Vol. 71, No. 1 * Pages 1–55 * January 2026

Carbon compounds, primarily carbon dioxide, play an essential role in shaping the Earth's climate. The study reviews the natural biogeochemical cycle of carbon and the changes caused by human activity. It presents the most important sources and sinks of atmospheric carbon dioxide, as well as the associated processes. Although human emissions are relatively small relative to the gross exchange between the atmosphere and other spheres, the study clarifies that they are of decisive importance in the net mass flow that shapes the carbon dioxide content of the atmosphere.

Mankind has long been interested in atmospheric phenomena. Both in science and in education these phenomena have played a central role since ancient times. However, the development and separation of the basic scientific disciplines (physics, chemistry, biology and geology) have limited the teaching of atmospheric phenomena to the field of geography. More recently, environmental issues are at the forefront of people's interest, and the teaching of meteorology has come to a new renaissance. The study of the fascinating phenomena of the atmosphere and the explanation of the impact of daily weather on our daily lives are a great motivation for students, while requiring teachers to summarise complex knowledge and apply the laws of the different subjects. Planning lesson and creating teaching resources for high school learners, present new challenges for teacher training and methodological research. In this paper we introduce the focal points of this research and the way in which the Department of Meteorology at Eötvös Loránd University participates in this process.

Timely forecasting of agricultural yields is crucial not only for food security, agricultural economic planning, and the management of climate risks, but also for market decisions - such as timing sales or responding to international commodity-market dynamics – since in the latter case it can have a concrete, monetarily measurable short-term impact. Alongside traditional statistical estimates, machine-learning methods are gaining increasing prominence, as they are capable of uncovering the complex relationships between agricultural and meteorological data. The goal of our study was to develop machine-learning-based models for multiple crop types (winter wheat, maize, sunflower) that can provide reliable yield forecasts based on continuously updated input data throughout the growing season (5- and 8-day temperature and precipitation indicators, soil moisture, and satellite-derived vegetation indices). Our results show that the model incorporating the largest number of predictors does not necessarily deliver the best performance, and that the most informative predictor variables differ substantially between crop types. During the development, our goal was to create an operational, real-time forecasting system.

Two unusually intense storms hit Europe in September and October 2025. The first cyclone was Storm Gabrielle, which was the remnant of a tropical cyclone and brought extreme rainfall and windstorm to the Iberian Peninsula. A few days later, another cyclone – named Storm Amy – swept across Western Europe. It can also be traced back to a former hurricane, Humberto. What these two severe events had in common was their tropical origin, former hurricanes.