We develop innovative tools, algorithms and methodologies at the intersection of climate and social sciences — grounded in the core of statistics and econometrics, and powered by artificial intelligence — to advance knowledge and support partners and countries across the world.
About Gauss Research
We exist to bridge rigorous academic research and the urgent, real-world demands of climate policy, sustainable development, and environmental governance — working alongside partners and institutions that share our commitment to science-driven change.
We develop innovative tools and algorithms grounded in the core sciences — and strengthened by statistics, econometrics, and artificial intelligence. Our work is not theoretical: it is deployed in real policy contexts, from EU-level reporting frameworks to national climate transparency systems in developing nations.
Our work spans climate change, energy, environment, environmental economics, climate finance, and policy analysis — partnering with institutions, governments, international organisations, and research networks across the world. We operate at every scale, from local initiatives to global frameworks, and across all levels of public administration.
Our tools are peer-reviewed, internationally published, and operationally validated — recognised by organisations such as the UNFCCC Secretariat and the IPCC.
Chemistry, physics, and engineering principles underpin our models, ensuring scientific rigour in all quantitative approaches.
Robust statistical methodologies for causal inference, uncertainty quantification, and evidence-based policy evaluation.
We develop and apply AI and machine learning across scientific domains — from climate projection to NLP-powered literature research, automated processing pipelines, and AI-assisted policy analysis.
Partnerships and research spanning European institutions, national governments, and international organisations across more than 70 countries.
Areas of expertise
Six interconnected knowledge domains, all contributing to a unified mission: generating knowledge that enables better decisions on climate and sustainability.
Simulation models, long-term projections, and complex systems integration — connecting physical, socioeconomic and environmental variables to inform policy and planning.
Advanced statistical frameworks for causal inference, quasi-experimental methods, uncertainty analysis, and rigorous policy impact evaluation.
Cost-benefit analysis of environmental and climate policies, carbon pricing, externality valuation, energy market modelling, and climate finance assessment.
Machine learning, deep learning, and NLP applied across scientific domains — from climate projections and automated data processing to AI-assisted literature review and policy analysis support.
Development of bespoke computational tools and software tailored to the specific needs of policy assessment, scientific reporting, and evidence-based decision-making.
Geospatial computation, remote sensing analysis, satellite data processing, and interactive dashboards that turn complex scientific outputs into accessible, actionable insights.
Modelling suite
We don't adapt existing software to our problems. We build the tools our research demands — validated, peer-reviewed, and deployable in real-world policy contexts.
Mitigation-Inventory Tool for Integrated Climate Action
The first modelling framework ensuring full consistency between all GHG reporting elements under the Paris Agreement — inventories, projections, mitigation policies (PAMs), and NDC tracking. Developed and operationally validated with acknowledgement from the UNFCCC GHG Support Unit.
At its core: ANNALIST, a proprietary hybrid AI algorithm combining LASSO regression, SARIMAX, and Random Forest with automated hyperparameter optimisation — outperforming classical methods across 14 countries.
A quantitative modelling framework linking physical climate projections to macroeconomic outcomes — GDP, sectoral output, employment, and welfare — under different warming and policy scenarios.
Combines climate science with econometric and input-output modelling to produce country-level economic damage functions that can inform adaptation and mitigation cost-benefit analysis.
Local land use from satellite Observation and Computational analysis
A remote sensing and machine learning pipeline extracting high-resolution local land use and land cover data from satellite imagery — providing ground-truth inputs for GHG inventories, REDD+ accounting, and environmental assessments in data-sparse contexts.
Addresses a critical bottleneck in climate reporting for developing nations: the absence of reliable, granular land use data at sub-national scale.
LTBM — Pathway modelling from net-zero targets
A quantitative backcasting framework that works backward from defined long-term mitigation targets — such as net-zero by 2050 — to derive consistent, sector-level policy pathways and milestone trajectories.
Grounded in the Tajikistan net-zero study, which demonstrated the need for target-led modelling where classical projection approaches cannot adequately capture structural transformation dynamics.
Projects
Active projects applying our methodologies in real policy contexts — from European institutional reporting to developing-country climate transparency.
Supporting the EEA in quality assurance and data flow management for energy and GHG reporting within the Energy Community framework — applying proprietary QA/QC methodologies to ensure compliance with European and international reporting standards.
