News

A large delegation from the AMC-Lahti early-career cohort — 9 PhD students and 3 post-docs — will travel to the 12th International Aerosol Conference (IAC 2026), held 30 August – 5 September 2026 in Xi'an under the theme "Breathe the same air, share a common fate".

Every member of the cohort has been accepted for an oral presentation, a remarkable result that reflects the breadth of the centre's activity in atmospheric chemistry, aerosol dynamics, inverse problems and machine-learning emulators.

• 🎤 Post-docs: Robin Wollesen de Jonge, Zihao Fu, Benjamin Foreback
• 🎤 PhD students: Valery Ashu, Zeqi Cui, Wenqing Peng, Haitong Zhang, Ammar Kheder, Jenni Köykkä, Samuel Agenorwoth, Xiang Li, Helmi Toropainen

Registration is now open for our biennial summer school "First Steps in Biosphere-Atmosphere Modelling", taking place 10–21 August 2026 at the Lahti University Campus.

During the two weeks, every participant will write their own one-dimensional atmospheric boundary-layer model from scratch, covering meteorology, gas-phase chemistry, BVOC emissions, deposition, and aerosol dynamics. Coding is done in Fortran 95; prior programming experience in any language is enough.

See the official poster below and head to the Education page for the registration details.

A delegation from AMC-Lahti visited Prof. Jia Chen and her group at the Technical University of Munich (TUM) for a two-day research stay, from 5 to 6 May 2026, marking the start of a new international collaboration.

The visiting team included Prof. Michael Boy, Assoc. Prof. Zhi-Song Liu, and doctoral researchers Wenqing Peng, Helmi Toropainen, and Ammar Kheder.

The discussions laid the foundation for a new joint collaboration focused on two main directions:

• AI-based air quality, climate downscaling and optimization
• ARCA box model — integrating and extending the framework for atmospheric chemistry and aerosol process modelling

AMC-Lahti members participated in the NOSA Symposium 2026: Impacts of Aerosol Science & Technology, held at Lund University.

This year, four members from AMC-Lahti presented their work:

• 🎤 Petri Clusius — Modelling the impact of anthropogenic aerosols on CCN concentrations
• 🎤 Haitong Zhang — Quantifying Residential Wood Combustion Emissions in Lahti
• 🎤 Robin Wollesen de Jonge — DMS-driven new particle formation in the remote marine boundary layer
• 🪧 Jenni Köykkä — Investigating Regional and Long-Range Transport of Aerosols to Greenland

Our PhD doctoral researcher Ammar Kheder visited his former program @but_sd_niort to share his academic and professional journey with current students.

Ammar is currently a second-year PhD student at AMC-Lahti, where he specializes in artificial intelligence applied to atmospheric chemistry 🌍🤖. His path perfectly illustrates how data science can lead to impactful research addressing major environmental challenges.

AMC-Lahti members participated in Finnish Inverse Days 2025 in Helsinki, a community event bringing together researchers and practitioners working on inverse problems.

This year, two doctoral researchers presented their work:

• 🎤 Jenni Köykkä — Inverse Perspectives on the SOSAA–FP Modelling Framework
• 🎤 Samuel Agenorwoth — Uncertainty and Computation in Acoustic Imaging of Pipes

Di Chen has successfully defended his PhD thesis titled "From atmospheric oxidation capacity at boreal forest to earth system tipping points under a changing climate" in Auditorium A129, Chemicum.

The opponent was Docent Tero Mielonen (Finnish Meteorological Institute), and the custos was Professor Michael Boy (University of Helsinki).

In this work, the one-dimensional model SOSAA was further developed to be suitable for long-term runs. Methods from complexity science were introduced to investigate possible connections between the Amazon Rainforest and the Tibetan Plateau under global warming.

Our PhD student Wenqing Peng presented his research paper "SPIN-ODE: Stiff Physics-Informed Neural ODE for Chemical Reaction Rate Estimation" at the 28th European Conference on Artificial Intelligence.

