Articles | Volume 13, issue 9
Geosci. Model Dev., 13, 4399–4412, 2020
https://doi.org/10.5194/gmd-13-4399-2020
© Author(s) 2020. This work is distributed under
the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Geosci. Model Dev., 13, 4399–4412, 2020
https://doi.org/10.5194/gmd-13-4399-2020
© Author(s) 2020. This work is distributed under
the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Model description paper 21 Sep 2020
Model description paper | 21 Sep 2020
RadNet 1.0: exploring deep learning architectures for longwave radiative transfer
Ying Liu et al.
Related authors
No articles found.
Sonja Murto, Rodrigo Caballero, Gunilla Svensson, and Lukas Papritz
Weather Clim. Dynam., 3, 21–44, https://doi.org/10.5194/wcd-3-21-2022, https://doi.org/10.5194/wcd-3-21-2022, 2022
Short summary
Short summary
This study uses reanalysis data to investigate the role of atmospheric blocking, prevailing high-pressure systems and mid-latitude cyclones in driving high-Arctic wintertime warm extreme events. These events are mainly preceded by Ural and Scandinavian blocks, which are shown to be significantly influenced and amplified by cyclones in the North Atlantic. It also highlights processes that need to be well captured in climate models for improving their representation of Arctic wintertime climate.
Joy Merwin Monteiro, Jeremy McGibbon, and Rodrigo Caballero
Geosci. Model Dev., 11, 3781–3794, https://doi.org/10.5194/gmd-11-3781-2018, https://doi.org/10.5194/gmd-11-3781-2018, 2018
Short summary
Short summary
In the same way that the fruit fly or the yeast cell serve as model systems in biology, climate scientists use a range of computer models to gain a fundamental understanding of our climate system. These models range from extremely simple models that can run on your phone to those that require supercomputers. Sympl and climt are packages that make it easy for climate scientists to build a hierarchy of such models using Python, which facilitates easy to read and self-documenting models.
David K. Hutchinson, Agatha M. de Boer, Helen K. Coxall, Rodrigo Caballero, Johan Nilsson, and Michiel Baatsen
Clim. Past, 14, 789–810, https://doi.org/10.5194/cp-14-789-2018, https://doi.org/10.5194/cp-14-789-2018, 2018
Short summary
Short summary
The Eocene--Oligocene transition was a major cooling event 34 million years ago. Climate model studies of this transition have used low ocean resolution or topography that roughly approximates the time period. We present a new climate model simulation of the late Eocene, with higher ocean resolution and topography which is accurately designed for this time period. These features improve the ocean circulation and gateways which are thought to be important for this climate transition.
Henrik Carlson and Rodrigo Caballero
Clim. Past, 13, 1037–1048, https://doi.org/10.5194/cp-13-1037-2017, https://doi.org/10.5194/cp-13-1037-2017, 2017
Short summary
Short summary
Climate models are able to simulate the surface temperature of the early Eocene as reconstructed from paleoclimatology data, but only by using extremely high CO2 concentrations or clouds that are more transparent to solar radiation. We explore the potential for distinguishing among these two forcing agents via their impact on regional climate. Better constraining the radiative forcing that led to Eocene warmth has important implications for understanding Earth's climate sensitivity.
Daniel J. Lunt, Matthew Huber, Eleni Anagnostou, Michiel L. J. Baatsen, Rodrigo Caballero, Rob DeConto, Henk A. Dijkstra, Yannick Donnadieu, David Evans, Ran Feng, Gavin L. Foster, Ed Gasson, Anna S. von der Heydt, Chris J. Hollis, Gordon N. Inglis, Stephen M. Jones, Jeff Kiehl, Sandy Kirtland Turner, Robert L. Korty, Reinhardt Kozdon, Srinath Krishnan, Jean-Baptiste Ladant, Petra Langebroek, Caroline H. Lear, Allegra N. LeGrande, Kate Littler, Paul Markwick, Bette Otto-Bliesner, Paul Pearson, Christopher J. Poulsen, Ulrich Salzmann, Christine Shields, Kathryn Snell, Michael Stärz, James Super, Clay Tabor, Jessica E. Tierney, Gregory J. L. Tourte, Aradhna Tripati, Garland R. Upchurch, Bridget S. Wade, Scott L. Wing, Arne M. E. Winguth, Nicky M. Wright, James C. Zachos, and Richard E. Zeebe
Geosci. Model Dev., 10, 889–901, https://doi.org/10.5194/gmd-10-889-2017, https://doi.org/10.5194/gmd-10-889-2017, 2017
Short summary
Short summary
In this paper we describe the experimental design for a set of simulations which will be carried out by a range of climate models, all investigating the climate of the Eocene, about 50 million years ago. The intercomparison of model results is called 'DeepMIP', and we anticipate that we will contribute to the next IPCC report through an analysis of these simulations and the geological data to which we will compare them.
M. Löfverström, R. Caballero, J. Nilsson, and J. Kleman
Clim. Past, 10, 1453–1471, https://doi.org/10.5194/cp-10-1453-2014, https://doi.org/10.5194/cp-10-1453-2014, 2014
A. Goldner, M. Huber, and R. Caballero
Clim. Past, 9, 173–189, https://doi.org/10.5194/cp-9-173-2013, https://doi.org/10.5194/cp-9-173-2013, 2013
Related subject area
Climate and Earth system modeling
ChAP 1.0: a stationary tropospheric sulfur cycle for Earth system models of intermediate complexity
Non-Hydrostatic RegCM4 (RegCM4-NH): model description and case studies over multiple domains
Robustness of neural network emulations of radiative transfer parameterizations in a state-of-the-art general circulation model
NorCPM1 and its contribution to CMIP6 DCPP
ENSO-ASC 1.0.0: ENSO deep learning forecast model with a multivariate air–sea coupler
Topography-based local spherical Voronoi grid refinement on classical and moist shallow-water finite-volume models
Decadal climate predictions with the Canadian Earth System Model version 5 (CanESM5)
The Simplified Chemistry-Dynamical Model (SCDM V1.0)
Iodine chemistry in the chemistry–climate model SOCOL-AERv2-I
Coupling interactive fire with atmospheric composition and climate in the UK Earth System Model
Fast and accurate learned multiresolution dynamical downscaling for precipitation
A parameterization of sub-grid topographical effects on solar radiation in the E3SM Land Model (version 1.0): implementation and evaluation over the Tibetan Plateau
WETMETH 1.0: a new wetland methane model for implementation in Earth system models
Effect of horizontal resolution on the simulation of tropical cyclones in the Chinese Academy of Sciences FGOALS-f3 climate system model
Grid-stretching capability for the GEOS-Chem 13.0.0 atmospheric chemistry model
PARASO, a circum-Antarctic fully-coupled ice-sheet - ocean - sea-ice - atmosphere - land model involving f.ETISh1.7, NEMO3.6, LIM3.6, COSMO5.0 and CLM4.5
Performance of the Adriatic Sea and Coast (AdriSC) climate component – a COAWST V3.3-based one-way coupled atmosphere–ocean modelling suite: ocean results
Validation of terrestrial biogeochemistry in CMIP6 Earth system models: a review
FAMOUS version xotzt (FAMOUS-ice): a general circulation model (GCM) capable of energy- and water-conserving coupling to an ice sheet model
EC-Earth3-AerChem: a global climate model with interactive aerosols and atmospheric chemistry participating in CMIP6
Enhancing the Accessibility of Unified Modeling Systems: GFDL SHiELD v2021b in a Container
Coupling the Community Land Model version 5.0 to the parallel data assimilation framework PDAF: Description and applications
Vertical grid refinement for stratocumulus clouds in the radiation scheme of the global climate model ECHAM6.3-HAM2.3-P3
Cloud Feedbacks from CanESM2 to CanESM5.0 and their influence on climate sensitivity
ATTRICI v1.1 – counterfactual climate for impact attribution
Mitigation of the double ITCZ syndrome in BCC-CSM2-MR through improving parameterizations of boundary-layer turbulence and shallow convection
COSMO-CLM regional climate simulations in the Coordinated Regional Climate Downscaling Experiment (CORDEX) framework: a review
TempestExtremes v2.1: a community framework for feature detection, tracking, and analysis in large datasets
The Earth Model Column Collaboratory (EMC2) v1.1: An Open-Source Ground-Based Lidar and Radar Instrument Simulator and Subcolumn Generator for Large-Scale Models
ICONGETM v1.0 – flexible NUOPC-driven two-way coupling via ESMF exchange grids between the unstructured-grid atmosphere model ICON and the structured-grid coastal ocean model GETM
A permafrost implementation in the simple carbon–climate model Hector v.2.3pf
The SMHI Large Ensemble (SMHI-LENS) with EC-Earth3.3.1
Oil palm modelling in the global land surface model ORCHIDEE-MICT
Testing the reliability of interpretable neural networks in geoscience using the Madden–Julian oscillation
Climate-model-informed deep learning of global soil moisture distribution
fv3gfs-wrapper: a Python wrapper of the FV3GFS atmospheric model
C-LLAMA v1.0: traceable model for food, agriculture and land-use
Recalibrating decadal climate predictions – what is an adequate model for the drift?
