The Agricultural Model Intercomparison and Improvement Project (AgMIP) coordinates protocol-based model intercomparison studies, analyzing model performance in order to build trust in models, identify deficiencies and strategies for model improvements, and to quantify model uncertainties. Within AgMIP, the Global Gridded Crop Model Intercomparison (GGCMI) focuses on global-scale crop model applications, development and evaluation. Several phases of GGCMI are conducted in coordination with the Intersectoral Impact Model Intercomparison (ISIMIP). In this special issue, we summarize protocol and experiment description papers, as well as model development and evaluation papers of the different GGCMI phases. Participation in GGCMI is voluntary and open to all groups that are able to conduct the minimum set of simulations. Different crops and management systems are included in the analyses, and output data are published in open repositories.

A large number of machine learning studies are currently performed in Earth System science to extract information from observations or model data, to learn the dynamics of components of the Earth system, or to build tools to improve weather and climate predictions (for example in model post-processing). These studies require both the domain knowledge of Earth system scientists and the expertise of machine learning experts. The astounding success of machine learning in other scientific disciplines – such as computer vision, speech recognition, robotics, and autonomous driving – is not least a result of the free and open sharing of benchmark datasets and software, which allows easy reproducibility by the community as well as the quantitative comparison of the quality of machine learning solutions. Contrary to many other machine learning disciplines, Earth system datasets and validation methods require thorough documentation to achieve the goal of providing a fair and firm basis for model comparisons according to the accepted community standards. Therefore, the Earth system science community has expressed a need for dedicated benchmark datasets and machine learning tools to process Earth system data.

With this special issue, we offer a platform to develop and openly share these benchmark datasets together with specific machine learning tasks and evaluation metrics as well as innovative solutions for machine learning applications in Earth system science with the wider community of machine learners and environmental scientists.

The special issue is managed jointly by ESSD and GMD and invites manuscripts focusing either on the description of new benchmark datasets (ESSD) or new machine learning model developments (GMD). Dataset papers can reference model papers and vice versa, but this is not a requirement as long as both the datasets and model codes are made freely available and easily accessible. To be accepted as special issue contribution a manuscript must define a specific machine learning task including a loss function and evaluation metric and the authors must provide executable code which demonstrates the machine learning problem with a vanilla solution. The usual guidelines of both journals apply.

The climate record of the last millennium holds much promise for identifying the links between forcings and responses at the global, hemispheric and regional scale. The specifications for the Paleoclimate Model Intercomparison Project Phase III (PMIP3) and the Coupled Model Intercomparison Project Phase V (CMIP5) include controlled experiments for the period 850 to 1850 CE. The implementation of those experiments, the collation and development of the climate drivers over this period, and the assessment of the model responses are the focus of this special issue of Geoscientific Model Development.

In order to improve our current knowledge on the budgets of the two most important anthropogenic greenhouse gases, CO₂ and CH₄, a co-ordinated measurement campaign in the central European region was successfully carried out in May-June 2018 with the German research aircraft HALO and two smaller Cessna aircraft as its main experimental platforms. The goal of CoMet is to combine a suite of state-of-the-art airborne active (lidar) and passive remote sensors (spectrometer) with in situ instruments to provide regional-scale data about greenhouse gases (GHGs) which are urgently required for their accurate modelling.

During the intensive observation period, HALO research flights were performed along extended latitudinal transects to capture GHG gradients, over known regions of strong emissions, and over the ground-based remote-sensing sites of the Total Carbon Column Observing Network (TCCON). While HALO provides a larger-scale picture, the two Cessna aircraft concentrated on a region of prime interest: the Upper Silesian Coal Basin in Poland, which, due to hard coal mining activities, is one of the hot spots of anthropogenic methane emissions in Europe. Here, a variety of ground-based instruments additionally supported the CoMet mission: FTIR spectrometers, wind lidars, mobile vans, and small drones provided valuable information to quantify CH₄ emissions from coal mining activities. A model infrastructure (regional inverse modelling, chemistry-climate modelling with regional refinement) has been employed in forecast mode to optimize flight strategies and is now used to hindcast the campaign period to evaluate GHG fluxes and transport. The CoMet mission is also part of validation activities for existing passive remote sounding GHG satellites (GOSAT, Sentinel-5P, and OCO-2) and preparation of the first active CH₄ satellite mission MERLIN. For that reason, the CHARM-F lidar developed at DLR as an airborne MERLIN demonstrator was one of the key instruments. In addition, CoMet is intended to investigate methodologies for the synergistic combination for GHG measurements using lidar and passive remote sensing.

CMIP5 represents the most ambitious and computer-intensive model inter-comparison project ever attempted. Integrating a new generation of Earth system models and sharing the model results with a broad community has brought with it many significant technical challenges, along with new community-wide efforts to provide the necessary software infrastructure. This special issue will focus on the software that supports the scientific enterprise for CMIP5, including: couplers and coupling frameworks for Earth system models; the Common Information Model and Controlled Vocabulary for describing models and data; The development of the Earth System Grid Federation; the development of new portals for providing data access to different end-user communities; the scholarly publishing of datasets, and studies of the software development and testing processes used for the CMIP5 models. We especially welcome papers that offer comparative studies of the software approaches taken by different groups, and lessons learnt from community efforts to create shareable software components and frameworks.

In April 2017, the Senate of the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) established the Priority Programme “Polarimetric Radar Observations meet Atmospheric Modelling (PROM)” (SPP 2115). The programme is designed to run for 6 years and is now in the third funding year. The overall motivation of the SPP PROM is a better understanding and representation of cloud and precipitation processes in numerical weather prediction and climate models. Since 2015 the whole atmosphere over Germany has been monitored by 17 state-of-the-art polarimetric Doppler weather radar observations suitable to challenge the representation of cloud and precipitation processes in atmospheric models. Data assimilation merges observations and models for state estimation as a prerequisite for prediction and can be regarded as a smart interpolation between observations while exploiting the physical consistency of atmospheric models as mathematical constraints. However, considerable knowledge gaps exist both in radar polarimetry and atmospheric models, which impede the full exploitation of the triangle radar-polarimetry–atmospheric-models–data-assimilation. Objectives of SPP PROM and thus the envisioned publications within the special issue are (1) the exploitation of radar polarimetry for quantitative process detection in precipitating clouds and for model evaluation, (2) the improvement of cloud and precipitation schemes in atmospheric models based on process fingerprints detectable in polarimetric observations, (3) monitoring of the energy budget evolution due to phase changes in the cloudy, precipitating atmosphere for a better understanding of its dynamics, (4) the generation of precipitation system analyses by the assimilation of polarimetric radar observations into atmospheric models for weather forecasting, and (5) radar-based detection of the initiation of convection for the improvement of thunderstorm prediction.