The project directly leverages our core competencies in GHG data quality, statistical validation, and transparent environmental reporting.
Full technical support to the Guyana Forestry Commission for the development of Guyana's BTR2 under the UNFCCC Enhanced Transparency Framework, including the REDD+ Technical Annex — the critical component for forest carbon accounting.
A direct application of MITICA: enabling a developing country to meet its Paris Agreement reporting obligations with consistent, inventory-linked methodology.
Research insights
Thought pieces from our researchers on the science, methods, and policy implications behind our work.
Most countries submit NDCs with projections methodologically disconnected from their own GHG inventories. We explain why this happens — and what it takes to solve it.
Read more →A candid assessment of which ML methods perform best on small, noisy climate datasets — and where classical statistics still win.
Read more →Why self-reported land use data falls short — and how satellite-derived classification transforms what developing nations can actually measure.
Read more →IAMs systematically underestimate climate damages. Here is why — and what a more rigorous damage function approach looks like.
Read more →A structural analysis of why most NDCs cannot be credibly tracked — and the conditions that would make them meaningful accountability instruments.
Read more →A novel five-step backcasting approach to assess alternative mitigation scenarios for carbon neutrality by 2050 — providing a starting point for LT-LEDS development under the Paris Agreement. The methodological foundation for our LTBM model.
Read paper → SpringerApplying ARDL methodology to test the Environmental Kuznets Curve by economic sector across all 27 EU countries — using audited UNFCCC inventory data from 1990 to 2018.
Read paper → SpringerThe team
A team combining academic expertise and hands-on experience in international climate policy, quantitative modelling, and applied research.
Over 15 years of expertise in climate change, energy, and environmental economics. IPCC author for the forthcoming 2027 Methodology Report on Inventories for Short-lived Climate Forcers. UNFCCC Lead Reviewer and Task Manager for GHG projections at ETC/CM (EEA). Principal developer of MITICA. Has directed or contributed to over 100 international projects across more than 70 countries.
Data Scientist with 4+ years in R&D applying machine and deep learning to environmental challenges. Specialised in end-to-end data pipelines for heterogeneous datasets and predictive models for risk assessment and early-warning systems. Awarded a HIDA Helmholtz Visiting Researcher Grant at GFZ German Research Centre for Geosciences, developing ML methods based on hyperspectral imagery for environmental monitoring.
Senior Data Scientist and climate-risk modeller with 10+ years applying advanced statistical, econometric, and ML techniques to climate impact assessment. Co-developer of MITICA and ANNALIST. Expertise in climate variable modelling, early-warning analytics, and Loss & Damage risk-layering and attribution analytics.
Environmental engineer with a Master in Natural Risk Planning and Management and a Master in Renewable Energies, Smart Grids and Electric Mobility. Expertise in climate change, energy transition, and renewable technologies. Author of scientific publications applying AI to climate topics. Expert in water resources, numerical modelling, and GIS.
Innovation profile
Our innovation is not incremental. We develop original methodologies, proprietary software, and novel scientific approaches where no market equivalent exists.
MITICA and ANNALIST are internally developed tools. MITICA is peer-reviewed and published in an international indexed journal (MDPI Sustainability, 2024), constituting demonstrable intellectual property and scientific know-how.
MITICA is the first integrated framework ensuring consistency across all GHG reporting elements of the Paris Agreement — a scientifically novel approach with no prior equivalent, formally documented in the peer-reviewed publication.
The team combines an IPCC author, a UNFCCC Lead Reviewer, a PhD in theoretical physics with 10+ years in climate-risk modelling, and a PhD in microbiology with expertise in ML for environmental monitoring. Research published in indexed international journals and operationally acknowledged by the UNFCCC Secretariat.
Two active projects with high-profile international public sector partners: the European Environment Agency (EU) and the Guyana Forestry Commission (national government), demonstrating real-world applicability and scalability.
MITICA is universally applicable across countries and sectors, with decreasing marginal deployment costs. Explicitly designed for scaling to developing nations — a large, growing area driven by Paris Agreement reporting obligations.
Get in touch
We work alongside institutions, international organisations, research networks, and governments that share our commitment to rigorous, science-driven approaches to climate policy, environmental reporting, and sustainable development.
Whether you are developing a BTR, designing an NDC, building a GHG projection system, or evaluating a climate policy — we'd like to explore how we can advance the science together.