CARDAMOM-FluxVal Version 1.0: a FLUXNET-based Validation System for CARDAMOM Carbon and Water Flux Estimates
Multi-variate factorisation of numerical simulations
Inclusion of a suite of weathering tracers in the cGENIE Earth system model – muffin release v.0.9.23
The ENEA-REG system (v1.0), a multi-component regional Earth system model: sensitivity to different atmospheric components over the Med-CORDEX (Coordinated Regional Climate Downscaling Experiment) region
CM2Mc-LPJmL v1.0: biophysical coupling of a process-based dynamic vegetation model with managed land to a general circulation model
ESM-Tools version 5.0: a modular infrastructure for stand-alone and coupled Earth system modelling (ESM)
Impact of increased resolution on long-standing biases in HighResMIP-PRIMAVERA climate models
Performance of the Adriatic Sea and Coast (AdriSC) climate component – a COAWST V3.3-based coupled atmosphere–ocean modelling suite: atmospheric dataset
Evaluation and optimisation of the I/O scalability for the next generation of Earth system models: IFS CY43R3 and XIOS 2.0 integration as a case study
Step-wise modifications of the Vegetation Optimality Model
Model of Early Diagenesis in the Upper Sediment with Adaptable complexity – MEDUSA (v. 2): a time-dependent biogeochemical sediment module for Earth system models, process analysis and teaching
A Markov chain method for weighting climate model ensembles
Alexey V. Eliseev, Rustam D. Gizatullin, and Alexandr V. Timazhev
Geosci. Model Dev., 14, 7725–7747, https://doi.org/10.5194/gmd-14-7725-2021, https://doi.org/10.5194/gmd-14-7725-2021, 2021
Short summary
Short summary
A stationary, computationally efficient scheme, ChAP 1.0 (Chemical and Aerosol Processes, version 1.0), is developed for the sulfur cycle in the troposphere. This scheme is designed for Earth system models of intermediate complexity (EMICs). The scheme model reasonably reproduces characteristics of the tropospheric sulfur cycle. Despite its simplicity, ChAP may be successfully used to simulate anthropogenic sulfur pollution in the atmosphere at coarse spatial scales and timescales.
Erika Coppola, Paolo Stocchi, Emanuela Pichelli, Jose Abraham Torres Alavez, Russell Glazer, Graziano Giuliani, Fabio Di Sante, Rita Nogherotto, and Filippo Giorgi
Geosci. Model Dev., 14, 7705–7723, https://doi.org/10.5194/gmd-14-7705-2021, https://doi.org/10.5194/gmd-14-7705-2021, 2021
Short summary
Short summary
In this work we describe the development of a non-hydrostatic version of the regional climate model RegCM4-NH, implemented to allow simulations at convection-permitting scales of <4 km for climate applications. The new core is described, and three case studies of intense convection are carried out to illustrate the model performances. Comparison with observations is much improved with respect to with coarse grid runs. RegCM4-NH offers a promising tool for climate investigations at a local scale.
Alexei Belochitski and Vladimir Krasnopolsky
Geosci. Model Dev., 14, 7425–7437, https://doi.org/10.5194/gmd-14-7425-2021, https://doi.org/10.5194/gmd-14-7425-2021, 2021
Short summary
Short summary
There is a lot interest in using machine learning (ML) techniques to improve environmental models by replacing physically based model components with ML-derived ones. The latter ordinarily demonstrate excellent results when tested in a stand-alone setting but can break their host model either outright when coupled to it or eventually when the model changes. We built an ML component that not only does not destabilize its host model but is also robust with respect to substantial changes in it.
Ingo Bethke, Yiguo Wang, François Counillon, Noel Keenlyside, Madlen Kimmritz, Filippa Fransner, Annette Samuelsen, Helene Langehaug, Lea Svendsen, Ping-Gin Chiu, Leilane Passos, Mats Bentsen, Chuncheng Guo, Alok Gupta, Jerry Tjiputra, Alf Kirkevåg, Dirk Olivié, Øyvind Seland, Julie Solsvik Vågane, Yuanchao Fan, and Tor Eldevik
Geosci. Model Dev., 14, 7073–7116, https://doi.org/10.5194/gmd-14-7073-2021, https://doi.org/10.5194/gmd-14-7073-2021, 2021
Short summary
Short summary
The Norwegian Climate Prediction Model version 1 (NorCPM1) is a new research tool for performing climate reanalyses and seasonal-to-decadal climate predictions. It adds data assimilation capability to the Norwegian Earth System Model version 1 (NorESM1) and has contributed output to the Decadal Climate Prediction Project (DCPP) as part of the sixth Coupled Model Intercomparison Project (CMIP6). We describe the system and evaluate its baseline, reanalysis and prediction performance.
Bin Mu, Bo Qin, and Shijin Yuan
Geosci. Model Dev., 14, 6977–6999, https://doi.org/10.5194/gmd-14-6977-2021, https://doi.org/10.5194/gmd-14-6977-2021, 2021
Short summary
Short summary
Considering the sophisticated energy exchanges and multivariate coupling in ENSO, we subjectively incorporate the prior physical knowledge into the modeling process and build up an ENSO deep learning forecast model with a multivariate air–sea coupler, named ENSO-ASC, the performance of which outperforms the other state-of-the-art models. The extensive experiments indicate that ENSO-ASC is a powerful tool for both the ENSO prediction and for the analysis of the underlying complex mechanisms.
Luan F. Santos and Pedro S. Peixoto
Geosci. Model Dev., 14, 6919–6944, https://doi.org/10.5194/gmd-14-6919-2021, https://doi.org/10.5194/gmd-14-6919-2021, 2021
Short summary
Short summary
The Andes act as a wall in atmospheric flows and play an important role in the weather of South America but are currently underrepresented in weather and climate models. In this work, we propose grids that better capture the mountains and, using idealized dynamical models, study the effects caused by the use of such grids. While possibly improving forecasts for short periods, the grids introduce spurious numerical (nonphysical) effects, which can demand added caution from model developers.
Reinel Sospedra-Alfonso, William J. Merryfield, George J. Boer, Viatsheslav V. Kharin, Woo-Sung Lee, Christian Seiler, and James R. Christian
Geosci. Model Dev., 14, 6863–6891, https://doi.org/10.5194/gmd-14-6863-2021, https://doi.org/10.5194/gmd-14-6863-2021, 2021
Short summary
Short summary
CanESM5 decadal predictions that started from observed climate states represent the observed evolution of upper-ocean temperatures, surface climate, and the carbon cycle better than ones not started from observed climate states for several years into the forecast. This is due both to better representations of climate internal variability and to corrections of the model response to external forcing including changes in GHG emissions and aerosols.
