We seek to understand how global change affects terrestrial ecosystems. We ask fundamental questions: How do global change drivers—from rising temperatures to urban expansion and deforestation—affect ecosystem seasonality, function, and resilience? We combine satellite remote sensing, field and greenhouse experiments, Earth system modeling, and AI/ML approaches to reveal patterns and mechanisms across scales. 

1. Anthropogenic drivers of vegetation phenology

The timing of seasonal biological events — such as when trees leaf out, flowers open, and leaves turn yellow — is called phenology. Phenology involves incredibly complex processes, but temperature and day length have been identified as critical environmental cues for trees to track time. While climate warming is a well-recognized driver of phenological change, our research shows that anthropogenic pressures such as urban warming and artificial light at night (ALAN) can exert equally strong or stronger influences on forest ecosystems. 

We found that the urban heat island effect leads to an earlier start of the season, but urban phenology showed less temperature sensitivity, shifting earlier but at a slower rate compared to rural areas (Meng et al. 2020, PNAS). At the same time, artificial light caused earlier leaf bud break in spring and delayed leaf coloring in autumn at the national scale (Meng et al. 2022 PNAS NEXUS). Building on these findings, we extended it globally and found ALAN had a greater impact on phenology than temperature across 428 cities worldwide (Wang et al. 2025, Nature Cities). 

Our research connects directly to management, policy and public health. We analyzed municipal legislation on ALAN reduction and found that cities with targeted lighting policies (e.g., New York City and Washington, D.C.) exhibited the greatest decreases in blue-spectrum light, alongside reduced ecosystem impact (Geist, under review). Building on this, we are investigating the effects of ALAN on pollen dynamics to assess implications for allergen exposure and human health (Geist, PNAS Nexus, 2026). These findings unveiled a previously unrecognized mechanism through which anthropogenic drivers influence urban ecosystem seasonality, forest health, and resilience, with cascading implications for ecosystem services, carbon cycling, and public health. 

2. Vegetation phenology and climate feedback

My research uses plant phenology as a lens to understand forest responses to climate change and shapes the strength and stability of terrestrial carbon sinks. I investigate the environmental drivers and feedbacks of phenology, with the goal of disentangling mechanisms and improving predictive models for forest ecosystems. For example, I demonstrated that asymmetric diurnal warming (i.e., nighttime temperatures rising faster than daytime) has distinct impacts on phenology, and incorporating this effect into models significantly improved forecast accuracy (Meng et al. 2020, AFM). I further disentangled the roles of temperature and photoperiod by leveraging the unique topography of the northern Alps, where elevation decreases with latitude, as a natural experiment. This work demonstrated that photoperiod moderates the advance of spring phenology under climate change, leading to the development of two photoperiod-based models (Meng et al. 2021, GCB). 

Phenology is not only a sensitive indicator but also feedbacks to the climate system. I introduced new seasonal-deciduous phenology schemes into the land model of the U.S. Department of Energy’s (DOE) Energy Exascale Earth System Model (ELM of E3SM), highlighting the crucial role of phenology in terrestrial-climate interactions (Meng et al. 2021, AFM). Further, we examine how phenology regulates local climate through two competing biophysical processes, evapotranspiration-induced cooling and albedo-caused warming, and found that cooling consistently outweighs warming (Li et al. 2025, PNAS).

3. Fire and land use in tropical ecosystems

Understanding how disturbances alter the temporal dynamics of ecosystems is crucial for predicting their resilience and carbon cycling under future climate scenarios. We focus on on the temporal shifts in these ecosystem processes induced by drought and fire. Historically, transpiration in the central Amazon was considered light-limited rather than water-limited, owing to abundant soil water even during dry seasons. However, my research revealed a significant temporal shift, with transpiration being constrained from light to water during the 2015 El Niño drought (Meng et al, 2022, ERL). Integrating satellite, field, and modeling data, we further showed that “hot droughts”, defined by elevated vapor pressure deficit and temperature, represent a critical new disturbance regime, projected to increase markedly after 2040 under CMIP6 multi-model ensembles (Chambers et al, Nature, 2026). Currently, we are investigating how climate change and deforestation alter fire regimes in the Amazon.