​

The Atmospheric Composition group

Earth Sciences Discovery Days

27/10/2025

Carlos Perez Garcia-Pando (BSC and ICREA)

Earth Sciences Department

• ~200 people
• Funding from EC, Copernicus, private sector, ESA, Spanish and regional governments
• Four ICREA, close link to local universities

Environmental modelling and forecasting using process-based and artificial intelligence models, with a particular focus on weather, climate and air quality. This includes transferring solutions to support the main societal environmental challenges through data applications

Computational Earth Sciences

Climate Variability and Change

Atmospheric Composition

Earth System Services

Global

Health

Resilience

Atmospheric Composition Group

• Main goal: to understand, constrain and predict the spatiotemporal variations of atmospheric pollutants across scales along with their effects upon air quality, health, weather and climate.

​

• Support to service activities:
• WMO SDS-WAS Barcelona Dust Regional Center
• Copernicus Regional Production Service
• CALIOPE air quality forecasts for Spain
• ICAP Global aerosol forecasting

​

• Staff: 50 people composed of senior researchers, postdocs, PhD and MSc students, and support engineers

​

• Competitive projects: 9 contracts, 7 national, 8 Europe, 2 others

Strong collaboration with other groups of the Earth Sciences Dept.

Air quality

Atmospheric Composition Group

Understand, constrain and predict the spatiotemporal variations of atmospheric pollutants across scales along with their effects upon air quality, health, weather and climate.

Strong collaboration with other groups of the Earth Sciences Dept.

Atmospheric Composition Group (October 2025 48 staff)

​

5

Group leaders

Carlos Pérez García-Pando

Oriol Jorba

AC-Communications

Karinna Matozinhos (1)

AC-EMIS

Marc Guevara (8)

AC-CHEM

Hervé Petetin (12)

AC-CLIM

María Gonçalves (12)

AC-DAFA

Jerónimo Escribano

Emanuele Emili (13)

WG Obs & Eval

Dene Bowdalo

WG MODELS

Oriol Jorba

WG AI

Hervé Petetin

WG OPER

Miriam Olid

Emanuele Emili

Ruben Alba Camille Alessio

Roger Isidre Patricia Georgia

Montse Marios Emily Luka Fernando

Vincenzo Carla Alba Melani

Dene Elina Calum Andres

Guillaume Raphael Rafaela Carlotta

Paula Kevin James

Oliver Paula Òscar Ivan

Francesco Francesco

Elena Yara Michael

Jayoung Miguel Tito

EC-Earth3-Iron

Atmospheric iron cycle in EC-Earth

Model and Tool developments

Model Output Statistics

Including machine learning

Guevara et al. (2019, GMD)

Guevara et al. (2020, GMD)

Guevara et al. (2021, ESSD)

Badia et al. (2017, GMD)

Peng et al. (2019, QJMRS)

Klose et al. (2021, GMD)

Di Tomaso et al. (2017, GMD)

Escribano et al. (2022, ACP)

Di Tomaso et al. (2022, ESSD)

Benavides et al. (2019, GMD)

Benavides et al. (2021, ERL)

Rodriguez-Rey et al. (2021, TR-RES)

Rodriguez-Rey et al. (2021, STOTEN)

GHOST

Harmonised treatment of

observations

Myriokefalitakis et al. (2021, GMD)

Bergas-Masso et al. (2023, Earth’s Future)

Petetin et al. (2020, ACP)

Petetin et al. (2022, ACP)

Bowdalo et al. (2023, subm. ESSD)

Bowdalo et al. (2021, ERL)

In collaboration with CES, ESS and CVC

Emissions

Quantify and perform the spatial and temporal distribution of primary anthropogenic emissions to support modeling efforts, and as input to policy support, at urban, national European and global scales

Global and EU emission products to support Copernicus CAMS and CO2MVS

​

Coordination of WG in international forums

National air Emissions Systems to suPport air quality modellIng and GHG monitoRing Efforts

Bottom-up versus satellite-based emissions

Guevara et al. (2023 and 2025, ACP); Guevara et al. (2022 and 2024, ESSD); Soulie et al. (2024, ESSD); Urraca et al., (2024, ESSD)

Oliveira et al. (2023, STOTEN)

Oliveira et al. (2024, ACP)

Oliveira et al. (2025, Env. Poll.)