Hao-Jhe Hong and Thomas Reichler
Geosci. Model Dev., 14, 6647–6660, https://doi.org/10.5194/gmd-14-6647-2021, https://doi.org/10.5194/gmd-14-6647-2021, 2021
Short summary
Short summary
The Arctic wintertime circulation of the stratosphere has pronounced impacts on the troposphere and surface climate. Changes in the stratospheric circulation can lead to either increases or decreases in Arctic ozone. Understanding the interactions between ozone and the circulation will have the benefit of model prediction for the climate. This study introduces an economical and fast simplified model that represents the realistic distribution of ozone and its interaction with the circulation.
Arseniy Karagodin-Doyennel, Eugene Rozanov, Timofei Sukhodolov, Tatiana Egorova, Alfonso Saiz-Lopez, Carlos A. Cuevas, Rafael P. Fernandez, Tomás Sherwen, Rainer Volkamer, Theodore K. Koenig, Tanguy Giroud, and Thomas Peter
Geosci. Model Dev., 14, 6623–6645, https://doi.org/10.5194/gmd-14-6623-2021, https://doi.org/10.5194/gmd-14-6623-2021, 2021
Short summary
Short summary
Here, we present the iodine chemistry module in the SOCOL-AERv2 model. The obtained iodine distribution demonstrated a good agreement when validated against other simulations and available observations. We also estimated the iodine influence on ozone in the case of present-day iodine emissions, the sensitivity of ozone to doubled iodine emissions, and when considering only organic or inorganic iodine sources. The new model can be used as a tool for further studies of iodine effects on ozone.
João C. Teixeira, Gerd A. Folberth, Fiona M. O'Connor, Nadine Unger, and Apostolos Voulgarakis
Geosci. Model Dev., 14, 6515–6539, https://doi.org/10.5194/gmd-14-6515-2021, https://doi.org/10.5194/gmd-14-6515-2021, 2021
Short summary
Short summary
Fire constitutes a key process in the Earth system, being driven by climate as well as affecting climate. However, studies on the effects of fires on atmospheric composition and climate have been limited to date. This work implements and assesses the coupling of an interactive fire model with atmospheric composition, comparing it to an offline approach. This approach shows good performance at a global scale. However, regional-scale limitations lead to a bias in modelling fire emissions.
Jiali Wang, Zhengchun Liu, Ian Foster, Won Chang, Rajkumar Kettimuthu, and V. Rao Kotamarthi
Geosci. Model Dev., 14, 6355–6372, https://doi.org/10.5194/gmd-14-6355-2021, https://doi.org/10.5194/gmd-14-6355-2021, 2021
Short summary
Short summary
Downscaling, the process of generating a higher spatial or time dataset from a coarser observational or model dataset, is a widely used technique. Two common methodologies for performing downscaling are to use either dynamic (physics-based) or statistical (empirical). Here we develop a novel methodology, using a conditional generative adversarial network (CGAN), to perform the downscaling of a model's precipitation forecasts and describe the advantages of this method compared to the others.
Dalei Hao, Gautam Bisht, Yu Gu, Wei-Liang Lee, Kuo-Nan Liou, and L. Ruby Leung
Geosci. Model Dev., 14, 6273–6289, https://doi.org/10.5194/gmd-14-6273-2021, https://doi.org/10.5194/gmd-14-6273-2021, 2021
Short summary
Short summary
Topography exerts significant influence on the incoming solar radiation at the land surface. This study incorporated a well-validated sub-grid topographic parameterization in E3SM land model (ELM) version 1.0. The results demonstrate that sub-grid topography has non-negligible effects on surface energy budget, snow cover, and surface temperature over the Tibetan Plateau and that the ELM simulations are sensitive to season, elevation, and spatial scale.
Claude-Michel Nzotungicimpaye, Kirsten Zickfeld, Andrew H. MacDougall, Joe R. Melton, Claire C. Treat, Michael Eby, and Lance F. W. Lesack
Geosci. Model Dev., 14, 6215–6240, https://doi.org/10.5194/gmd-14-6215-2021, https://doi.org/10.5194/gmd-14-6215-2021, 2021
Short summary
Short summary
In this paper, we describe a new wetland methane model (WETMETH) developed for use in Earth system models. WETMETH consists of simple formulations to represent methane production and oxidation in wetlands. We also present an evaluation of the model performance as embedded in the University of Victoria Earth System Climate Model (UVic ESCM). WETMETH is capable of reproducing mean annual methane emissions consistent with present-day estimates from the regional to the global scale.
Jinxiao Li, Qing Bao, Yimin Liu, Lei Wang, Jing Yang, Guoxiong Wu, Xiaofei Wu, Bian He, Xiaocong Wang, Xiaoqi Zhang, Yaoxian Yang, and Zili Shen
Geosci. Model Dev., 14, 6113–6133, https://doi.org/10.5194/gmd-14-6113-2021, https://doi.org/10.5194/gmd-14-6113-2021, 2021
Short summary
Short summary
The configuration and simulated performance of tropical cyclones (TCs) in FGOALS-f3-L/H will be introduced firstly. The results indicate that the simulated performance of TC activities is improved globally with the increased horizontal resolution especially in TC counts, seasonal cycle, interannual variabilities and intensity aspects. It is worth establishing a high-resolution coupled dynamic prediction system based on FGOALS-f3-H (~ 25 km) to improve the prediction skill of TCs.
Liam Bindle, Randall V. Martin, Matthew J. Cooper, Elizabeth W. Lundgren, Sebastian D. Eastham, Benjamin M. Auer, Thomas L. Clune, Hongjian Weng, Jintai Lin, Lee T. Murray, Jun Meng, Christoph A. Keller, William M. Putman, Steven Pawson, and Daniel J. Jacob
Geosci. Model Dev., 14, 5977–5997, https://doi.org/10.5194/gmd-14-5977-2021, https://doi.org/10.5194/gmd-14-5977-2021, 2021
Short summary
Short summary
Atmospheric chemistry models like GEOS-Chem are versatile tools widely used in air pollution and climate studies. The simulations used in such studies can be very computationally demanding, and thus it is useful if the model can simulate a specific geographic region at a higher resolution than the rest of the globe. Here, we implement, test, and demonstrate a new variable-resolution capability in GEOS-Chem that is suitable for simulations conducted on supercomputers.
Charles Pelletier, Thierry Fichefet, Hugues Goosse, Konstanze Haubner, Samuel Helsen, Pierre-Vincent Huot, Christoph Kittel, François Klein, Sebastien Le clec'h, Nicole P. M. van Lipzig, Sylvain Marchi, François Massonnet, Pierre Mathiot, Ehsan Moravveji, Eduardo Moreno-Chamarro, Pablo Ortega, Frank Pattyn, Niels Souverijns, Guillian Van Achter, Sam Vanden Broucke, Alexander Vanhulle, Deborah Verfaillie, and Lars Zipf
Geosci. Model Dev. Discuss., https://doi.org/10.5194/gmd-2021-315, https://doi.org/10.5194/gmd-2021-315, 2021
Revised manuscript accepted for GMD
Short summary
Short summary
We present PARASO, a circumpolar model for simulating the Antarctic climate. PARASO features five distinct models, each covering different Earth system subcomponents (ice sheet, atmosphere, land, sea ice, ocean). In this technical article, we describe how this tool has been developed, with a focus on the "coupling interfaces" representing the feedbacks between the distinct models put to contribution. PARASO is stable and ready-to-use, but is still characterized by significant biases.
