Our research aims to understand and better model terrestrial ecosystem carbon and water cycling under a multitude of global change factors. We tackle important ecological questions and test hypotheses with multi-disciplinary tools, including ground-based  experiments and inventory data, remote sensing, statistical and process-based modeling, machine learning and AI. Major research themes include: (1) vegetation dynamics and ecosystem functioning, (2) water cycle and ecosystem drought impacts, (3) land use change and carbon cycle, and (4) model-data integration. We are also interested in topics such as climatic and hydrological feedbacks of vegetation changes and fires, theories on plant community assembly, and conservation ecology. 
​Our research has been funded by the U.S. Department of Energy (DOE), USDA NIFA, NASA, USGS, and CSU. 

​Vegetation Dynamics and Ecosystem Functioning
Understanding how vegetation responds to climate change is critical for predicting shifts in ecosystem functioning, carbon cycling, and biodiversity. Our research addresses several big-picture questions: How do changing climates alter the timing and duration of key phenological events such as leaf green-up and senescence? What are the spatially varied implications of extreme climate events, such as droughts and heatwaves, on vegetation structure and productivity? How do large-scale climatic gradients, such as warming rates across continents, influence vegetation redistribution and dynamics? To address these questions, we combined remote sensing, ground-based observations, and modeling approaches. For instance, using multi-source satellite data, we found that the ratio of time allocated to vegetation green-up versus senescence remained remarkably stable across northern ecosystems, even as growing seasons extended under warming. This finding suggests intrinsic biotic controls that regulate phenology independently of climate variation, challenging the assumption that climate entirely dictates phenological responses (Science Advances, 10(23), eadn2487). We also demonstrated how the choice of data and methods influences our understanding of phenology by comparing solar-induced chlorophyll fluorescence (SIF) with vegetation indices like normalized difference vegetation index (NDVI), as SIF—directly tied to photosynthesis—showed shorter growing seasons than those indicated by NDVI (Agricultural and Forest Meteorology, 323, 109027).
To assess the impacts of climate extremes on vegetation dynamics, we investigated how droughts, heatwaves, and other climatic events influence productivity and growth at multiple spatial scales. Analysis of satellite-derived NDVI data revealed that while localized tree mortality events leave distinct imprints on productivity, their effects are often masked by long-term greening trends at broader spatial resolutions, highlighting the importance of spatiotemporal scales (Nature Ecology & Evolution, 8(5), 912-923). Furthermore, we identified the drivers of negative extreme anomalies in vegetation growth (NEGs) globally, showing that 70% of NEGs are attributable to compound and individual climate extremes, with dominant drivers varying by biome and region. For example, cold and wet extremes affect temperate ecosystems, while soil drought and compound extremes dominate in tropical and semi-arid regions (Global Change Biology, 29(8), 2351-2362). Additionally, we examined the 2015–2016 El Niño, one of the strongest on record, to reveal an unusual phenomenon of simultaneous increases in atmospheric CO2 growth rate and seasonal-cycle amplitude that could signal a decoupling between seasonal and annual carbon cycle dynamics (Global Change Biology, 27(16), 3798-3809). Together, these findings provide critical insights into the local and global effects of climate extremes, informing strategies for ecosystem resilience.

Meng et al. (2024) Science Advances

• Meng, F., Felton, A. J., Mao, J., Cong, N., Smith, W. K., Körner, C., Hu, Z., Hong, S., Knott, J., Yan, Y., Guo, B., Deng, Y., Leisz, S., Dorji, T., Wang, S., & Chen, A. (2024). Consistent time allocation fraction to vegetation green-up versus senescence across northern ecosystems despite recent climate change. Science Advances, 10(23), eadn2487. https://doi.org/10.1126/sciadv.adn2487
• Chen, A., Meng, F., Mao, J., Ricciuto, D., & Knapp, A. K. (2022). Photosynthesis phenology, as defined by solar-induced chlorophyll fluorescence, is overestimated by vegetation indices in the extratropical Northern Hemisphere. Agricultural and Forest Meteorology, 323, 109027. https://doi.org/10.1016/j.agrformet.2022.109027
• Yan, Y., Piao, S., Hammond, W. M., Chen, A., Hong, S., Xu, H., Munson, S. M., Myneni, R. B., & Allen, C. D. (2024). Climate-induced tree-mortality pulses are obscured by broad-scale and long-term greening. Nature Ecology & Evolution, 8(5), 912–923. https://doi.org/10.1038/s41559-024-02372-1
• Yang, H., Munson, S. M., Huntingford, C., Carvalhais, N., Knapp, A. K., Li, X., Peñuelas, J., Zscheischler, J., & Chen, A. (2023). The detection and attribution of extreme reductions in vegetation growth across the global land surface. Global Change Biology, 29(8), 2351–2362. https://doi.org/10.1111/gcb.16595
• Wang, K., Wang, X., Piao, S., Chevallier, F., Mao, J., Shi, X., Huntingford, C., Bastos, A., Ciais, P., Xu, H., Keeling, R. F., Pacala, S. W., & Chen, A. (2021). Unusual characteristics of the carbon cycle during the 2015−2016 El Niño. Global Change Biology, 27(16), 3798–3809. https://doi.org/10.1111/gcb.15669