Rodriguez-Rey et al., (2022, STOTEN)

Ding et al. (2024, ACP)

NMVOC emissions – improving speciation

Assess the impact of revised NMVOC speciation on emissions of individual species and on the modelling of BTX concentrations across Europe

1. Compiled and compared speciation profiles (Oliveira et al., 2023)

2. NMVOC speciated emissions comparison: CAMS default vs. OLIV23

2. NMVOC speciation sensitivity analysis: Impact on modelled BTX concentrations

ESA World emissions: Bottom-up versus top-dow estimates

Spatiotemporal intercomparison between official NH3 bottom-up emissions compiled by the Ministry against IASI-based estimates

Global point source catalogue to support CO2MVS Copernicus service

Plant-level annual NO_x emissions (kt/year)

TROPOMI-based annual NOx emissions for a power plant using the divergence method. Integration radius defined by checking the curvature of the radial profiles and select the point that maximally contains the source without including the next one

USA

South Africa

Lebanon

Bangladesh

Enhancing the country’s strategic capabilities to support better decision-making for air pollution control and contribute to international climate change mitigation efforts.

Modelling of atmospheric pollutant emissions in Spain

National monitoring system for GHG emissions in Spain

high-Resolution air Emissions Systems to suPport modellIng and monitoring Efforts

The project is part of the Recovery, Transformation and Resilience Plan (Plan de Recuperación, Transformación y Resiliencia, PRTyR) funded by the European Commission – NextGenerationEU.

High-Elective Resolution Modelling Emission System, version Delta (HERMES_Δ)

Pre-processing

​

​

​

​

​

​

​

• NUTS2 level
• Detailed sectoral level (~225 emission sources).
• SNAP/NFR (AP, GHG).

Distribution�

• Horizontal: ~140 spatial proxies.
• Temporal: ~800 monthly, weekly and hourly profiles.
• Vertical: stack height.
• Speciation: splitting emissions (NMVOC, NO_x and PM_2.5) into chemical mechanism species.

Outputs�

Ready-to-use emission maps for modelling applications (1km, hourly), and for supporting public administration.

​

Total annual CO₂ emissions (Gg), 2023

Four pillars integrating activity-based and observation-based high-resolution CO₂ and CH₄ emissions.

Atmospheric chemistry and air quality

Understanding tropospheric Ozone reactive chemistry and the role of volatile organic compounds for policy and health assessment.

Advancing current representation of chemistry in models and enabling AI solutions.

Ozone sensitivity

Health impact

AI and air quality modelling

Brown carbon and nitrate aerosol chemistry

​

GPU

Scientific basis for an O3 plan for Spain

Overall scientific objective, specific research lines and tools/data

To improve our understanding and modeling capabilities of the physical and chemical processes driving air quality and their impact on human health and ecosystems

Ozone formation and precursors

Secondary inorganic aerosols and precursors

Secondary organic aerosols and multiphase chemistry

Heat waves, droughts, vegetation fires and climate change

Machine learning

MONARCH chemistry-weather model

CAMP for multiphase chemistry

GECKO-A for explicit chemistry

Surface in-situ observations

Satellite observations

+

Auto-ML for ML developments

Aerosol studies, from 0D to global scale

Characterizing Brown carbon absorption �across Europe (and now also at global scale…)

Exploring the differences of sensitivity to SOA formation among state-of-the-art chemical schemes

Unraveling nitrate heterogeneous formation on desert dust

Sousse-Villa et al., Atmos Chem Phys (2024); Navarro-Barboza et al., EnvInt (2023); Navarro-Barboza et al., Atmos Chem Phys (2024)

Leveraging machine learning capabilities for…

… emulating the entire MONARCH to support air quality planning

… downscaling MONARCH to increase spatial resolution�

… emulating the gas-phase chemistry in order to allow more detailed chemical schemes and/or faster air quality models