Petra Pranić, Cléa Denamiel, and Ivica Vilibić
Geosci. Model Dev., 14, 5927–5955, https://doi.org/10.5194/gmd-14-5927-2021, https://doi.org/10.5194/gmd-14-5927-2021, 2021
Short summary
Short summary
The Adriatic Sea and Coast model was developed due to the need for higher-resolution climate models and longer-term simulations to capture coastal atmospheric and ocean processes at climate scales in the Adriatic Sea. The ocean results of a 31-year-long simulation were compared to the observational data. The evaluation revealed that the model is capable of reproducing the observed physical properties with good accuracy and can be further used to study the dynamics of the Adriatic–Ionian basin.
Lynsay Spafford and Andrew H. MacDougall
Geosci. Model Dev., 14, 5863–5889, https://doi.org/10.5194/gmd-14-5863-2021, https://doi.org/10.5194/gmd-14-5863-2021, 2021
Short summary
Short summary
Land biogeochemical cycles influence global climate change. Their influence is examined through complex computer models that account for the interaction of the land, ocean, and atmosphere. Improved models used in the recent round of model intercomparison used inconsistent validation methods to compare simulated land biogeochemistry to datasets. For the next round of model intercomparisons we recommend a validation protocol with explicit reference datasets and informative performance metrics.
Robin S. Smith, Steve George, and Jonathan M. Gregory
Geosci. Model Dev., 14, 5769–5787, https://doi.org/10.5194/gmd-14-5769-2021, https://doi.org/10.5194/gmd-14-5769-2021, 2021
Short summary
Short summary
Many of the complex computer models used to study the physics of the natural world treat ice sheets as fixed and unchanging, capable of only simple interactions with the rest of the climate. This is partly because it is technically very difficult to usefully do anything more realistic. We have adapted a climate model so it can be joined together with a dynamical model of the Greenland ice sheet. This gives us a powerful tool to help us better understand how ice sheets and the climate interact.
Twan van Noije, Tommi Bergman, Philippe Le Sager, Declan O'Donnell, Risto Makkonen, María Gonçalves-Ageitos, Ralf Döscher, Uwe Fladrich, Jost von Hardenberg, Jukka-Pekka Keskinen, Hannele Korhonen, Anton Laakso, Stelios Myriokefalitakis, Pirkka Ollinaho, Carlos Pérez García-Pando, Thomas Reerink, Roland Schrödner, Klaus Wyser, and Shuting Yang
Geosci. Model Dev., 14, 5637–5668, https://doi.org/10.5194/gmd-14-5637-2021, https://doi.org/10.5194/gmd-14-5637-2021, 2021
Short summary
Short summary
This paper documents the global climate model EC-Earth3-AerChem, one of the members of the EC-Earth3 family of models participating in CMIP6. We give an overview of the model and describe in detail how it differs from its predecessor and the other EC-Earth3 configurations. The model's performance is characterized using coupled simulations conducted for CMIP6. The model has an effective equilibrium climate sensitivity of 3.9 °C and a transient climate response of 2.1 °C.
Kai-Yuan Cheng, Lucas M. Harris, and Yong Qiang Sun
Geosci. Model Dev. Discuss., https://doi.org/10.5194/gmd-2021-239, https://doi.org/10.5194/gmd-2021-239, 2021
Revised manuscript accepted for GMD
Short summary
Short summary
This paper presents the implementation of container technology for the System for High‐resolution prediction on Earth‐to‐Local Domains (SHiELD), a unified atmospheric model that can be used as a global, a global-nest, and a regional model for weather-to-seasonal prediction. Container technology makes SHiELD cross-platform and easy to use, which opens opportunities for collaborative research and development. Performance and scalability of the containerized SHiELD are evaluated and discussed.
Lukas Strebel, Heye Bogena, Harry Vereecken, and Harrie-Jan Hendricks Franssen
Geosci. Model Dev. Discuss., https://doi.org/10.5194/gmd-2021-38, https://doi.org/10.5194/gmd-2021-38, 2021
Revised manuscript accepted for GMD
Short summary
Short summary
We present the technical coupling between a land surface model (CLM5) and the Parallel Data Assimilation Framework (PDAF). This coupling enables measurement data to update simulated model states and parameters in a statistically optimal way. We demonstrate the viability of the model framework using an application in a forested catchment where the inclusion of soil water measurements significantly improved the simulation quality.
Paolo Pelucchi, David Neubauer, and Ulrike Lohmann
Geosci. Model Dev., 14, 5413–5434, https://doi.org/10.5194/gmd-14-5413-2021, https://doi.org/10.5194/gmd-14-5413-2021, 2021
Short summary
Short summary
Stratocumulus are thin clouds whose cloud cover is underestimated in climate models partly due to overly low vertical resolution. We develop a scheme that locally refines the vertical grid based on a physical constraint for the cloud top. Global simulations show that the scheme, implemented only in the radiation routine, can increase stratocumulus cloud cover. However, this effect is poorly propagated to the simulated cloud cover. The scheme's limitations and possible ways forward are discussed.
John G. Virgin, Christopher G. Fletcher, Jason N. S. Cole, Knut von Salzen, and Toni Mitovski
Geosci. Model Dev., 14, 5355–5372, https://doi.org/10.5194/gmd-14-5355-2021, https://doi.org/10.5194/gmd-14-5355-2021, 2021
Short summary
Short summary
Equilibrium climate sensitivity, or the amount of warming the Earth would exhibit a result of a doubling of atmospheric CO2, is a common metric used in assessments of climate models. Here, we compare climate sensitivity between two versions of the Canadian Earth System Model. We find the newest iteration of the model (version 5) to have higher climate sensitivity due to reductions in low-level clouds, which reflect radiation and cool the planet, as the surface warms.
Matthias Mengel, Simon Treu, Stefan Lange, and Katja Frieler
Geosci. Model Dev., 14, 5269–5284, https://doi.org/10.5194/gmd-14-5269-2021, https://doi.org/10.5194/gmd-14-5269-2021, 2021
Short summary
Short summary
To identify the impacts of historical climate change it is necessary to separate the effect of the different impact drivers. To address this, one needs to compare historical impacts to a counterfactual world with impacts that would have been without climate change. We here present an approach that produces counterfactual climate data and can be used in climate impact models to simulate counterfactual impacts. We make these data available through the ISIMIP project.
Yixiong Lu, Tongwen Wu, Yubin Li, and Ben Yang
Geosci. Model Dev., 14, 5183–5204, https://doi.org/10.5194/gmd-14-5183-2021, https://doi.org/10.5194/gmd-14-5183-2021, 2021
Short summary
Short summary
The spurious precipitation in the tropical southeastern Pacific and southern Atlantic is one of the most prominent systematic biases in coupled atmosphere–ocean general circulation models. This study significantly promotes the marine stratus simulation and largely alleviates the excessive precipitation biases through improving parameterizations of boundary-layer turbulence and shallow convection, providing an effective solution to the long-standing bias in the tropical precipitation simulation.
Silje Lund Sørland, Roman Brogli, Praveen Kumar Pothapakula, Emmanuele Russo, Jonas Van de Walle, Bodo Ahrens, Ivonne Anders, Edoardo Bucchignani, Edouard L. Davin, Marie-Estelle Demory, Alessandro Dosio, Hendrik Feldmann, Barbara Früh, Beate Geyer, Klaus Keuler, Donghyun Lee, Delei Li, Nicole P. M. van Lipzig, Seung-Ki Min, Hans-Jürgen Panitz, Burkhardt Rockel, Christoph Schär, Christian Steger, and Wim Thiery
Geosci. Model Dev., 14, 5125–5154, https://doi.org/10.5194/gmd-14-5125-2021, https://doi.org/10.5194/gmd-14-5125-2021, 2021
Short summary
Short summary
We review the contribution from the CLM-Community to regional climate projections following the CORDEX framework over Europe, South Asia, East Asia, Australasia, and Africa. How the model configuration, horizontal and vertical resolutions, and choice of driving data influence the model results for the five domains is assessed, with the purpose of aiding the planning and design of regional climate simulations in the future.
