Global challenges such as biodiversity loss and climate change are compounded by human activities, leading to increasingly complex ecological dynamics and urgent knowledge gaps. While advances in high-resolution satellite imaging, sensor networks, and citizen science now generate vast spatial and temporal datasets, traditional methods struggle to capture the nonlinear processes and interactions that shape ecosystems. Cutting-edge approaches in geospatial data science, artificial intelligence, and spatio-temporal modeling offer powerful ways to integrate these diverse data sources, yielding deeper insights to address relevant research questions, why biodiversity changes vary across space and time, how they are linked to climate shifts and other anthropogenic developments, how dynamics of ecosystems can provide insights/signals into their health status (and other risks to humans) and resilience, whether and how these insights are linked to capacity of ecosystems to maintain biodiversity, or can signal the risk of collapses in species communities and ecosystem functions. By identifying early warning signals of critical transitions and quantifying ecosystem resilience, researchers can better anticipate tipping points and guide more proactive management. My motivation lies in harnessing these methodological innovations to not only improve our understanding of ecosystem functioning and its implications for human well-being, but also inform effective interventions to reverse biodiversity loss.
In my research group, we have established two primary research themes that align with our overarching goal of integrating advanced computational methods with ecological data to address pressing environmental challenges:
- Monitoring Ecosystem Health through Spatio-Temporal Remote Sensing
This research line focuses on detecting and understanding symptoms of ecosystem degradation by leveraging spatio-temporal remote sensing technologies. We analyze time-series satellite imagery to monitor ecosystem dynamics, assess health indicators such as resilience, and develop early warning systems for critical transitions. By identifying patterns and anomalies over time, we aim to provide timely insights that can inform conservation strategies and mitigate ecosystem collapse. For instance, our work contributes to understanding the rapid declines in wildlife populations, as highlighted by recent reports indicating a 73% decrease over the past 50 years.
The Guardian
- Advancing Biodiversity Change Research
The second theme centers on both methodological and applied aspects of biodiversity change. We strive to enhance modeling techniques, such as species distribution models, to improve biodiversity assessments and uncertainty evaluations. Our research also delves into the geographical patterns of biodiversity change and investigates the drivers behind these changes, including habitat fragmentation, over-exploitation, and invasive species. By refining these models, we aim to better predict and manage the impacts of these drivers on biodiversity. Our approach aligns with the principles of ecological forecasting, which utilizes knowledge of ecological processes to predict future changes in ecosystems.