This special issue groups together documentation papers for successive releases of the Met Office Unified Model (UM) Global Atmosphere (GA) and JULES Global Land (GL) configurations. GA and GL are science configurations of the UM and JULES developed for use across weather prediction and climate research timescales. Each paper presents a scientific description of the latest configuration, a fuller description of incremental changes since the previous configuration and a brief summary of their performance.

The Joint UK Land Environment Simulator (JULES) is a process-based, community model that simulates the fluxes of carbon, water, energy and momentum between the land surface and the atmosphere. JULES-crop is a parameterisation of crops within JULES which was developed with the dual aim of being able to simulate the impact of weather and climate on crop productivity and the impact that croplands have on weather and climate. This ongoing special issue collects together papers on model development, evaluation and application of JULES-crop.

The Joint UK Land Environment Simulator (JULES) is a process-based community model that simulates the fluxes of carbon, water, energy and momentum between the land surface and the atmosphere. Configurations of JULES underpin important science applications including the Earth system, global land, hydrology, UK environmental prediction and climate impact modelling. This indefinite special issue covers all publications covering JULES model development, evaluation and benchmarking; the documentation of the code includes schemes and parameterisations. Also included are papers covering the major scientific configurations of JULES that should be a starting point for model application and development. This special issue supports the information available through the JULES code repository (https://code.metoffice.gov.uk/trac/jules). Licences are available here http://jules-lsm.github.io/access_req/JULES_Licence.pdf.

Geoscience model infrastructure provides architecture and services for building physics-based numerical models and combining multiple models into a coupled system. Examples of model infrastructure are software libraries and frameworks that provide data structures and functions for parallel data transfer and grid interpolation, define a common interface for model components, manage a model's control flow, abstract details of parallel programming, or provide supporting capabilities such as configuration management and file I/O. The special issue of GMD is dedicated to exploring all aspects of model infrastructure integration and interoperability. Model infrastructure integration is the software engineering process of adopting an infrastructure package into a numerical model's codebase, to address new scientific requirements, enable interoperability with other models, or to improve a model's performance. Model infrastructure interoperability is required whenever multiple infrastructure packages must interact in some way in a coupled system, for example, to couple model components that originate from different scientific communities into a single cohesive system. Topics of interest for this issue include but are not limited to the following: the design and implementation of geoscience model infrastructure software; experience reports on integrating infrastructure into models; modelling systems that span scientific communities; framework interoperability, metadata and semantic mediation; performance on emerging hardware platforms; configuration and build management; and tools that support infrastructure integration, code refactoring, verification, or debugging.

NEMO (Nucleus for European Modelling of the Ocean) is a state-of-the-art modelling framework for oceanographic research, operational oceanography seasonal forecast and climate studies. The NEMO components are NEMO-OPA (the "blue" ocean, modelling the ocean dynamics and solving the primitive equations); NEMO-LIM (the "white ocean" for modelling sea-ice), and NEMO-TOP (the "green ocean" for modelling biogeochemistry). NEMO also includes grid refinement software (AGRIF) and an assimilation component (linear-tangent NEMO-TAM, Observational operators NEMO-OBS, and increment NEMO-ASM). The "blue ocean" component is fundamental to all users. NEMO can also be interfaced to a number of other components such as atmosphere models or alternative other models of sea-ice or biogeochemistry, to enable Earth system modelling.

This Special Issue aims to collect technical and scientific manuscripts dealing with evaluation of model skill and performance as well as development of NEMO components. Submitted manuscripts can cover a wide variety of topics, including process studies, new parameterizations, implementation of new model features and new NEMO configurations. The main scope is to collect relevant and state of the art manuscripts to provide the NEMO users with a single portal to search, discover and understand about the NEMO modelling framework potential and evolution and submit their contributions.

Unlike in traditional bulk and bin microphysical approaches, atmospheric particles like hydrometeors and aerosol particles are not described by field variables but by computational particles. Over the last years, such particle-based methods have become more and more popular in aerosol and cloud physics. The special issue is open to all submissions with a clear focus on model description or evaluation or on studies specifically exploiting the advantages of particle-based methods. Contributions of any particle-based model related to aerosol and cloud modelling are welcome. This inter-journal special issue of GMD and ACP serves as a thematic collection and intends to increase visibility and foster collaboration among researchers.

The 2019 Raikoke eruption emitted an estimated 1.5±0.2 Tg of volcanic sulfur dioxide (SO₂) into the atmosphere, which makes it the largest volcanic SO₂ perturbation in the upper troposphere and lower stratosphere since the 2011 eruption of Nabro. The 2019 Raikoke eruption triggered numerous discussions among scientists involved in the SPARC Volcano Response (VolRes) initiative and provides an excellent opportunity to test our understanding of processes involved in the formation, growth and removal of volcanic aerosols in the atmosphere. Remote sensing measurements from geostationary and low-orbiting satellites with ultraviolet, visible, and infrared sensors (HIMAWARI-8, OMPS/NPP, OMI/Aura, TROPOMI/SP5, IASI/MeteopA-B-C, AIRS/Aqua, CALIOP/CALIPSO, SAGE III/ISS) provided a wealth of information on the SO₂ concentration, ash and sulfate aerosol particle optical properties from the initial injection through to the growth and removal of aerosols. This eruption therefore provides an excellent opportunity to inter-compare satellite observations and validate the different retrievals for SO₂ and volcanic aerosol properties from new satellite sensors such as TROPOMI. After the initial injection into altitudes up to 15–16 km, the plume reached 25–26 km altitude in late 2019, which corresponds to a fast vertical ascent that appears unusual compared to previous eruptions. The fast vertical ascent of the 2019 Raikoke plume may suggest an influence of smoke aerosol from large wildfires that took place in Canada and Siberia around the time of the eruption. The veil of aerosol particles from Raikoke 2019 has noticeably reduced the transparency of the stratosphere and detailed information and measurements on the aerosol properties will be useful to constrain atmospheric and climate model simulations.

The Australian Community Climate and Earth System Simulator, ACCESS, is Australia’s national modelling system for all timescales from weather forecasting to climate projections. ACCESS is a family of related models, configured for specific applications, to meet operational and research needs. ACCESS is built from core component models: the UK Met Office Unified Model (UM) for the atmosphere, the Community Atmosphere Biosphere Land Exchange (CABLE) model for the land, the US NOAA/GFDL Modular Ocean Model (MOM) for the ocean and the LANL CICE model for sea ice. Additional modules support biogeochemistry and atmospheric chemistry. Different ACCESS model configurations may use a subset of the core models for atmosphere-only or ocean-only simulations or may replace component models to meet the needs of a specific application. This special issue brings together papers that describe ACCESS model configurations and code development, as well as papers that evaluate the performance of the model across any ACCESS application.

