Understanding the future climate is crucial for informed policy decisions on climate change prevention and mitigation. Earth system models play an important role in predicting future climate, requiring accurate representation of complex sub-processes that span multiple time scales and spatial scales. One such process that links seasonal and interannual climate variability to cyclical biological events is tree phenology in deciduous broadleaf forests. Phenological dates, such as the start and end of the growing season, are critical for understanding the exchange of carbon and water between the biosphere and the atmosphere. Mechanistic prediction of these dates is challenging. Hybrid modelling, which integrates data-driven approaches into complex models, offers a solution. In this work, as a first step towards this goal, train a deep neural network to predict a phenological index from meteorological time series. We find that this approach outperforms traditional process-based models. This highlights the potential of data-driven methods to improve climate predictions. We also analyze which variables and aspects of the time series influence the predicted onset of the season, in order to gain a better understanding of the advantages and limitations of our model.

Climate change is global, yet its concrete impacts can strongly vary between different locations in the same region. Seasonal weather forecasts currently operate at the mesoscale (> 1 km). For more targeted mitigation and adaptation, modelling impacts to < 100 m is needed. Yet, the relationship between driving variables and Earth's surface at such local scales remains unresolved by current physical models. Large Earth observation datasets now enable us to create machine learning models capable of translating coarse weather information into high-resolution Earth surface forecasts. Here, we define high-resolution Earth surface forecasting as video prediction of satellite imagery conditional on mesoscale weather forecasts. Video prediction has been tackled with deep learning models. Developing such models requires analysis-ready datasets. We introduce EarthNet2021, a new, curated dataset containing target spatio-temporal Sentinel 2 satellite imagery at 20 m resolution, matched with high-resolution topography and mesoscale (1.28 km) weather variables. With over 32000 samples it is suitable for training deep neural networks. Comparing multiple Earth surface forecasts is not trivial. Hence, we define the EarthNetScore, a novel ranking criterion for models forecasting Earth surface reflectance. For model intercomparison we frame EarthNet2021 as a challenge with four tracks based on different test sets. These allow evaluation of model validity and robustness as well as model applicability to extreme events and the complete annual vegetation cycle. In addition to forecasting directly observable weather impacts through satellite-derived vegetation indices, capable Earth surface models will enable downstream applications such as crop yield prediction, forest health assessments, coastline management, or biodiversity monitoring. Find data, code, and how to participate at www.earthnet.tech .

In the past decades mainly static attributes have been investigated using machine learning approaches, such as the distribution of soil properties from the local to the global scale. For some time now, it has been possible to tackle more dynamic processes by using more sophisticated deep learning techniques. This allows for example to quantify the global photosynthesis on land with simultaneous consideration of seasonal and short term variations. "From a plethora of sensors, a deluge of Earth system data has become available, but so far we've been lagging behind in analysis and interpretation," explains Markus Reichstein, managing director of the Max Planck Institute for Biogeochemistry in Jena, directory board member of the Michael-Stifel-Center Jena (MSCJ) and first author of the publication. "This is where deep learning techniques become a promising tool, beyond the classical machine learning applications such as image recognition, natural language processing or AlphaGo," adds co-author Joachim Denzler from the Computer Vision Group of the Friedrich Schiller University Jena (FSU) and member of MSCJ.

A study by German scientists from Jena and Hamburg, published today in the journal Nature, shows that artificial intelligence (AI) can substantially improve our understanding of the climate and the Earth system. Especially the potential of deep learning has only partially been exhausted so far. In particular, complex dynamic processes such as hurricanes, fire propagation, and vegetation dynamics can be better described with the help of AI. As a result, climate and Earth system models will be improved, with new models combining artificial intelligence and physical modeling. In the past decades mainly static attributes have been investigated using machine learning approaches, such as the distribution of soil properties from the local to the global scale.