​

​

… improving our estimation of NOx emissions based on satellite observations (AC-EMIS)

Additionally: development of auto-ML for facilitating our ML developments on BSC’s HPC

… building surface air quality reanalysis for health studies

… forecasting dust AOD based on satellite observations (AC-DAFA)

Emission scenario analysis to support the design of a national ozone mitigation plan

Additionally, investigating the climate penalty in Catalonia and Europe

​

​

​

Ozone regime evaluation against TROPOMI satellite

​

​

​

Petetin et al., Atmos Chem Phys (2023); Petetin et al., Sci Tot Env (2023); Elshorbany et al., Atmos Chem Phys (2024, under press); Fadnavis et al., Atmos Chem Phys (2024, under review)

Impact of national plans (PNIECC and PNCAA) on surface ozone in Spain

Scientific basis for an O3 plan for Spain

Spanish- and European-scale source apportionment of �ozone pollution and associated health and policy assessment

Contribution from boundary conditions (left), maritime (middle) and national (right) contribution to ozone in Spain

​

​

​

Achebak et al., Nature Medicine (2024); Garatachea et al., Comm Earth Environ (2024)

Scientific basis for an O3 plan for Spain

O3 pollution (left) and associated mortality across Europe (right)

Import/export of ozone pollution across European countries

​

Trends in population exposure to compound extreme-risk temperature and air pollution

​

Chen et al., Nature Communications (2024)

Chen et al., The Lancet Planetary Health (2025

AC Clim: Dust, aerosols and climate

Aim: To improve our understanding of the interactions between short-term lived atmospheric species, with a focus on aerosols and mineral dust, and different components of the Earth System.

​

Models and observations.

Dust, aerosols and climate

Field campaigns, in-lab analyses: soil and dust mineralogy, emission.

Dust modelling at different spatio-temporal scales.

Model development.

Dust and bio-aerosols - mixed-phase clouds

​

Dust mineralogy, shape and size - radiation

​

Future fires - ocean biogeochemistry

​

… these impacts are modulated by mineralogy, size and shape.

Dust (and other aerosols)

23

Image credits: NASA, NOAA, Krueger et al. (2004)

Dust campaigns

Dust size distribution, mineralogy, emission processes.

Geological characterization of the dust sources: soils, morphology.

METHODS

RESULTS

CONCLUSIONS

PI: C. Pérez García-Pando

Frontiers in dust minerAloGical coMposition and its Effects upoN climaTe

Morocco

“El Bour” (M’hamid)

(Aug/Sept. 2019)

Iceland

Vatnajökull National Park

(Aug. 2021)

González-Romero et al (2024a,b), González-Flórez et al. (2023; 2024 in prep.), Dupont et al. (2024), Panta et al. (2023).

​

Mineralogy maps from NASA EMIT

VSWIR Spectra of Dust Source Minerals

Dust Minerals have distinct spectral signatures

The EMIT instrument is measuring from the ISS since July 14, 2022.

Target mask for EMIT retrievals covering arid land regions

Jet Propulsion Laboratory

California Institute

of Technology

Level 3 products – map of 10 (+2) minerals to be used within ESMs

(Green et al. 2020)

EMIT data products

​

Image courtesy of Phil Brodrick (NASA-JPL)

25

Dust radiation interaction

Obiso et al. (2023, 2024; in prep), Song et al. (2024), Li et al. (2024), Adebidyi et al. (2023), Ilic et al., (2025, in prep.), Kok et al., 2025, in review),

Dust mineralogy, mixing state, shape and size - radiation

​

26

Dust and bio-aerosol - clouds

Dust

MPOA

PBPAs

Ice crystal number concentration

(mixed-phase clouds, column burden <10km)

Meyers

Harrison + Wilson + RaFSIPv2

Chatziparaschos et al. (2023; 2025.); Costa-Surós et al. (2025); Thomas et al. (2024), Riipinen et al. (2024, in prep)

New INP sources in models + effect on clouds and climate

​

27

LWC

IWC

Results for 12-year nudged simulations

28

ICNC_MPC

LWP

IWP

Meyers

Aerosol-sensitive param. impact: tendency to cool the system respect to Meyers

CLT

Implementation of aerosol-sensitive + RaFSIPv1 in ECE3 FORCeS version

+ tuning

Dust, fires and ocean biogeochemistry

Bergas et al. (2025.); Bergas et al. (2022), Myriokefaltakis et al. (2021)

Bergas et al. (2025.) Nature Climate Change

29

Impact of NTCFs on the climate system

Arctic Cooling

Labrador Sea Variability

Equatorial Precipitation

• Cooler temperatures.