The CSIRO Mk3L climate system model is a computationally-efficient coupled general circulation model, designed primarily for the study of climate variability and change on millennial timescales. The model distribution is freely available to the research community. This Special Issue allows the history and evolution of Mk3L to be comprehensively documented within a single issue of a journal. Papers are welcome which describe and evaluate components of the model, and which describe and evaluate subsequent enhancements to the modelling system. The issue is also intended for papers which document specific experiments, particularly those which contribute towards Model Intercomparison Projects.

The Canadian Earth System Model, CanESM, is a global coupled Earth System Model designed to simulate climate variability and change on timescales from seasons to centuries. CanESM is developed at the Canadian Centre for Climate Modelling and Analysis (CCCma), Environment and Climate Change Canada, with inputs from other government and university stakeholders. A significant update to the model, CanESM5, has been developed for inclusion in the upcoming Coupled Model Intercomparison Project Phase 6. Beyond CMIP6, the new version of the model will provide an updated tool for climate science applications at CCCma and within Canada and will be submitted for possible integration into the operational Canadian Seasonal to Interseasonal Prediction System (CanSIPS). CanESM5 consists of the Canadian atmospheric model (CanAM) with an embedded land surface scheme and terrestrial biogeochemistry (CLASS-CTEM) and an off-the-shelf but customized ocean model (CanNEMO), which includes sea-ice and Canadian-developed ocean biogeochemistry modules (CMOC and CanOE). The components communicate through a new custom coupler, CanCPL. This special issue hosts the documentation and model evaluation papers for CanESM5.

FAMOUS is climate model based on the widely used "HadCM3" atmosphere–ocean general circulation code, a version of the UK Met Office Unified Model. Run at a lower resolution than HadCM3, its computational requirements make it suitable for large ensembles and millennial-scale climate simulations. This ongoing special issue collects technical documentation and evaluations of the model climatology as FAMOUS is developed and coupled to models of other Earth system components.

This special issue gathers together papers documenting the development of the Firedrake system, and geoscientific models built on Firedrake.

Firedrake provides a model development system which is both high productivity and high performance. Users write high-level code in Python describing the mathematical formulation of a model. The low-level, high-performance, parallel implementation of the algorithm is then automatically generated by a sequence of domain-specific compilers. The user writes maths and gets simulation. Firedrake provides users with a vast range of finite element discretisations, including the compatible finite-element methods which accurately represent the critical force balances in large-scale geoscientific problems. Other important features for the geoscientific user include curved elements and layered meshes, which are key to accurate atmosphere and ocean modelling.

The Lagrangian particle dispersion model FLEXPART was originally designed for calculating the long-range and mesoscale dispersion of air pollutants from point sources, such as after an accident in a nuclear power plant. The model has now evolved into a comprehensive tool for atmospheric transport modelling and analysis. Its application fields are extended to a range of atmospheric transport processes for both atmospheric gases and aerosols, e.g. greenhouse gases, ozone depletion substances, short-lived climate forces like black carbon, volcanic ash and gases as well as studies of the water cycle. It supports ECMWF, NCEP, mesoscale (WRF) and climate (NorESM, CAM) wind fields in both forward and backwards modes. FLEXTPART is open source (https://www.flexpart.eu).

Loop is an open-source integrated and interoperable platform enabling 3D stochastic geological modelling. This special issue groups contributions to the Loop platform in the form of new algorithmic developments and applications. Contributions are open to anyone using, developing, or contributing to the Loop platform.

This special issue is a collection of subsequent extensions and developments of the LPJmL model as well as model evaluation papers. The idea is to have collected in one place all the development stages of the LPJmL model that are described (and will continue to be described) in individual GMD papers.

Volcanic eruptions are one of the major natural factors influencing climate variability at interannual to multidecadal timescales. However, simulating volcanically forced climate variability is a challenging task for climate models and one of the major uncertainties in near-term climate predictions. The Model Intercomparison Project on the climatic response to Volcanic forcing (VolMIP) is an endorsed contribution to the sixth phase of the Coupled Model Intercomparison Project. This multi-journal special issue on VolMIP aims at collecting relevant research results obtained within the VolMIP framework, and specifically concerning different aspects of the radiative and dynamical climatic response to volcanic forcing, detailed description of effects of different implementation of volcanic forcing in current climate models, aspects concerning the dynamical and chemical atmospheric response to volcanic aerosols simulated by global aerosol models, and comparison between reconstructed and simulated climate evolution after major eruptions.

The Modular Earth Submodel System (MESSy) is a multi-institutional project providing a strategy and the software for developing Earth System Models (ESMs) with highly flexible complexity.

The strategy follows a bottom-up approach, meaning that the various processes and diagnostic tools are implemented as so-called submodels, which are technically independent of each other and strictly separated from the underlying technical model infrastructure, such as memory management, input/output, flow-control, etc.

The MESSy software provides generalized interfaces for the standardized control and interconnection (coupling) of these submodels.

The present time-unlimited Special Issue hosts scientific and technical documentation and evaluation manuscripts concerned with the Modular Earth Submodel System and the models build upon it. Moreover, it comprises manuscripts about scientific applications involving these models.

The Norwegian Earth System Model (NorESM) is a global, coupled model system for the physical climate system, which can be run with various degrees of interactions with biogeochemical processes in the earth system. NorESM is developed as a nationally coordinated effort in Norway, but important parts of the model code are imported from the Community Climate System Model (CCSM) and Community Earth System Model (CESM) projects operated at NCAR on behalf of UCAR in the USA.

The NorESM publications in this special issue address a range of NorESM versions. The first set of model versions delivered results to CMIP5. NorESM1-M is run concentration-driven for greenhouse gases (GHGs) and is based on CCSM4 (released 1 April 2010), while NorESM1-ME can be run emission-driven for GHGs and is based on CESM1 (released 1 July 2010). A low-resolution version, NorESM1-L, was developed mainly for paleo-climate simulations. New versions of NorESM are underway: NorESM1.X, where X indicates updates of the NorESM1 versions, and NorESM2, which is intended to contribute to CMIP6. Further versions will follow thereafter. NorESM includes the following: its own developed code for chemistry–aerosol–cloud–radiation interactions (CAM-Oslo) and enhancements of the dynamics/physics of the atmospheric module; alternative parameterization of surface turbulent fluxes; an isopycnic coordinate ocean model originating from the Miami Isopycnic Coordinate Ocean Model (MICOM) but developed further; and the HAMburg Ocean Carbon Cycle (HAMOCC) model developed at the Max Plank Institute for Meteorology, Hamburg and adapted to the isopycnic coordinate ocean model framework.