Paul A. Ullrich, Colin M. Zarzycki, Elizabeth E. McClenny, Marielle C. Pinheiro, Alyssa M. Stansfield, and Kevin A. Reed
Geosci. Model Dev., 14, 5023–5048, https://doi.org/10.5194/gmd-14-5023-2021, https://doi.org/10.5194/gmd-14-5023-2021, 2021
Short summary
Short summary
TempestExtremes (TE) is a multifaceted framework for feature detection, tracking, and scientific analysis of regional or global Earth system datasets. Version 2.1 of TE now provides extensive support for nodal and areal features. This paper describes the algorithms that have been added to the TE framework since version 1.0 and gives several examples of how these can be combined to produce composite algorithms for evaluating and understanding atmospheric features.
Israel Silber, Robert C. Jackson, Ann M. Fridlind, Andrew S. Ackerman, Scott Collis, Johannes Verlinde, and Jiachen Ding
Geosci. Model Dev. Discuss., https://doi.org/10.5194/gmd-2021-194, https://doi.org/10.5194/gmd-2021-194, 2021
Revised manuscript accepted for GMD
Short summary
Short summary
The Earth Model Column Collaboratory (EMC2) is an open-source ground-based (and air- or space-borne) lidar and radar simulator and subcolumn generator designed for large-scale models, in particular climate models, applicable also for high-resolution models. EMC2 emulates measurements while remaining faithful to large-scale models' physical assumptions implemented in their cloud or radiation schemes. We demonstrate the use of EMC2 to compare AWARE measurements with the NASA GISS ModelE3 and LES.
Tobias Peter Bauer, Peter Holtermann, Bernd Heinold, Hagen Radtke, Oswald Knoth, and Knut Klingbeil
Geosci. Model Dev., 14, 4843–4863, https://doi.org/10.5194/gmd-14-4843-2021, https://doi.org/10.5194/gmd-14-4843-2021, 2021
Short summary
Short summary
We present the coupled atmosphere–ocean model system ICONGETM. The added value and potential of using the latest coupling technologies are discussed in detail. An exchange grid handles the different coastlines from the unstructured atmosphere and the structured ocean grids. Due to a high level of automated processing, ICONGETM requires only minimal user input. The application to a coastal upwelling scenario demonstrates significantly improved model results compared to uncoupled simulations.
Dawn L. Woodard, Alexey N. Shiklomanov, Ben Kravitz, Corinne Hartin, and Ben Bond-Lamberty
Geosci. Model Dev., 14, 4751–4767, https://doi.org/10.5194/gmd-14-4751-2021, https://doi.org/10.5194/gmd-14-4751-2021, 2021
Short summary
Short summary
We have added a representation of the permafrost carbon feedback to the simple, open-source global carbon–climate model Hector and calibrated the results to be consistent with historical data and Earth system model projections. Our results closely match previous work, estimating around 0.2 °C of warming from permafrost this century. This capability will be useful to explore uncertainties in this feedback and for coupling with integrated assessment models for policy and economic analysis.
Klaus Wyser, Torben Koenigk, Uwe Fladrich, Ramon Fuentes-Franco, Mehdi Pasha Karami, and Tim Kruschke
Geosci. Model Dev., 14, 4781–4796, https://doi.org/10.5194/gmd-14-4781-2021, https://doi.org/10.5194/gmd-14-4781-2021, 2021
Short summary
Short summary
This paper describes the large ensemble done by SMHI with the EC-Earth3 climate model. The ensemble comprises 50 realizations for each of the historical experiments after 1970 and four different future projections for CMIP6. We describe the creation of the initial states for the ensemble and the reduced set of output variables. A first look at the results illustrates the changes in the climate during this century and puts them in relation to the uncertainty from the model's internal variability.
Yidi Xu, Philippe Ciais, Le Yu, Wei Li, Xiuzhi Chen, Haicheng Zhang, Chao Yue, Kasturi Kanniah, Arthur P. Cracknell, and Peng Gong
Geosci. Model Dev., 14, 4573–4592, https://doi.org/10.5194/gmd-14-4573-2021, https://doi.org/10.5194/gmd-14-4573-2021, 2021
Short summary
Short summary
In this study, we implemented the specific morphology, phenology and harvest process of oil palm in the global land surface model ORCHIDEE-MICT. The improved model generally reproduces the same leaf area index, biomass density and life cycle fruit yield as observations. This explicit representation of oil palm in a global land surface model offers a useful tool for understanding the ecological processes of oil palm growth and assessing the environmental impacts of oil palm plantations.
Benjamin A. Toms, Karthik Kashinath, Prabhat, and Da Yang
Geosci. Model Dev., 14, 4495–4508, https://doi.org/10.5194/gmd-14-4495-2021, https://doi.org/10.5194/gmd-14-4495-2021, 2021
Short summary
Short summary
We test whether a type of machine learning called neural networks can be used trustfully within the geosciences. We do so by challenging the networks to understand the spatial patterns of a commonly studied geoscientific phenomenon. The neural networks can correctly identify the spatial patterns, which lends confidence that similar networks can be used for more uncertain problems. The results of this study may give geoscientists confidence when using neural networks in their research.
Klaus Klingmüller and Jos Lelieveld
Geosci. Model Dev., 14, 4429–4441, https://doi.org/10.5194/gmd-14-4429-2021, https://doi.org/10.5194/gmd-14-4429-2021, 2021
Short summary
Short summary
Soil moisture is of great importance for weather and climate. We present a machine learning model that produces accurate predictions of satellite-observed surface soil moisture, based on meteorological data from a climate model. It can be used as soil moisture parametrisation in climate models and to produce comprehensive global soil moisture datasets. Moreover, it may motivate similar applications of machine learning in climate science.
Jeremy McGibbon, Noah D. Brenowitz, Mark Cheeseman, Spencer K. Clark, Johann P. S. Dahm, Eddie C. Davis, Oliver D. Elbert, Rhea C. George, Lucas M. Harris, Brian Henn, Anna Kwa, W. Andre Perkins, Oliver Watt-Meyer, Tobias F. Wicky, Christopher S. Bretherton, and Oliver Fuhrer
Geosci. Model Dev., 14, 4401–4409, https://doi.org/10.5194/gmd-14-4401-2021, https://doi.org/10.5194/gmd-14-4401-2021, 2021
Short summary
Short summary
FV3GFS is a weather and climate model written in Fortran. It uses Fortran so that it can run fast, but this makes it hard to add features if you do not (or even if you do) know Fortran. We have written a Python interface to FV3GFS that lets you import the Fortran model as a Python package. We show examples of how this is used to write
modelscripts, which reproduce or build on what the Fortran model can do. You could do this same wrapping for any compiled model, not just FV3GFS.
Thomas S. Ball, Naomi E. Vaughan, Thomas W. Powell, Andrew Lovett, and Timothy M. Lenton
Geosci. Model Dev. Discuss., https://doi.org/10.5194/gmd-2021-169, https://doi.org/10.5194/gmd-2021-169, 2021
Revised manuscript accepted for GMD
Short summary
Short summary
We present C-LLAMA 1.0 (Country-Level Land Availability Model for Agriculture), a simple model of the global food and agriculture system. C-LLAMA uses publicly available dataset to make projections of diet, food demand, food production and land-use. Operating at the national level, each country acts as a 'box' with inputs and outputs. The model is designed to be simple to understand and modify, making it well-placed to explore the food system as a driver of land-use in future scenarios.
Alexander Pasternack, Jens Grieger, Henning W. Rust, and Uwe Ulbrich
Geosci. Model Dev., 14, 4335–4355, https://doi.org/10.5194/gmd-14-4335-2021, https://doi.org/10.5194/gmd-14-4335-2021, 2021
Short summary
Short summary
Decadal climate ensemble forecasts are increasingly being used to guide adaptation measures. To ensure the applicability of these probabilistic predictions, inherent systematic errors of the prediction system must be adjusted. Since it is not clear which statistical model is optimal for this purpose, we propose a recalibration strategy with a systematic model selection based on non-homogeneous boosting for identifying the most relevant features for both ensemble mean and ensemble spread.
