US4

In times of rapid global change, the ability to faithfully predict the development of vegetation on larger scales is of key relevance to society. However, ecosystem models that incorporate enough process understanding for being applicable under future and non-analog conditions are often restricted to finer spatial scales due to data and computational constraints. Recent breakthroughs in machine learning, particularly in the field of deep learning, allow bridging this scale mismatch by providing new means for analyzing data, e.g., in remote sensing, but also new modelling approaches. We here present a novel approach for Scaling Vegetation Dynamics (SVD) which uses a deep neural network for predicting large-scale vegetation development. In a first step, the network learns its representation of vegetation dynamics as a function of current vegetation state and environmental drivers from process-based models and empirical data. The trained model is then used within of a dynamic simulation on large spatial scales. In this contribution we introduce the conceptual approach of SVD and show results for example applications in Europe and the US. More broadly we discuss aspects of applying deep learning in the context of ecological modeling.

How to cite: Rammer, W. and Seidl, R.: New deep-learning based approaches for forest modeling beyond landscape scale, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-16528, https://doi.org/10.5194/egusphere-egu21-16528, 2021.

Addressing the impact of 3D vegetation structure on shortwave radiation transfer in Earth System Models (ESMs) is important for accurate weather forecasting, carbon budget estimates, and climate predictions. While leaf-level photosynthesis is well characterized and understood, estimates of global level carbon assimilation in the literature range from 110 to 175 PgC.yr-1. I will explore how neglecting canopy structure leads to significant uncertainties in shortwave radiation partitioning, as well as second order derived canopy properties, such as leaf area index (LAI). I will also cover how modeled carbon assimilation of the terrestrial biosphere is impacted when a satellite derived clumping index is incorporated into the UKESM. Finally, I will touch on how the clumping index might be integrated into hyperspectral ESMs to explore the theoretical relationship between canopy structure and photosynthesis.

How to cite: Braghiere, R.: Better representing vegetation canopy structure in Earth System Models, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-16531, https://doi.org/10.5194/egusphere-egu21-16531, 2021.

Ecosystems are facing unprecedented pressures as a result of human activities. At the same time, ecology as a discipline is increasingly demanding more mechanistic understanding of what causes observed ecological patterns, in part for the development of the science but also to help mitigate impacts. Here, I will present the Madingley Model (www.madingleymodel.org), a General Ecosystem Model that aims to provide a mechanistic understanding of how ecosystems, on land and in the seas, are structured and how they function, and for how anthropogenic changes might alter that structure and function. I will discuss the model’s current capabilities, how it is being used, and highlight some necessary and exciting future directions for development.

How to cite: Harfoot, M., Tittensor, D., Hoeks, S., Krause, J., Arneth, A., Doughty, C., and Abraham, A.: General Ecosystem Models, moving towards modelling responses and effects of whole ecosystems, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-16569, https://doi.org/10.5194/egusphere-egu21-16569, 2021.

Over the last years, the role of forests in climate change has received increased attention. This is due to the observation that not only the atmosphere has a principal impact on vegetation growth but also that vegetation is contributing to local variations of weather resulting in diverse microclimates. The interconnection of plant ecosystems and weather is described and studied as ecoclimates. In this work we simulate ecoclimates by modeling (1) the local climate-response of individual plants in large-scale ecosystems, (2) the vegetation impact on the atmosphere, and (3) the soil hydrology. We employ interactive state-of-the-art methods for simulating ecosystem growth and weather dynamics to enable a realistic animation of vegetation growth, plant competition and cooperation mediated by light and soil water, as well as cloud transitions. Our plant ecosystem model simulates the growth of individual trees with branch-level geometry. We couple an ecosystem with a weather model to locally sample weather variations over time, which enables us to simulate the long-term climate-response of individual tree models. Simultaneously, the composition of an ecosystem affects the development of weather: individual trees vary in how they release vapor or transfer heat to the air. Our framework allows us to interactively explore the growth- and climate-response of individual trees and of an ecosystem as a whole.

How to cite: Makowski, M., Hädrich, T., Michels, D., Pirk, S., Palubicki, W., and Skowronska, W.: Interactive Simulation of 3-D Ecoclimates, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-16600, https://doi.org/10.5194/egusphere-egu21-16600, 2021.