Papers developed for full validation (CMIP-DECK) or more specific evaluation of the NorESM versions and further developments of these are welcome in this special issue. Authors intending to contribute papers to this special issue should contact the coordinators (Mats Bentsen and Michael Schulz), e.g., to ensure the consistency of version names and numbers.

SimSphere is a one-dimensional soil–vegetation–atmosphere transfer model devoted to the study of land surface interactions of the Earth’s system. Since its early development, the model has become highly variable in its application use.

Apart from its use as an educational tool at several universities worldwide, SimSphere is used in a number of research studies related to the examination of hypothetical scenarios examining land surface processes and feedbacks. It is also used synergistically with Earth observation (EO) data to retrieve spatiotemporal estimates of energy fluxes and surface soil moisture, involving exploration studies on the development of related operational products.

This special issue hosts contributions concerned with descriptions of further upgrades of SimSphere or its exploitation in any way. It comprises articles on model developments or applications involving the model; this includes – but is not limited to – studies exploring hypothetical scenario examination, model validation, sensitivity analysis and synergies of it with EO data.

Scope: The Transport Matrix Method (TMM) is a computationally efficient ``offline'' scheme for simulation of ocean biogeochemical tracers. The TMM essentially encapsulates the circulation of any ocean model, including sub-grid-scale parameterizations, as a sparse matrix, and advection-diffusion as a sequence of sparse matrix-vector products that can be efficiently carried out on modern distributed-memory computers. Simulations carried out with the TMM can potentially be orders of magnitude faster than the underlying ocean circulation model. The flexible architecture of the TMM framework also allows ``mixing and matching'' of circulations and biogeochemical models. The TMM code, various biogeochemical models, and transport matrices from different ocean models are freely available to download from https://github.com/samarkhatiwala/tmm.

The aim of this Special Issue is to bring together under one roof papers using the TMM as the underlying simulation method. This can range from manuscripts documenting various technical aspects of the TMM framework to those describing new biogeochemical models/parameterizations and their application.

The atmospheric chemistry box modeling system CAABA/MECCA is open-source code published under the GNU General Public License (GPL). The system is based on the box model CAABA (Chemistry As A Boxmodel Application). The chemistry code is provided by the sub-model MECCA (Module Efficiently Calculating the Chemistry of the Atmosphere). Additional sub-models provide the code to calculate photolysis, trajectories, isotope tagging and other features.

iLOVECLIM is an intermediate complexity fully coupled climate Earth system model that aims at computation and understanding of the climate system on a millennial timescale. It is a code fork from the LOVECLIM climate model version 1.2. From its forerunner, iLOVECLIM retains only the physical climate components (atmosphere–ocean–terrestrial vegetation modules). It is developed further to progressively include the components necessary for multi-millennia palaeoclimate and future climate experiments. As such, iLOVECLIM is a tool designed to enhance the integration of model simulations and (palaeo-)data, with an emphasis on the simulation of isotopic tracers throughout all components of the climate system, as indicated by the i- prefix. The present, time-unlimited special issue hosts the technical documentation of the current version of iLOVECLIM as well as model evaluation manuscripts.

Fire is an essential feature of terrestrial ecosystems and plays an important role in the Earth system. Fires are regulated by climate, vegetation characteristics, and human activity and also feedback to them in multiple ways. For example, fires can shape vegetation composition and structure; adjust land carbon, nutrient, water, and energy cycles; change atmospheric composition, chemistry, and physics; and affect air quality and human health. Elevated fire activity in 2019 across the arctic, Amazon, and other regions underscores the urgency of a quantitative understanding of fire as an Earth system process that interacts with vegetation, climate, and humans. The EGU2020 B3.17 session on the role of fire in the Earth system received the most abstracts in the Biogeosciences division this year and was highly attended during the live chat. The abstracts cover several aspects of interactions between fire and the biosphere, atmosphere, and humans across various temporal and spatial scales using modelling, field and laboratory observations, and remote sensing. Given that the topic is highly interdisciplinary and overarching, we propose an inter-journal (ESD/BG/ACP/GMD/NHESS) special issue for this session with ESD as the lead journal, and we also encourage the submission of topic-related studies outside the EGU fire session. The special issue will bring together new advances in understanding the feedbacks and interactions between fire and other components of the Earth system at all temporal and spatial scales using various methods, including (1) impacts of fire on weather, climate, and atmospheric chemistry; (2) interactions between fire, the biogeochemical cycle, vegetation composition and structure, and land water and energy budgets; (3) influence of humans on fire and vice versa; (4) fire characteristics (e.g., fire duration, emission factor, emission height, smoke transport); (5) spatial and temporal changes of fires in the past, present, and future; (6) fire products and models, their validation, and error/bias assessment; and (7) analytical tools designed to enhance situational awareness among fire practitioners and early warning systems, addressing specific needs of operational fire behaviour modelling.

The prediction of the winter weather over complex terrain is quite challenging due to the highly variable nature of winds, visibility, and snowfall. As a World Meteorological Organization (WMO) World Weather Research Program (WWRP) Research Demonstration Project (RDP) and Forecast Demonstration Project (FDP), ICE-POP 2018 (International Collaborative Experiments for PyeongChang 2018 Olympic and Paralympic winter games) was held in the PyeongChang region from November 2017 to April 2018 with contributions from 29 agencies from 12 countries. The region was quite unique for observing winter weathers that are influenced by cold air and warm ocean interaction, sudden uplifting by steep terrains near the coast, and modulation by complex terrains. The main scientific goal was to understand the precipitation processes in this unique region during the cold season and to evaluate/improve forecasting from numerical models based on intensive observations. Dense observational networks of upper air observation (eight soundings, two wind profilers, shipborne sounding, and dropsonde), remote sensing (three X-Pol radars, one Ku/Ka-Pol radar and three S-Pol, one S-band, two C-band, and three Doppler lidars), microphysical observation (2DVD, MASC, PIP, Parsivel, MRR, POSS, Pluvio), and surface stations (64 stations) were implemented, in particular, to observe the evolution of precipitation along and across atmospheric flows. The field experiment and real-time forecast demonstration ended and the second phase of the experiment has started for better understanding of the microphysical processes, their better representation in the numerical modeling, and further improvement of winter weather prediction through various international collaborations.

The main purposes of the special issue are

1) to document the scientific findings on the winter weather during the forecast demonstration project

2) to share scientific knowledge on processes of winter weathers that have been investigated with unprecedented dense observational networks,

3) to share current status and improved knowledge of forecasting of winter weathers, and

4) to document new retrieval and quality control methods of the operational and advanced instruments.

The special issue will include all manuscripts related to observational data, products, NWP modeling, researches on observational instrumentation, process/mechanism study, reanalysis, integration of observation and numerical modeling, and prediction of the winter weathers.

