BG4.5 | Experimental Approaches, Machine Learning, and Data Analysis: Understanding the Paleo- and Future of Marine Environments in Laboratory and Nature Settings
EDI
Experimental Approaches, Machine Learning, and Data Analysis: Understanding the Paleo- and Future of Marine Environments in Laboratory and Nature Settings
Co-sponsored by JpGU
Convener: Petra Heinz | Co-conveners: Hiroshi Kitazato, Sinatrya Diko PrayudiECSECS, Christiane Schmidt, Takashi Toyofuku

To understand marine realms of the Earth and answer questions about biotic evolution and ecosystem functioning in the past, present and future, laboratory- or natural-based experimental approaches can contribute substantially. This includes experiments controlling environmental variables, using stable or radioactive isotopic biomarkers, breeding experiments, genetic analyses, biomineralization experiments of calcareous or siliceous organisms, experiments for proxy development, theoretical experiments up to modelling, machine learning and data analysis approaches or even natural laboratories (e.g., areas of biological invasion, vent and seep activities, tsunami landslides and turbidites, and many other natural situations strongly influencing the environment). Altogether, they decode faunal and ecosystem functional responses to changing connectivity patterns, habitat change or global change threats. These experimental approaches and their effective evaluation using various data analysis tools contribute greatly to a better understanding of how biotic evolution takes place in nature, how ecosystems also act as functional labs, and how Earth systems have moved and can move dynamically. They enable us to make more robust projections into the future or decipher past ecosystem trajectories with potential analogues to future change.