Yan Yang, A. Anthony Bloom, Shuang Ma, Paul Levine, Alexander Norton, Nicholas C. Parazoo, John T. Reager, John Worden, Gregory R. Quetin, T. Luke Smallman, Mathew Williams, Liang Xu, and Sassan Saatchi
Geosci. Model Dev. Discuss., https://doi.org/10.5194/gmd-2021-190, https://doi.org/10.5194/gmd-2021-190, 2021
Revised manuscript accepted for GMD
Short summary
Short summary
Global carbon and water have large uncertainties that are hard to quantify in current regional and global models. Field observations provide opportunities for better calibration and validation of current modeling of carbon and water. With the unique structure of CARDAMOM model, we have utilized the data assimilation capability and designed the benchmarking framework by taking field observations into modeling. Results show that data assimilation improves model performance in different aspects.
Daniel J. Lunt, Deepak Chandan, Alan M. Haywood, George M. Lunt, Jonathan C. Rougier, Ulrich Salzmann, Gavin A. Schmidt, and Paul J. Valdes
Geosci. Model Dev., 14, 4307–4317, https://doi.org/10.5194/gmd-14-4307-2021, https://doi.org/10.5194/gmd-14-4307-2021, 2021
Short summary
Short summary
Often in science we carry out experiments with computers in which several factors are explored, for example, in the field of climate science, how the factors of greenhouse gases, ice, and vegetation affect temperature. We can explore the relative importance of these factors by
swapping in and outdifferent values of these factors, and can also carry out experiments with many different combinations of these factors. This paper discusses how best to analyse the results from such experiments.
Markus Adloff, Andy Ridgwell, Fanny M. Monteiro, Ian J. Parkinson, Alexander J. Dickson, Philip A. E. Pogge von Strandmann, Matthew S. Fantle, and Sarah E. Greene
Geosci. Model Dev., 14, 4187–4223, https://doi.org/10.5194/gmd-14-4187-2021, https://doi.org/10.5194/gmd-14-4187-2021, 2021
Short summary
Short summary
We present the first representation of the trace metals Sr, Os, Li and Ca in a 3D Earth system model (cGENIE). The simulation of marine metal sources (weathering, hydrothermal input) and sinks (deposition) reproduces the observed concentrations and isotopic homogeneity of these metals in the modern ocean. With these new tracers, cGENIE can be used to test hypotheses linking these metal cycles and the cycling of other elements like O and C and simulate their dynamic response to external forcing.
Alessandro Anav, Adriana Carillo, Massimiliano Palma, Maria Vittoria Struglia, Ufuk Utku Turuncoglu, and Gianmaria Sannino
Geosci. Model Dev., 14, 4159–4185, https://doi.org/10.5194/gmd-14-4159-2021, https://doi.org/10.5194/gmd-14-4159-2021, 2021
Short summary
Short summary
The Mediterranean Basin is a complex region, characterized by the presence of pronounced topography and a complex land–sea distribution including a considerable number of islands and straits; these features generate strong local atmosphere–sea interactions.
Regional Earth system models have been developed and used to study both present and future Mediterranean climate systems. The main aims of this paper are to present and evaluate the newly developed regional Earth system model ENEA-REG.
Markus Drüke, Werner von Bloh, Stefan Petri, Boris Sakschewski, Sibyll Schaphoff, Matthias Forkel, Willem Huiskamp, Georg Feulner, and Kirsten Thonicke
Geosci. Model Dev., 14, 4117–4141, https://doi.org/10.5194/gmd-14-4117-2021, https://doi.org/10.5194/gmd-14-4117-2021, 2021
Short summary
Short summary
In this study, we couple the well-established and comprehensively validated state-of-the-art dynamic LPJmL5 global vegetation model to the CM2Mc coupled climate model (CM2Mc-LPJmL v.1.0). Several improvements to LPJmL5 were implemented to allow a fully functional biophysical coupling. The new climate model is able to capture important biospheric processes, including fire, mortality, permafrost, hydrological cycling and the the impacts of managed land (crop growth and irrigation).
Dirk Barbi, Nadine Wieters, Paul Gierz, Miguel Andrés-Martínez, Deniz Ural, Fatemeh Chegini, Sara Khosravi, and Luisa Cristini
Geosci. Model Dev., 14, 4051–4067, https://doi.org/10.5194/gmd-14-4051-2021, https://doi.org/10.5194/gmd-14-4051-2021, 2021
Eduardo Moreno-Chamarro, Louis-Philippe Caron, Saskia Loosveldt Tomas, Oliver Gutjahr, Marie-Pierre Moine, Dian Putrasahan, Christopher D. Roberts, Malcolm J. Roberts, Retish Senan, Laurent Terray, Etienne Tourigny, and Pier Luigi Vidale
Geosci. Model Dev. Discuss., https://doi.org/10.5194/gmd-2021-209, https://doi.org/10.5194/gmd-2021-209, 2021
Revised manuscript accepted for GMD
Short summary
Short summary
Climate models do not fully reproduce observations: they show differences (biases) in regional temperature, precipitation, or cloud cover. Reducing model biases is important to increase our confidence in the climate models' ability to reproduce the future climate change. A model's realism is set by its resolution: the finer it is, the more physical processes and interactions it can resolve. Our paper shows that increasing resolution up to ~25 km can help reduce model biases but not totally.
Cléa Denamiel, Petra Pranić, Damir Ivanković, Iva Tojčić, and Ivica Vilibić
Geosci. Model Dev., 14, 3995–4017, https://doi.org/10.5194/gmd-14-3995-2021, https://doi.org/10.5194/gmd-14-3995-2021, 2021
Short summary
Short summary
The atmospheric results of the Adriatic Sea and Coast (AdriSC) climate simulation (1987–2017) are evaluated against available observational datasets in the Adriatic region. Generally, the AdriSC model performs better than regional climate models that have resolutions that are 4 times more coarse, except concerning summer temperatures, which are systematically underestimated. High-resolution climate models may thus provide new insights about the local impacts of global warming in the Adriatic.
Xavier Yepes-Arbós, Gijs van den Oord, Mario C. Acosta, and Glenn D. Carver
Geosci. Model Dev. Discuss., https://doi.org/10.5194/gmd-2021-65, https://doi.org/10.5194/gmd-2021-65, 2021
Revised manuscript accepted for GMD
Short summary
Short summary
Climate prediction models produce a large volume of simulated data that sometimes might not be efficiently managed. In this paper we present an approach to address this issue by reducing the computing time and storage space. As a case study, we analyse the output writing process of the ECMWF atmospheric model called IFS, and we integrate into it a data writing tool called XIOS. The results suggest that the integration between both components achieves an adequate computational performance.
Remko Christiaan Nijzink, Jason Beringer, Lindsay Beaumont Hutley, and Stanislaus Josef Schymanski
Geosci. Model Dev. Discuss., https://doi.org/10.5194/gmd-2021-151, https://doi.org/10.5194/gmd-2021-151, 2021
Revised manuscript accepted for GMD
Short summary
Short summary
The Vegetation Optimality Model (VOM) is a coupled water-vegetation model that predicts vegetation properties, rather than determining them based on observations. A range of updates to previous applications of the VOM have been made for increased generality and improved comparability with conventional models. This showed that there is a large effect in the simulated water and carbon fluxes caused by the assumption of deep groundwater tables and updated soil profiles in the model.
Guy Munhoven
Geosci. Model Dev., 14, 3603–3631, https://doi.org/10.5194/gmd-14-3603-2021, https://doi.org/10.5194/gmd-14-3603-2021, 2021
Short summary
Short summary
Sea-floor sediments play an important role in biogeochemical cycling of elements (e.g. carbon, silicon, nutrients) in the ocean. Realistic sediment modules are, however, not yet commonly used in global ocean biogeochemical models. Here we present MEDUSA, a model of the processes taking place in the surface sea-floor sediments which control the interaction between the sediments and the ocean. MEDUSA can be configured to meet the exact needs of any given ocean biogeochemical model.