Unlike in traditional bulk and bin microphysical approaches, atmospheric particles like hydrometeors and aerosol particles are not described by field variables but by computational particles. Over the last years, such particle-based methods have become more and more popular in aerosol and cloud physics. The special issue is open to all submissions with a clear focus on model description or evaluation or on studies specifically exploiting the advantages of particle-based methods. Contributions of any particle-based model related to aerosol and cloud modelling are welcome. This inter-journal special issue of GMD and ACP serves as a thematic collection and intends to increase visibility and foster collaboration among researchers.

In April 2017, the Senate of the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) established the Priority Programme “Polarimetric Radar Observations meet Atmospheric Modelling (PROM)” (SPP 2115). The programme is designed to run for 6 years and is now in the third funding year. The overall motivation of the SPP PROM is a better understanding and representation of cloud and precipitation processes in numerical weather prediction and climate models. Since 2015 the whole atmosphere over Germany has been monitored by 17 state-of-the-art polarimetric Doppler weather radar observations suitable to challenge the representation of cloud and precipitation processes in atmospheric models. Data assimilation merges observations and models for state estimation as a prerequisite for prediction and can be regarded as a smart interpolation between observations while exploiting the physical consistency of atmospheric models as mathematical constraints. However, considerable knowledge gaps exist both in radar polarimetry and atmospheric models, which impede the full exploitation of the triangle radar-polarimetry–atmospheric-models–data-assimilation. Objectives of SPP PROM and thus the envisioned publications within the special issue are (1) the exploitation of radar polarimetry for quantitative process detection in precipitating clouds and for model evaluation, (2) the improvement of cloud and precipitation schemes in atmospheric models based on process fingerprints detectable in polarimetric observations, (3) monitoring of the energy budget evolution due to phase changes in the cloudy, precipitating atmosphere for a better understanding of its dynamics, (4) the generation of precipitation system analyses by the assimilation of polarimetric radar observations into atmospheric models for weather forecasting, and (5) radar-based detection of the initiation of convection for the improvement of thunderstorm prediction.

A large number of machine learning studies are currently performed in Earth System science to extract information from observations or model data, to learn the dynamics of components of the Earth system, or to build tools to improve weather and climate predictions (for example in model post-processing). These studies require both the domain knowledge of Earth system scientists and the expertise of machine learning experts. The astounding success of machine learning in other scientific disciplines – such as computer vision, speech recognition, robotics, and autonomous driving – is not least a result of the free and open sharing of benchmark datasets and software, which allows easy reproducibility by the community as well as the quantitative comparison of the quality of machine learning solutions. Contrary to many other machine learning disciplines, Earth system datasets and validation methods require thorough documentation to achieve the goal of providing a fair and firm basis for model comparisons according to the accepted community standards. Therefore, the Earth system science community has expressed a need for dedicated benchmark datasets and machine learning tools to process Earth system data.

With this special issue, we offer a platform to develop and openly share these benchmark datasets together with specific machine learning tasks and evaluation metrics as well as innovative solutions for machine learning applications in Earth system science with the wider community of machine learners and environmental scientists.

The special issue is managed jointly by ESSD and GMD and invites manuscripts focusing either on the description of new benchmark datasets (ESSD) or new machine learning model developments (GMD). Dataset papers can reference model papers and vice versa, but this is not a requirement as long as both the datasets and model codes are made freely available and easily accessible. To be accepted as special issue contribution a manuscript must define a specific machine learning task including a loss function and evaluation metric and the authors must provide executable code which demonstrates the machine learning problem with a vanilla solution. The usual guidelines of both journals apply.

Loop is an open-source integrated and interoperable platform enabling 3D stochastic geological modelling. This special issue groups contributions to the Loop platform in the form of new algorithmic developments and applications. Contributions are open to anyone using, developing, or contributing to the Loop platform.

The prediction of the winter weather over complex terrain is quite challenging due to the highly variable nature of winds, visibility, and snowfall. As a World Meteorological Organization (WMO) World Weather Research Program (WWRP) Research Demonstration Project (RDP) and Forecast Demonstration Project (FDP), ICE-POP 2018 (International Collaborative Experiments for PyeongChang 2018 Olympic and Paralympic winter games) was held in the PyeongChang region from November 2017 to April 2018 with contributions from 29 agencies from 12 countries. The region was quite unique for observing winter weathers that are influenced by cold air and warm ocean interaction, sudden uplifting by steep terrains near the coast, and modulation by complex terrains. The main scientific goal was to understand the precipitation processes in this unique region during the cold season and to evaluate/improve forecasting from numerical models based on intensive observations. Dense observational networks of upper air observation (eight soundings, two wind profilers, shipborne sounding, and dropsonde), remote sensing (three X-Pol radars, one Ku/Ka-Pol radar and three S-Pol, one S-band, two C-band, and three Doppler lidars), microphysical observation (2DVD, MASC, PIP, Parsivel, MRR, POSS, Pluvio), and surface stations (64 stations) were implemented, in particular, to observe the evolution of precipitation along and across atmospheric flows. The field experiment and real-time forecast demonstration ended and the second phase of the experiment has started for better understanding of the microphysical processes, their better representation in the numerical modeling, and further improvement of winter weather prediction through various international collaborations.

The main purposes of the special issue are

1) to document the scientific findings on the winter weather during the forecast demonstration project

2) to share scientific knowledge on processes of winter weathers that have been investigated with unprecedented dense observational networks,

3) to share current status and improved knowledge of forecasting of winter weathers, and

4) to document new retrieval and quality control methods of the operational and advanced instruments.

The special issue will include all manuscripts related to observational data, products, NWP modeling, researches on observational instrumentation, process/mechanism study, reanalysis, integration of observation and numerical modeling, and prediction of the winter weathers.

Fire is an essential feature of terrestrial ecosystems and plays an important role in the Earth system. Fires are regulated by climate, vegetation characteristics, and human activity and also feedback to them in multiple ways. For example, fires can shape vegetation composition and structure; adjust land carbon, nutrient, water, and energy cycles; change atmospheric composition, chemistry, and physics; and affect air quality and human health. Elevated fire activity in 2019 across the arctic, Amazon, and other regions underscores the urgency of a quantitative understanding of fire as an Earth system process that interacts with vegetation, climate, and humans. The EGU2020 B3.17 session on the role of fire in the Earth system received the most abstracts in the Biogeosciences division this year and was highly attended during the live chat. The abstracts cover several aspects of interactions between fire and the biosphere, atmosphere, and humans across various temporal and spatial scales using modelling, field and laboratory observations, and remote sensing. Given that the topic is highly interdisciplinary and overarching, we propose an inter-journal (ESD/BG/ACP/GMD/NHESS) special issue for this session with ESD as the lead journal, and we also encourage the submission of topic-related studies outside the EGU fire session. The special issue will bring together new advances in understanding the feedbacks and interactions between fire and other components of the Earth system at all temporal and spatial scales using various methods, including (1) impacts of fire on weather, climate, and atmospheric chemistry; (2) interactions between fire, the biogeochemical cycle, vegetation composition and structure, and land water and energy budgets; (3) influence of humans on fire and vice versa; (4) fire characteristics (e.g., fire duration, emission factor, emission height, smoke transport); (5) spatial and temporal changes of fires in the past, present, and future; (6) fire products and models, their validation, and error/bias assessment; and (7) analytical tools designed to enhance situational awareness among fire practitioners and early warning systems, addressing specific needs of operational fire behaviour modelling.