SPIN-ODE introduces a three-stage optimisation process: a black-box neural ODE fits concentration trajectories, a Chemical Reaction Neural Network (CRNN) is pre-trained, and rate coefficients are fine-tuned by integrating with the pre-trained CRNN.

A three-day retreat bringing together members of the MSM and AMC groups in Hyytiälä, combining scientific presentations, thematic discussions, and social activities.

Day 1: welcome session, introduction, and individual presentations. Dinner and sauna in the evening.

Day 2: presentations, visit to the SMEAR II station, breakout sessions on AI applications (led by Zhi-Song) and the FLEXPART/SOSAA modeling system (led by Petri). The day wrapped up with a lake-side barbecue and sauna.

Day 3: strategic discussions on merging groups and websites, and on increasing visibility scientifically and publicly.

Every year, our group takes part in EAC, which brings together researchers from across the world to discuss the latest advances in aerosol science, atmospheric chemistry, climate, air quality, and their societal impacts.

This year, six of our PhD students presented their work:

• 🖼️ Wenqing Peng — Physics-informed neural ODEs for chemical reaction rate estimation
• 🖼️ Zeqi Cui — Machine learning to optimize OH-oxidation of naphthalene
• 🖼️ Valery Ashu — Uncertainty quantification of reaction coefficients with MCMC
• 🖼️ Ammar Kheder — Deep spatio-temporal neural network for air quality reanalysis
• 🖼️ Haitong Zhang — Residential wood combustion emissions in Lahti, Finland
• 🎤 Benjamin Foreback — FLEXPART atmospheric back-trajectories for pollution transport to Beijing

The University of Helsinki (UH), INAR, Lappeenranta-Lahti University of Technology (LUT), and the Lahti University Campus held the summer school "Application of AI/ML techniques in Atmospheric Science".

The course provided introductory lectures on atmospheric chemistry, aerosol dynamics, Earth System models, and numerical weather prediction, alongside computational methods for data science and machine learning. Mixed groups trained end-to-end neural networks (LSTM, RNN, Transformer) on previously created atmospheric datasets using PyTorch.

Doctoral researcher Samuel Agenorwoth and Associate Professor Emilia Blåsten attended AIP 2025 in Rio de Janeiro, hosted by FGV EMAp.

The conference was a valuable opportunity to follow recent developments in inverse problems, exchange ideas, and connect with researchers across the international community.

PhD student Ammar Kheder presented his first research paper "Deep Spatio-Temporal Neural Network for Air Quality Reanalysis" at the 23rd Scandinavian Conference on Image Analysis.

AQ-Net is a spatiotemporal reanalysis model for both observed and unobserved stations. It uses LSTM and multi-head attention for temporal regression, cyclic encoding for continuous time representation, and a neural kNN to fill spatial gaps. Experiments use PM data from 2013–2017 in northern China.

Benjamin Foreback has successfully defended his PhD thesis titled "Extreme air pollution events in Beijing: Examining their origins and dynamics" in Aalto Auditorium at Lahti University Campus.

The opponent was Professor John Lin (University of Utah), and the custos was Professor Michael Boy (University of Helsinki / LUT University).

The thesis examines two different types of extreme air pollution events: Chinese New Year fireworks effects on Beijing air quality (2013–2019), and a severe wintertime haze episode in November 2018.

Petri Clusius has successfully defended his PhD thesis titled "Process modelling of emissions, chemistry and SOA formation in real environments" in Physicum, Auditorio E204.

The opponent was Senior Scientist Ulas Im from Århus University, and the Custos was Professor Michael Boy.

The thesis investigates how aerosols influence weather and global climate through cloud interactions, focusing on cloud condensation nuclei (CCN) and their effects on cloud lifetime, optical properties, and precipitation.

Four of our PhD students presented their research at the LUT Doctoral School (DS) Science Conference 2025 in Lappeenranta, Finland.

• Jenni Köykkä — Particle-phase module for organic chemistry using Bayesian analysis and generative AI
• Samuel Agenorwoth — Uncertainty quantification in pipe imaging
• Valery Ashu — Optimizing autoxidation chemistry codes with neural networks
• Ammar Kheder — Physics-guided deep learning for air quality prediction