Max Kulinich, Yanan Fan, Spiridon Penev, Jason P. Evans, and Roman Olson
Geosci. Model Dev., 14, 3539–3551, https://doi.org/10.5194/gmd-14-3539-2021, https://doi.org/10.5194/gmd-14-3539-2021, 2021
Short summary
Short summary
We present a novel stochastic approach based on Markov chains to estimate climate model weights of multi-model ensemble means. This approach showed improved performance (better correlation with observations) over existing alternatives during cross-validation and model-as-truth tests. The results of this comparative analysis should serve to motivate further studies in applications of Markov chain and other nonlinear methods to find optimal model weights for constructing ensemble means.
Cited articles
Brenowitz, N. D. and Bretherton, C. S.: Prognostic Validation of a Neural
Network Unified Physics Parameterization, Geophys. Res. Lett., 45, 6289–6298, https://doi.org/10.1029/2018GL078510, 2018. a
Brenowitz, N. D. and Bretherton, C. S.: Spatially Extended Tests of a
Neural Network Parametrization Trained by Coarse-Graining,
J. Adv. Model. Earth Sy., 11, 2728–2744,
https://doi.org/10.1029/2019MS001711,
2019. a
Chevallier, F., Chéruy, F., Scott, N. A., and Chédin, A.: A Neural
Network Approach for a Fast and Accurate Computation of a
Longwave Radiative Budget, J. Appl. Meteorol., 37,
1385–1397, https://doi.org/10.1175/1520-0450(1998)037<1385:ANNAFA>2.0.CO;2, 1998. a
Chollet, F.: Xception: Deep Learning with Depthwise Separable Convolutions,
CoRR, abs/1610.02357, 2016. a
Dee, D. P., Uppala, S. M., Simmons, A. J., Berrisford, P., Poli, P.,
Kobayashi, S., Andrae, U., Balmaseda, M. A., Balsamo, G., Bauer, P.,
Bechtold, P., Beljaars, A. C. M., van de Berg, I., Biblot, J., Bormann, N.,
Delsol, C., Dragani, R., Fuentes, M., Greer, A. J., Haimberger, L., Healy, S.
B., Hersbach, H., Holm, E. V., Isaksen, L., Kallberg, P., Kohler, M.,
Matricardi, M., McNally, A. P., Mong-Sanz, B. M., Morcette, J.-J., Park,
B.-K., Peubey, C., de Rosnay, P., Tavolato, C., Thepaut, J. N., and Vitart,
F.: The ERA-Interim reanalysis: Configuration and performance of the data
assimilation system, Q. J. Roy. Meteorol. Soc., 137, 553–597, https://doi.org/10.1002/qj.828, 2011. a, b
Dennis, J. M. and Loft, R. D.: Refactoring Scientific Applications for Massive
Parallelism, in: Numerical Techniques for Global Atmospheric Models, Lecture
Notes in Computational Science and Engineering, Springer,
Berlin, Heidelberg, 539–556, https://doi.org/10.1007/978-3-642-11640-7_16, 2011. a
Dueben, P. D. and Bauer, P.: Challenges and design choices for global weather and climate models based on machine learning, Geosci. Model Dev., 11, 3999–4009, https://doi.org/10.5194/gmd-11-3999-2018, 2018. a
Emanuel, K. A. and Zivkovic-Rothman, M.: Development and Evaluation of a
Convection Scheme for Use in Climate Models, J.
Atmos. Sci., 56, 1766–1782,
https://doi.org/10.1175/1520-0469(1999)056<1766:DAEOAC>2.0.CO;2, 00553, 1999. a
Fu, Q. and Liou, K. N.: On the Correlated k-Distribution Method for Radiative
Transfer in Nonhomogeneous Atmospheres, J. Atmos. Sci., 49, 2139–2156,
https://doi.org/10.1175/1520-0469(1992)049<2139:OTCDMF>2.0.CO;2, 1992. a
Gardner, M. and Dorling, S.: Artificial Neural Networks (The Multilayer
Perceptron) – A Review of Applications in the Atmospheric Sciences,
Atmos. Environ., 32, 2627–2636, 1998. a
Gentine, P., Pritchard, M., Rasp, S., Reinaudi, G., and Yacalis, G.: Could
Machine Learning Break the Convection Parameterization Deadlock?,
Geophys. Res. Lett., 45, 5742–5751, https://doi.org/10.1029/2018GL078202,
2018. a
Goodfellow, I., Bengio, Y., and Courville, A.: Deep learning, MIT press, 2016. a
He, K., Zhang, X., Ren, S., and Sun, J.: Delving Deep into Rectifiers:
Surpassing Human-Level Performance on ImageNet Classification, CoRR,
abs/1502.01852,
2015a. a
He, K., Zhang, X., Ren, S., and Sun, J.: Deep Residual Learning for Image
Recognition, CoRR, abs/1512.03385, 2015b. a
Howard, A. G., Zhu, M., Chen, B., Kalenichenko, D., Wang, W., Weyand, T.,
Andreetto, M., and Adam, H.: MobileNets: Efficient Convolutional Neural
Networks for Mobile Vision Applications, CoRR, abs/1704.04861, 2017. a
Iacono, M. J., Delamere, J. S., Mlawer, E. J., Shephard, M. W., Clough, S. A., and Collins, W. D.: Radiative forcing by long-lived greenhouse gases:
Calculations with the AER radiative transfer models, J. Geophys. Res.-Atmos., 113, D13103, https://doi.org/10.1029/2008JD009944,
2008a. a
Iacono, M. J., Delamere, J. S., Mlawer, E. J., Shephard, M. W., Clough, S. A.,
and Collins, W. D.: Radiative forcing by long-lived greenhouse gases:
Calculations with the AER radiative transfer models, J. Geophys. Res.-Atmos., 113, D13103, https://doi.org/10.1029/2008JD009944, 2008b. a
Innamorati, C., Ritschel, T., Weyrich, T., and Mitra, N. J.: Learning on the
Edge: Explicit Boundary Handling in CNNs, CoRR, abs/1805.03106, 2018. a
Kingma, D. P. and Ba, J.: Adam: A Method for Stochastic Optimization, CoRR,
abs/1412.6980, 2014. a
Krasnopolsky, V. M., Fox-Rabinovitz, M. S., and Chalikov, D. V.: New Approach to Calculation of Atmospheric Model Physics: Accurate and Fast
Neural Network Emulation of Longwave Radiation in a Climate
Model, Mon. Weather Rev., 133, 1370–1383, https://doi.org/10.1175/MWR2923.1,
2005. a
Krasnopolsky, V. M., Fox-Rabinovitz, M. S., and Belochitski, A. A.: Decadal
Climate Simulations Using Accurate and Fast Neural Network
Emulation of Full, Longwave and Shortwave, Radiation, Mon.