The 2019 Raikoke eruption emitted an estimated 1.5±0.2 Tg of volcanic sulfur dioxide (SO₂) into the atmosphere, which makes it the largest volcanic SO₂ perturbation in the upper troposphere and lower stratosphere since the 2011 eruption of Nabro. The 2019 Raikoke eruption triggered numerous discussions among scientists involved in the SPARC Volcano Response (VolRes) initiative and provides an excellent opportunity to test our understanding of processes involved in the formation, growth and removal of volcanic aerosols in the atmosphere. Remote sensing measurements from geostationary and low-orbiting satellites with ultraviolet, visible, and infrared sensors (HIMAWARI-8, OMPS/NPP, OMI/Aura, TROPOMI/SP5, IASI/MeteopA-B-C, AIRS/Aqua, CALIOP/CALIPSO, SAGE III/ISS) provided a wealth of information on the SO₂ concentration, ash and sulfate aerosol particle optical properties from the initial injection through to the growth and removal of aerosols. This eruption therefore provides an excellent opportunity to inter-compare satellite observations and validate the different retrievals for SO₂ and volcanic aerosol properties from new satellite sensors such as TROPOMI. After the initial injection into altitudes up to 15–16 km, the plume reached 25–26 km altitude in late 2019, which corresponds to a fast vertical ascent that appears unusual compared to previous eruptions. The fast vertical ascent of the 2019 Raikoke plume may suggest an influence of smoke aerosol from large wildfires that took place in Canada and Siberia around the time of the eruption. The veil of aerosol particles from Raikoke 2019 has noticeably reduced the transparency of the stratosphere and detailed information and measurements on the aerosol properties will be useful to constrain atmospheric and climate model simulations.

The Agricultural Model Intercomparison and Improvement Project (AgMIP) coordinates protocol-based model intercomparison studies, analyzing model performance in order to build trust in models, identify deficiencies and strategies for model improvements, and to quantify model uncertainties. Within AgMIP, the Global Gridded Crop Model Intercomparison (GGCMI) focuses on global-scale crop model applications, development and evaluation. Several phases of GGCMI are conducted in coordination with the Intersectoral Impact Model Intercomparison (ISIMIP). In this special issue, we summarize protocol and experiment description papers, as well as model development and evaluation papers of the different GGCMI phases. Participation in GGCMI is voluntary and open to all groups that are able to conduct the minimum set of simulations. Different crops and management systems are included in the analyses, and output data are published in open repositories.

In order to improve our current knowledge on the budgets of the two most important anthropogenic greenhouse gases, CO₂ and CH₄, a co-ordinated measurement campaign in the central European region was successfully carried out in May-June 2018 with the German research aircraft HALO and two smaller Cessna aircraft as its main experimental platforms. The goal of CoMet is to combine a suite of state-of-the-art airborne active (lidar) and passive remote sensors (spectrometer) with in situ instruments to provide regional-scale data about greenhouse gases (GHGs) which are urgently required for their accurate modelling.

During the intensive observation period, HALO research flights were performed along extended latitudinal transects to capture GHG gradients, over known regions of strong emissions, and over the ground-based remote-sensing sites of the Total Carbon Column Observing Network (TCCON). While HALO provides a larger-scale picture, the two Cessna aircraft concentrated on a region of prime interest: the Upper Silesian Coal Basin in Poland, which, due to hard coal mining activities, is one of the hot spots of anthropogenic methane emissions in Europe. Here, a variety of ground-based instruments additionally supported the CoMet mission: FTIR spectrometers, wind lidars, mobile vans, and small drones provided valuable information to quantify CH₄ emissions from coal mining activities. A model infrastructure (regional inverse modelling, chemistry-climate modelling with regional refinement) has been employed in forecast mode to optimize flight strategies and is now used to hindcast the campaign period to evaluate GHG fluxes and transport. The CoMet mission is also part of validation activities for existing passive remote sounding GHG satellites (GOSAT, Sentinel-5P, and OCO-2) and preparation of the first active CH₄ satellite mission MERLIN. For that reason, the CHARM-F lidar developed at DLR as an airborne MERLIN demonstrator was one of the key instruments. In addition, CoMet is intended to investigate methodologies for the synergistic combination for GHG measurements using lidar and passive remote sensing.

This special issue is a collection of subsequent extensions and developments of the LPJmL model as well as model evaluation papers. The idea is to have collected in one place all the development stages of the LPJmL model that are described (and will continue to be described) in individual GMD papers.

The Australian Community Climate and Earth System Simulator, ACCESS, is Australia’s national modelling system for all timescales from weather forecasting to climate projections. ACCESS is a family of related models, configured for specific applications, to meet operational and research needs. ACCESS is built from core component models: the UK Met Office Unified Model (UM) for the atmosphere, the Community Atmosphere Biosphere Land Exchange (CABLE) model for the land, the US NOAA/GFDL Modular Ocean Model (MOM) for the ocean and the LANL CICE model for sea ice. Additional modules support biogeochemistry and atmospheric chemistry. Different ACCESS model configurations may use a subset of the core models for atmosphere-only or ocean-only simulations or may replace component models to meet the needs of a specific application. This special issue brings together papers that describe ACCESS model configurations and code development, as well as papers that evaluate the performance of the model across any ACCESS application.

The atmospheric chemistry box modeling system CAABA/MECCA is open-source code published under the GNU General Public License (GPL). The system is based on the box model CAABA (Chemistry As A Boxmodel Application). The chemistry code is provided by the sub-model MECCA (Module Efficiently Calculating the Chemistry of the Atmosphere). Additional sub-models provide the code to calculate photolysis, trajectories, isotope tagging and other features.

This special issue gathers together papers documenting the development of the Firedrake system, and geoscientific models built on Firedrake.

