Detection and Attribution of Changes in Soil Moisture and Temperature as Affected by Plant Diversity and Climate
Team
Overview
This project aims to study climate variability impacts on plant-soil interactions and carbon cycling that occur across different temporal scales along a gradient of plant diversity. To this end, we aim to combine deep learning capabilities, causal inference theories, and domain knowledge to learn the cause-effect interaction patterns in the observed plant-soil-climate system. We use a unique, highly-resolved dataset on soil moisture and soil temperature from a long-term biodiversity experiment (the Jena Experiment).
Over a time-coarse of 18 years, we will explore if effects of climate extremes on soil abiotic conditions are buffered by plant diversity. This hypothesis can be tested with unprecedently detailed long-term data on climate, plant community properties, soil microclimatic conditions, and soil microbial biomass and respiration with likely consequences for soil carbon sequestration. This project is highly interdisciplinary by combining cutting edge machine learning approaches with unique long-term data from an iDiv experimental platform.
Publications
Investigating the Effects of Plant Diversity on Soil Thermal Diffusivity Using Physics- Informed Neural Networks.
ICLR Workshop on AI4DifferentialEquations In Science (ICLR-WS). 2024.
[bibtex] [pdf] [web]
Enhanced Stability of Grassland Soil Temperature by Plant Diversity.
Nature Geoscience. pp. 1-7. 2023.
[bibtex] [doi] [abstract]
Plant Diversity Stabilizes Soil Temperature.
bioRxiv. pp. 2023-03. 2023.
[bibtex] [pdf]