Weather Rev., 136, 3683–3695, https://doi.org/10.1175/2008MWR2385.1, 2008. a, b, c
Krasnopolsky, V. M., Fox-Rabinovitz, M. S., Hou, Y. T., Lord, S. J., and
Belochitski, A. A.: Accurate and Fast Neural Network Emulations of
Model Radiation for the NCEP Coupled Climate Forecast System:
Climate Simulations and Seasonal Predictions, Mon. Weather Rev.,
138, 1822–1842, https://doi.org/10.1175/2009MWR3149.1, 2009. a
Krizhevsky, A., Sutskever, I., and Hinton, G. E.: ImageNet Classification withDeep Convolutional Neural Networks, in: Proceedings of the 25th International Conference on Neural Information Processing Systems – Volume 1, NIPS'12, 1097–1105, Curran Associates Inc., USA, available at: http://dl.acm.org/citation.cfm?id=2999134.2999257 (last access: 1 December 2019),
2012a. a
Krizhevsky, A., Sutskever, I., and Hinton, G. E.: Imagenet classification with deep convolutional neural networks, in: Advances in neural information processing systems, 1097–1105, 2012b. a
Krizhevsky, A., Sutskever, I., and Hinton, G. E.: ImageNet Classification with Deep Convolutional Neural Networks, Commun. ACM, 60, 84–90, 2017. a
Lin, H. W., Tegmark, M., and Rolnick, D.: Why Does Deep and Cheap
Learning Work So Well?, J. Stat. Phys., 168,
1223–1247, https://doi.org/10.1007/s10955-017-1836-5, 2017. a
Liu, Y., Monteiro, J. M., and Caballero, R.: RadNet release for GMD, Zenodo, https://doi.org/10.5281/zenodo.3884964, 2019. a
Malik, M., Grosheintz, L., Mendonça, J. M., Grimm, S. L., Lavie, B., Kitzmann,
D., Tsai, S.-M., Burrows, A., Kreidberg, L., Bedell, M., Bean, J. L.,
Stevenson, K. B., and Heng, K.: Helios: An Open-Source,
Gpu-Accelerated Radiative Transfer Code For Self-Consistent
Exoplanetary Atmospheres, Astron. J., 153, 56,
https://doi.org/10.3847/1538-3881/153/2/56, 2017. a
Meador, W. E. and Weaver, W. R.: Two-Stream Approximations to Radiative
Transfer in Planetary Atmospheres: A Unified Description of
Existing Methods and a New Improvement, J. Atmos.
Sci., 37, 630–643,
https://doi.org/10.1175/1520-0469(1980)037<0630:TSATRT>2.0.CO;2, 1980. a
Mhaskar, H. N. and Poggio, T.: Deep vs. shallow networks: An approximation
theory perspective, Anal. Appl., 14, 829–848,
https://doi.org/10.1142/S0219530516400042, 2016. a
Mielikainen, J., Price, E., Huang, B., Huang, H. A., and Lee, T.: GPU
Compute Unified Device Architecture (CUDA)-based Parallelization
of the RRTMG Shortwave Rapid Radiative Transfer Model, IEEE J. Sel. Top. Appl.,
9, 921–931, https://doi.org/10.1109/JSTARS.2015.2427652, 2016. a, b, c
Monteiro, J. M. and Caballero, R.: The Climate Modelling Toolkit, in:
Proceedings of the 15th Python in Science Conference,
Austin, USA, 69–74, available at: http://conference.scipy.org/proceedings/scipy2016/joy_monteiro.html (last access: 1 December 2019),
2016. a
Monteiro, J. M., McGibbon, J., and Caballero, R.: sympl (v. 0.4.0) and climt (v. 0.15.3) – towards a flexible framework for building model hierarchies in Python, Geosci. Model Dev., 11, 3781–3794, https://doi.org/10.5194/gmd-11-3781-2018, 2018. a, b
O'Gorman, P. A. and Dwyer, J. G.: Using Machine Learning to Parameterize
Moist Convection: Potential for Modeling of Climate, Climate
Change, and Extreme Events, J. Adv. Model. Earth Sy., 10, 2548–2563, https://doi.org/10.1029/2018MS001351, 2018. a
O'Shea, K. and Nash, R.: An Introduction to Convolutional Neural Networks,
CoRR, abs/1511.08458, 2015. a
Pal, A., Mahajan, S., and Norman, M. R.: Using Deep Neural Networks as
Cost-Effective Surrogate Models for Super-Parameterized E3SM
Radiative Transfer, Geophys. Res. Lett., 46, 6069–6079,
https://doi.org/10.1029/2018GL081646, 2019. a, b, c
Palmer, T. N.: More reliable forecasts with less precise computations: a
fast-track route to cloud-resolved weather and climate simulators?, Philos. T. R. Soc. A, 372,
20130391, https://doi.org/10.1098/rsta.2013.0391, 2014.
a
Pincus, R. and Stevens, B.: Paths to accuracy for radiation parameterizations
in atmospheric models, J. Adv. Model. Earth. Sy., 5, 225–233, https://doi.org/10.1002/jame.20027, 2013. a
Price, E., Mielikainen, J., Huang, M., Huang, B., Huang, H. A., and Lee, T.:
GPU-Accelerated Longwave Radiation Scheme of the Rapid
Radiative Transfer Model for General Circulation Models
(RRTMG), IEEE J. Sel. Top. Appl., 7, 3660–3667, https://doi.org/10.1109/JSTARS.2014.2315771, 2014. a, b, c
Rasp, S., Pritchard, M. S., and Gentine, P.: Deep learning to represent subgrid
processes in climate models, P. Natl. Acad. Sci. USA,
115, 9684–9689, https://doi.org/10.1073/pnas.1810286115, 2018. a
Reed, K. A. and Jablonowski, C.: Idealized tropical cyclone simulations of
intermediate complexity: a test case for AGCMs, J. Adv. Model. Earth Sy., 4, M04001, https://doi.org/10.1029/2011MS000099, 2012. a
San, O. and Maulik, R.: Extreme learning machine for reduced order modeling of turbulent geophysical flows, Phys. Rev. E, 97, 042322,
https://doi.org/10.1103/PhysRevE.97.042322, 2018. a
Scher, S.: Toward Data-Driven Weather and Climate Forecasting:
Approximating a Simple General Circulation Model With Deep
Learning, Geophys. Res. Lett., 45, 12616–12622,
https://doi.org/10.1029/2018GL080704, 2018. a
Schneider, T., Lan, S., Stuart, A., and Teixeira, J.: Earth system modeling
2.0: A blueprint for models that learn from observations and targeted
high-resolution simulations, Geophys. Res. Lett., 44, 12–396, 2017. a
Sietsma, J. and Dow, R. J. F.: Creating artificial neural networks that
generalize, Neural Networks, 4, 67–79, https://doi.org/10.1016/0893-6080(91)90033-2,
1991. a
Simonyan, K. and Zisserman, A.: Very deep convolutional networks for
large-scale image recognition, arXiv preprint arXiv:1409.1556, 2014. a
Tan, M. and Le, Q. V.: EfficientNet: Rethinking Model Scaling for Convolutional
Neural Networks, CoRR, abs/1905.11946, 2019. a
Wang, Y., Zhao, Y., Jiang, J., and Zhang, H.: A Novel GPU-Based
Acceleration Algorithm for a Longwave Radiative Transfer Model,
Appl. Sci., 10, 649, https://doi.org/10.3390/app10020649, 2020. a, b
Wu, Y., Schuster, M., Chen, Z., Le, Q. V., Norouzi, M., Macherey, W., Krikun, M., Cao, Y., Gao, Q., Macherey, K., Klingner, J., Shah,
A., Johnson, M., Liu, X., Kaiser, L., Gouws, S., Kato, Y., Kudo, T., Kazawa, H., Stevens, K., Kurian, G., Patil, N., Wang, W., Young, C., Smith, J., Riesa, J., Rudnick, A., Vinyals, O., Corrado, G., Hughes, M., and Dean, J.: Google's Neural Machine Translation System:
Bridging the Gap between Human and Machine Translation, CoRR, abs/1609.08144,
2016. a
Yuval, J. and O'Gorman, P. A.: Use of machine learning to improve simulations
of climate, arXiv:2001.03151, 2020. a
Short summary
The calculation of atmospheric radiative transfer is the most computationally expensive part of climate models. Reducing this computational burden could potentially improve our ability to simulate the earth's climate at finer scales. We propose using a statistical model – a deep neural network – to compute approximate radiative transfer in the earth's atmosphere. We demonstrate a significant reduction in computational burden as compared to a traditional scheme, especially when using GPUs.
The calculation of atmospheric radiative transfer is the most computationally expensive part of...
