equation and coupled with an emissions inventory and atmospheric transportation processes. Acenaphthene was selected as a proxy for gaseous PAHs. The modeled acenaphthene concentrations were similar to the concentrations measured. Moreover, we used the model to identify the impact of meteorological factors on gaseous PAHs via scenario analysis and found that a narrow-range temperature change and even heavy rainfall may not significantly affect atmospheric gaseous PAH concentrations, whereas the wind played an important part in transferring PAHs and changing their geographic distribution.

Wind Shielding Impacts on Water Quality in an Urban Reservoir

This study applied an atmospheric Computational Fluid Dynamics (CFD) model and a three-dimensional hydrodynamic-ecological lake model to investigate the impacts of variable wind forcing on lake hydrodynamics and water quality. The CFD model was used to predict the wind field over the lake surface which served as the meteorological forcing for the lake model. The study was carried out for the Marina Reservoir in downtown Singapore where field measurements of the wind field showed that temporal and spatial differences in wind field were found to be significant. By comparing simulations using uniform and spatially variable wind inputs in terms of thermal stratification, Richardson number, vorticity, and spatial gradients, this study demonstrates that the variability in wind forcing in urban lakes that arise from shielding and seasonal effects can have significant impacts on the hydrodynamics and water quality in the reservoir.

VLTI-GRAVITY measurements of cool evolved stars

Context. Dynamic model atmospheres of Mira stars predict variabilities in the photospheric radius and in atmospheric molecular layers which are not yet strongly constrained by observations. Aims. Here we measure the variability of the oxygen-rich Mira star R Peg in near-continuum and molecular bands. Methods. We used near-infrared K-band spectro-interferometry with a spectral resolution of about 4000 obtained at four epochs between post-maximum and minimum visual phases employing the newly available GRAVITY beam combiner at the Very Large Telescope Interferometer (VLTI). Results. Our observations show a continuum radius that is anti-correlated with the visual lightcurve. Uniform disc (UD) angular diameters at a near-continuum wavelength of 2.25 μm are steadily increasing with values of 8.7 ± 0.1 mas, 9.4 ± 0.1 mas, 9.8 ± 0.1 mas, and 9.9 ± 0.1 mas at visual phases of 0.15, 0.36, 0,45, 0.53, respectively. UD diameters at a bandpass around 2.05 μm, dominated by water vapour, follow the near-continuum variability at larger UD diameters between 10.7 mas and 11.7 mas. UD diameters at the CO 2–0 bandhead, instead, are correlated with the visual lightcurve and anti-correlated with the near-continuum UD diameters, with values between 12.3 mas and 11.7 mas. Conclusions. The observed anti-correlation between continuum radius and visual lightcurve is consistent with an earlier study of the oxygen-rich Mira S Lac, and with recent 1D CODEX dynamic model atmosphere predictions. The amplitude of the variation is comparable to the earlier observations of S Lac, and smaller than predicted by CODEX models. The wavelength-dependent visibility variations at our epochs can be reproduced by a set of CODEX models at model phases between 0.3 and 0.6. The anti-correlation of water vapour and CO contributions at our epochs suggests that these molecules undergo different processes in the extended atmosphere along the stellar cycle. The newly available GRAVITY instrument is suited to conducting longer time series observations, which are needed to provide strong constraints on the model-predicted intra- and inter-cycle variability.

The lagged connection of the positive NAO with the MJO phase 3 in a simplified atmospheric model

Based on a simplified nonlinear model and reanalysis data, the lagged connection of the North Atlantic Oscillation (NAO) with the Madden–Julian Oscillation (MJO) in boreal winters is investigated. The positive NAO is observed to occur more frequently about 8–20 days after the onset of the MJO phase 3. A series of heating forcing experiments and initial-value experiments are conducted by utilizing the Geophysical Fluid Dynamics Laboratory (GFDL) dynamical core atmospheric model. The extratropical responses to the tropical heating associated with the MJO phase 3 are characterized by a wave train over the Pacific–North American region with an anticyclone anomaly over the northeastern Pacific and then followed by a positive-NAO-like pattern over the North Atlantic sector. These circulation anomalies generally match the observed lagged-connection well. At the earlier stage, the Rossby wave train excited by the MJO convection propagates into the North Atlantic, leading to a planetary wave anomaly with a low-over-high dipole prior to the positive NAO. At the later stage, the anomalous synoptic eddy vorticity forcing (EVF) streamfunction tendency has a negative-over-positive dipole, which plays a key role in the development of the positive NAO. Further analysis of the initial-value experiments indicates that, for the subsequent formation of the positive NAO, the anomalous circulation over the Indian Ocean aroused by the MJO phase 3 is more crucial than that over the northeastern Pacific.

The dynamical core of the Aeolus 1.0 statistical–dynamical atmosphere model: validation and parameter optimization

Abstract. We present and validate a set of equations for representing the atmosphere's large-scale general circulation in an Earth system model of intermediate complexity (EMIC). These dynamical equations have been implemented in Aeolus 1.0, which is a statistical–dynamical atmosphere model (SDAM) and includes radiative transfer and cloud modules (Coumou et al., 2011; Eliseev et al., 2013). The statistical dynamical approach is computationally efficient and thus enables us to perform climate simulations at multimillennia timescales, which is a prime aim of our model development. Further, this computational efficiency enables us to scan large and high-dimensional parameter space to tune the model parameters, e.g., for sensitivity studies. Here, we present novel equations for the large-scale zonal-mean wind as well as those for planetary waves. Together with synoptic parameterization (as presented by Coumou et al., 2011), these form the mathematical description of the dynamical core of Aeolus 1.0. We optimize the dynamical core parameter values by tuning all relevant dynamical fields to ERA-Interim reanalysis data (1983–2009) forcing the dynamical core with prescribed surface temperature, surface humidity and cumulus cloud fraction. We test the model's performance in reproducing the seasonal cycle and the influence of the El Niño–Southern Oscillation (ENSO). We use a simulated annealing optimization algorithm, which approximates the global minimum of a high-dimensional function. With non-tuned parameter values, the model performs reasonably in terms of its representation of zonal-mean circulation, planetary waves and storm tracks. The simulated annealing optimization improves in particular the model's representation of the Northern Hemisphere jet stream and storm tracks as well as the Hadley circulation. The regions of high azonal wind velocities (planetary waves) are accurately captured for all validation experiments. The zonal-mean zonal wind and the integrated lower troposphere mass flux show good results in particular in the Northern Hemisphere. In the Southern Hemisphere, the model tends to produce too-weak zonal-mean zonal winds and a too-narrow Hadley circulation. We discuss possible reasons for these model biases as well as planned future model improvements and applications.