Especially in the Arctic.

• More arctic sea ice.

Especially in the Barents Sea and Okhotsk Sea.

The Labrador Sea is an area of deep water formation. We see:

• Increased variability.
• Increased convection.

Greater vertical water mixing.

• Southward displacement of the Intertropical Convergence Zone (ITCZ).
• Reduced tropical rainfall.

Collaboration with CVC

Santos Espeso et al. (2025, in review.)

Dust trends preindustrial to present day

Kok et al., 2023

CMIP6 models dust emission trend

AEROCOM DURF experiments run with EC-Earth3

Elhacham et al. (2024, in prep.)

31

Data Assimilation, Forecasting and Applications Team showcases

Data assimilation, forecasting and applications

Research backbone of several international and national service initiatives delivering forecasting and reanalysis products. Major initiative harmonising observational data sets across multiple global/regional networks.

​

​

Support to service activities

Novel AC datasets

Data assimilation research and application

Dust emission inversions with satellite and data assimilation

Global aerosol forecasts

Satellite AOD, NO₂ and short-wave radiance aerosol assimilation

12yr dust reanalysis Public dataset

DAFA

Data Assimilation Forecasting and Applications

DATA

FORECASTING

APPLICATIONS

ASSIMILATION

Numerical models

• Satellite retrievals
• In-situ observations
• Data curation and standardization

​

​

• Global/regional daily forecasts of atmospheric composition
• Models evaluation
• Operational workflows

​

​

• Air quality, dust, smoke
• Reanalyses
• Evaluation of new satellites products
• Sources inversion
• Trainings

​

​

• Inverse methods
• Ensembles techniques
• Observation operators

​

​

Building and improving the bridges between atmospheric composition modeling, observations and end-users.

GHOST: Globally Harmonised Observations in Space & Time

​

N° Processed Measurements: ~7.3 billion

N° Components: 227

N° Networks: 38

Temporal range: 1970-2023

​

​

Data assimilation

​

Data Assimilation (DA) combines observations, simulations and their uncertainties to produce an optimal representation of the atmospheric state.

DA for concentrations

prior posterior/prior

F. Bouttier and P. Courtier, ECMWF notes, 1999

Barcelona Dust Regional Center

Dust daily forecasts ( https://dust.aemet.es ):

​

- 72 hours deterministic forecasts from MONARCH and 15 other models

- Multi-model ensemble average and probabilistic forecasts

- Dust AOD, total column load, surface concentration, dry and wet deposition, surface extinction

In collaboration with CES and ESS

Copernicus Atmosphere Monitoring Service

• Daily forecasts and analyses of air-quality over Europe (96 hours)
• Gases and aerosols
• MONARCH contributes to a multi-model ensemble of 11 EU models
• Assimilation of surface air-quality measurements plus yearly reanalyses

Observations

Ensemble

MONARCH

In collaboration with CES

International Cooperative for Aerosols Prediction

• Daily global forecasts (120 hours)
• Gases and aerosols
• MONARCH contributes to a multi-model ensemble of 9 global models
• Total and speciated aerosols

2021 yearly averages

Aerosol optical depth

SATELLITE

MONARCH

In collaboration with CES

Applications: inversion of NOx sources

Satellite based (TROPOMI) estimation of industrial NOx plumes over Canary Islands using lightweight methods*

​

* Kuhlmann, G., et al.: The ddeq Python library for point source quantification from remote sensing images (version 1.0), Geosci. Model Dev., 17, 4773–4789, 2024

THANK YOU!!