Seismic networks produce large volumes of data, that are hard to process through traditional methods. Here we create an automatic workflow using machine and deep learning techniques to process five years of data in the western coast of South America. This allows us to detect smaller earthquakes than before, creating a catalogue of seismicity that contains far more events than other regional catalogues.

This article describes a dataset of annual snow depth maximum across Europe, from 1961 to 2015, based on a regional reanalysis. It evaluates the performance of the dataset, against in-situ snow depth observations. This dataset is found to perform well in most environments, with challenges at high elevation and some coastal areas. Assessing the quality of this dataset is necessary in order to use it as a baseline to infer future changes of extreme snow loads under climate change.

Quantifying greenhouse gas emissions from wetlands is vital for climate mitigation. Our high-precision measurements of CO₂, CH₄, and N₂O fluxes reveal highly dynamic exchanges, especially for CO₂ and N₂O, driven by temperature and groundwater fluctuations across hourly to monthly scales. These insights advance understanding of wetland emission regulation and support improved biogeochemical models.

Understanding plant sunlight absorption is crucial for tracking global ecosystem health. We developed a 1995–2024 dataset that enhances satellite-based plant activity measurements by resolving data inconsistencies and improving resolution. Using advanced modeling, we harmonized signals from multiple satellites, cutting errors by 49 %. This offers clearer global photosynthesis trends, aiding climate research and vegetation monitoring.

We present a public dataset of weekly groundwater levels from more than 3000 wells across Germany, spanning 32 years. It combines weather data and site-specific environmental information to support forecasting groundwater changes. Three benchmark models of varying complexity show how data and modeling approaches influence predictions. This resource promotes open, reproducible research and helps guide future water management decisions.

This study integrates ground observations, satellite remote sensing, and reanalysis data, applying machine learning techniques to generate a high-resolution, daily multi-layer soil temperature dataset in China  from 2010 to 2020. The dataset accurately captures the spatiotemporal variations in soil temperature at multiple depths, offering valuable scientific insights for agricultural management and ecological research.

Small lakes and ponds provide many functions for the environment. Because of their size they are often not considered in scientific studies. We collected all the information on those lakes and ponds for Germany and combined it to a database. We described the ponds in detail, for example how deep they might by, how much water they can store or if they are connected to rivers. We found more than 260.000 lakes and ponds in Germany.

We compiled a global dataset of soil carbon release rates from lab incubations across 721 soils and various temperatures. Many experiments did not reflect real-world conditions, especially for deeper soils and realistic temperature ranges. Simulation experiments showed that both carbon properties and environmental constraints equally drive temperature sensitivity. Current climate models miss these patterns, but our findings can help improve predictions of soil carbon responses to warming.

Our study uses Google Earth Engine to create a dataset of Antarctic icebergs in the Southern Ocean (south of 55° S) from October 2018 to 2023. The dataset includes icebergs larger than 0.04 km², with details on their locations, sizes, and shapes. It shows significant changes in iceberg number and area, mainly driven by major ice shelf calving events – especially in the Weddell Sea. This resource fills key gaps in understanding iceberg impacts on the ocean and climate.