Firedrake provides a model development system which is both high productivity and high performance. Users write high-level code in Python describing the mathematical formulation of a model. The low-level, high-performance, parallel implementation of the algorithm is then automatically generated by a sequence of domain-specific compilers. The user writes maths and gets simulation. Firedrake provides users with a vast range of finite element discretisations, including the compatible finite-element methods which accurately represent the critical force balances in large-scale geoscientific problems. Other important features for the geoscientific user include curved elements and layered meshes, which are key to accurate atmosphere and ocean modelling.

The Joint UK Land Environment Simulator (JULES) is a process-based community model that simulates the fluxes of carbon, water, energy and momentum between the land surface and the atmosphere. Configurations of JULES underpin important science applications including the Earth system, global land, hydrology, UK environmental prediction and climate impact modelling. This indefinite special issue covers all publications covering JULES model development, evaluation and benchmarking; the documentation of the code includes schemes and parameterisations. Also included are papers covering the major scientific configurations of JULES that should be a starting point for model application and development. This special issue supports the information available through the JULES code repository (https://code.metoffice.gov.uk/trac/jules). Licences are available here http://jules-lsm.github.io/access_req/JULES_Licence.pdf.

The Canadian Earth System Model, CanESM, is a global coupled Earth System Model designed to simulate climate variability and change on timescales from seasons to centuries. CanESM is developed at the Canadian Centre for Climate Modelling and Analysis (CCCma), Environment and Climate Change Canada, with inputs from other government and university stakeholders. A significant update to the model, CanESM5, has been developed for inclusion in the upcoming Coupled Model Intercomparison Project Phase 6. Beyond CMIP6, the new version of the model will provide an updated tool for climate science applications at CCCma and within Canada and will be submitted for possible integration into the operational Canadian Seasonal to Interseasonal Prediction System (CanSIPS). CanESM5 consists of the Canadian atmospheric model (CanAM) with an embedded land surface scheme and terrestrial biogeochemistry (CLASS-CTEM) and an off-the-shelf but customized ocean model (CanNEMO), which includes sea-ice and Canadian-developed ocean biogeochemistry modules (CMOC and CanOE). The components communicate through a new custom coupler, CanCPL. This special issue hosts the documentation and model evaluation papers for CanESM5.

Volcanic eruptions are one of the major natural factors influencing climate variability at interannual to multidecadal timescales. However, simulating volcanically forced climate variability is a challenging task for climate models and one of the major uncertainties in near-term climate predictions. The Model Intercomparison Project on the climatic response to Volcanic forcing (VolMIP) is an endorsed contribution to the sixth phase of the Coupled Model Intercomparison Project. This multi-journal special issue on VolMIP aims at collecting relevant research results obtained within the VolMIP framework, and specifically concerning different aspects of the radiative and dynamical climatic response to volcanic forcing, detailed description of effects of different implementation of volcanic forcing in current climate models, aspects concerning the dynamical and chemical atmospheric response to volcanic aerosols simulated by global aerosol models, and comparison between reconstructed and simulated climate evolution after major eruptions.

Scope: The Transport Matrix Method (TMM) is a computationally efficient ``offline'' scheme for simulation of ocean biogeochemical tracers. The TMM essentially encapsulates the circulation of any ocean model, including sub-grid-scale parameterizations, as a sparse matrix, and advection-diffusion as a sequence of sparse matrix-vector products that can be efficiently carried out on modern distributed-memory computers. Simulations carried out with the TMM can potentially be orders of magnitude faster than the underlying ocean circulation model. The flexible architecture of the TMM framework also allows ``mixing and matching'' of circulations and biogeochemical models. The TMM code, various biogeochemical models, and transport matrices from different ocean models are freely available to download from https://github.com/samarkhatiwala/tmm.

The aim of this Special Issue is to bring together under one roof papers using the TMM as the underlying simulation method. This can range from manuscripts documenting various technical aspects of the TMM framework to those describing new biogeochemical models/parameterizations and their application.

The Lagrangian particle dispersion model FLEXPART was originally designed for calculating the long-range and mesoscale dispersion of air pollutants from point sources, such as after an accident in a nuclear power plant. The model has now evolved into a comprehensive tool for atmospheric transport modelling and analysis. Its application fields are extended to a range of atmospheric transport processes for both atmospheric gases and aerosols, e.g. greenhouse gases, ozone depletion substances, short-lived climate forces like black carbon, volcanic ash and gases as well as studies of the water cycle. It supports ECMWF, NCEP, mesoscale (WRF) and climate (NorESM, CAM) wind fields in both forward and backwards modes. FLEXTPART is open source (https://www.flexpart.eu).

The Joint UK Land Environment Simulator (JULES) is a process-based, community model that simulates the fluxes of carbon, water, energy and momentum between the land surface and the atmosphere. JULES-crop is a parameterisation of crops within JULES which was developed with the dual aim of being able to simulate the impact of weather and climate on crop productivity and the impact that croplands have on weather and climate. This ongoing special issue collects together papers on model development, evaluation and application of JULES-crop.

SimSphere is a one-dimensional soil–vegetation–atmosphere transfer model devoted to the study of land surface interactions of the Earth’s system. Since its early development, the model has become highly variable in its application use.

Apart from its use as an educational tool at several universities worldwide, SimSphere is used in a number of research studies related to the examination of hypothetical scenarios examining land surface processes and feedbacks. It is also used synergistically with Earth observation (EO) data to retrieve spatiotemporal estimates of energy fluxes and surface soil moisture, involving exploration studies on the development of related operational products.

This special issue hosts contributions concerned with descriptions of further upgrades of SimSphere or its exploitation in any way. It comprises articles on model developments or applications involving the model; this includes – but is not limited to – studies exploring hypothetical scenario examination, model validation, sensitivity analysis and synergies of it with EO data.

Geoscience model infrastructure provides architecture and services for building physics-based numerical models and combining multiple models into a coupled system. Examples of model infrastructure are software libraries and frameworks that provide data structures and functions for parallel data transfer and grid interpolation, define a common interface for model components, manage a model's control flow, abstract details of parallel programming, or provide supporting capabilities such as configuration management and file I/O. The special issue of GMD is dedicated to exploring all aspects of model infrastructure integration and interoperability. Model infrastructure integration is the software engineering process of adopting an infrastructure package into a numerical model's codebase, to address new scientific requirements, enable interoperability with other models, or to improve a model's performance. Model infrastructure interoperability is required whenever multiple infrastructure packages must interact in some way in a coupled system, for example, to couple model components that originate from different scientific communities into a single cohesive system. Topics of interest for this issue include but are not limited to the following: the design and implementation of geoscience model infrastructure software; experience reports on integrating infrastructure into models; modelling systems that span scientific communities; framework interoperability, metadata and semantic mediation; performance on emerging hardware platforms; configuration and build management; and tools that support infrastructure integration, code refactoring, verification, or debugging.

This special issue groups together documentation papers for successive releases of the Met Office Unified Model (UM) Global Atmosphere (GA) and JULES Global Land (GL) configurations. GA and GL are science configurations of the UM and JULES developed for use across weather prediction and climate research timescales. Each paper presents a scientific description of the latest configuration, a fuller description of incremental changes since the previous configuration and a brief summary of their performance.

NEMO (Nucleus for European Modelling of the Ocean) is a state-of-the-art modelling framework for oceanographic research, operational oceanography seasonal forecast and climate studies. The NEMO components are NEMO-OPA (the "blue" ocean, modelling the ocean dynamics and solving the primitive equations); NEMO-LIM (the "white ocean" for modelling sea-ice), and NEMO-TOP (the "green ocean" for modelling biogeochemistry). NEMO also includes grid refinement software (AGRIF) and an assimilation component (linear-tangent NEMO-TAM, Observational operators NEMO-OBS, and increment NEMO-ASM). The "blue ocean" component is fundamental to all users. NEMO can also be interfaced to a number of other components such as atmosphere models or alternative other models of sea-ice or biogeochemistry, to enable Earth system modelling.

This Special Issue aims to collect technical and scientific manuscripts dealing with evaluation of model skill and performance as well as development of NEMO components. Submitted manuscripts can cover a wide variety of topics, including process studies, new parameterizations, implementation of new model features and new NEMO configurations. The main scope is to collect relevant and state of the art manuscripts to provide the NEMO users with a single portal to search, discover and understand about the NEMO modelling framework potential and evolution and submit their contributions.

iLOVECLIM is an intermediate complexity fully coupled climate Earth system model that aims at computation and understanding of the climate system on a millennial timescale. It is a code fork from the LOVECLIM climate model version 1.2. From its forerunner, iLOVECLIM retains only the physical climate components (atmosphere–ocean–terrestrial vegetation modules). It is developed further to progressively include the components necessary for multi-millennia palaeoclimate and future climate experiments. As such, iLOVECLIM is a tool designed to enhance the integration of model simulations and (palaeo-)data, with an emphasis on the simulation of isotopic tracers throughout all components of the climate system, as indicated by the i- prefix. The present, time-unlimited special issue hosts the technical documentation of the current version of iLOVECLIM as well as model evaluation manuscripts.

The Norwegian Earth System Model (NorESM) is a global, coupled model system for the physical climate system, which can be run with various degrees of interactions with biogeochemical processes in the earth system. NorESM is developed as a nationally coordinated effort in Norway, but important parts of the model code are imported from the Community Climate System Model (CCSM) and Community Earth System Model (CESM) projects operated at NCAR on behalf of UCAR in the USA.

The NorESM publications in this special issue address a range of NorESM versions. The first set of model versions delivered results to CMIP5. NorESM1-M is run concentration-driven for greenhouse gases (GHGs) and is based on CCSM4 (released 1 April 2010), while NorESM1-ME can be run emission-driven for GHGs and is based on CESM1 (released 1 July 2010). A low-resolution version, NorESM1-L, was developed mainly for paleo-climate simulations. New versions of NorESM are underway: NorESM1.X, where X indicates updates of the NorESM1 versions, and NorESM2, which is intended to contribute to CMIP6. Further versions will follow thereafter. NorESM includes the following: its own developed code for chemistry–aerosol–cloud–radiation interactions (CAM-Oslo) and enhancements of the dynamics/physics of the atmospheric module; alternative parameterization of surface turbulent fluxes; an isopycnic coordinate ocean model originating from the Miami Isopycnic Coordinate Ocean Model (MICOM) but developed further; and the HAMburg Ocean Carbon Cycle (HAMOCC) model developed at the Max Plank Institute for Meteorology, Hamburg and adapted to the isopycnic coordinate ocean model framework.

Papers developed for full validation (CMIP-DECK) or more specific evaluation of the NorESM versions and further developments of these are welcome in this special issue. Authors intending to contribute papers to this special issue should contact the coordinators (Mats Bentsen and Michael Schulz), e.g., to ensure the consistency of version names and numbers.

CMIP5 represents the most ambitious and computer-intensive model inter-comparison project ever attempted. Integrating a new generation of Earth system models and sharing the model results with a broad community has brought with it many significant technical challenges, along with new community-wide efforts to provide the necessary software infrastructure. This special issue will focus on the software that supports the scientific enterprise for CMIP5, including: couplers and coupling frameworks for Earth system models; the Common Information Model and Controlled Vocabulary for describing models and data; The development of the Earth System Grid Federation; the development of new portals for providing data access to different end-user communities; the scholarly publishing of datasets, and studies of the software development and testing processes used for the CMIP5 models. We especially welcome papers that offer comparative studies of the software approaches taken by different groups, and lessons learnt from community efforts to create shareable software components and frameworks.

The climate record of the last millennium holds much promise for identifying the links between forcings and responses at the global, hemispheric and regional scale. The specifications for the Paleoclimate Model Intercomparison Project Phase III (PMIP3) and the Coupled Model Intercomparison Project Phase V (CMIP5) include controlled experiments for the period 850 to 1850 CE. The implementation of those experiments, the collation and development of the climate drivers over this period, and the assessment of the model responses are the focus of this special issue of Geoscientific Model Development.

The CSIRO Mk3L climate system model is a computationally-efficient coupled general circulation model, designed primarily for the study of climate variability and change on millennial timescales. The model distribution is freely available to the research community. This Special Issue allows the history and evolution of Mk3L to be comprehensively documented within a single issue of a journal. Papers are welcome which describe and evaluate components of the model, and which describe and evaluate subsequent enhancements to the modelling system. The issue is also intended for papers which document specific experiments, particularly those which contribute towards Model Intercomparison Projects.

FAMOUS is climate model based on the widely used "HadCM3" atmosphere–ocean general circulation code, a version of the UK Met Office Unified Model. Run at a lower resolution than HadCM3, its computational requirements make it suitable for large ensembles and millennial-scale climate simulations. This ongoing special issue collects technical documentation and evaluations of the model climatology as FAMOUS is developed and coupled to models of other Earth system components.

The Modular Earth Submodel System (MESSy) is a multi-institutional project providing a strategy and the software for developing Earth System Models (ESMs) with highly flexible complexity.

The strategy follows a bottom-up approach, meaning that the various processes and diagnostic tools are implemented as so-called submodels, which are technically independent of each other and strictly separated from the underlying technical model infrastructure, such as memory management, input/output, flow-control, etc.

The MESSy software provides generalized interfaces for the standardized control and interconnection (coupling) of these submodels.

The present time-unlimited Special Issue hosts scientific and technical documentation and evaluation manuscripts concerned with the Modular Earth Submodel System and the models build upon it. Moreover, it comprises manuscripts about scientific applications involving these models.