Water Cycle and Drought Impacts
Understanding the interplay between water availability and ecosystem function is essential for predicting resilience to drought and hydrological extremes under climate change. Our research focuses on understanding the intricate relationships between soil moisture, hydrological variability, and ecosystem productivity, particularly in the context of drought resilience. We aim to enhance methods for monitoring and modeling water availability to more accurately predict how ecosystems respond to shifts in precipitation patterns and climate extremes. We demonstrated that soil moisture is a key factor in understanding ecosystem resilience to drought, with pre-growing-season moisture playing a critical role in mitigating drought impacts on vegetation productivity. For instance, in the Mongolian Plateau, spring soil moisture carried over from the previous growing season was shown to sustain vegetation productivity during subsequent dry periods, underscoring the importance of seasonal soil moisture dynamics (Environmental Research Letters, 16(1), 014050). At the global scale, we examined the relationship between gross primary production (GPP) and SIF, revealing that moisture availability modulates this relationship. Our findings showed that GPP/SIF ratios consistently decrease in hot and dry climates due to stomatal responses to aridity, emphasizing the need to incorporate water availability into models of ecosystem productivity (Global Change Biology, 27(6), 1144-1156).
Additionally, to advance drought assessment frameworks, we underscored the importance of integrating ecosystem-specific water demands into traditional monitoring systems, highlighting the need to consider divergent water requirements across vegetation types and regions. By leveraging multi-source satellite and ground-based datasets, this approach provides a more nuanced understanding of how ecosystems respond to water scarcity and drought, offering critical insights into the hydrological processes underpinning ecosystem resilience and informs adaptive strategies for managing water resources in a changing climate (Nature Water, 2(3), 215-218). 

Cui et al. (2024) Nature Water

• Lu, Z., Peng, S., Slette, I., Cheng, G., Li, X., & Chen, A. (2021). Soil moisture seasonality alters vegetation response to drought in the Mongolian Plateau. Environmental Research Letters, 16(1), 014050. https://doi.org/10.1088/1748-9326/abd1a2
• Chen, A., Mao, J., Ricciuto, D., Xiao, J., Frankenberg, C., Li, X., Thornton, P. E., Gu, L., & Knapp, A. K. (2021). Moisture availability mediates the relationship between terrestrial gross primary production and solar-induced chlorophyll fluorescence: Insights from global-scale variations. Global Change Biology, 27(6), 1144–1156. https://doi.org/10.1111/gcb.15373
• Cui, J., Chen, A., Huntingford, C., & Piao, S. (2024). Integrating ecosystem water demands into drought monitoring and assessment under climate change. Nature Water, 2(3), 215–218. https://doi.org/10.1038/s44221-024-00217-6

Land Use Change, Carbon Cycle
Understanding the mechanisms driving carbon storage, sequestration, and emissions is essential for predicting ecosystem contributions to global carbon cycling and developing effective climate mitigation strategies. We study the impacts of land-use changes, such as afforestation and deforestation, on soil and biomass carbon storage. We also examine how natural disturbances like fires and droughts drive carbon emissions. A key focus is uncovering the fundamental processes that govern carbon dynamics and sequestration across diverse ecosystems. For example, by investigating the impacts of afforestation on soil carbon pools with an extensive field database, we uncovered context-dependent effects on both soil inorganic carbon (SIC) and soil organic carbon (SOC). Afforestation increases SIC in acidic soils but decreases it in alkaline soils, emphasizing the importance of soil chemistry and tree species selection in determining carbon storage outcomes (Global Biogeochemical Cycles, e2021GB007038). We further showed that afforestation can neutralize soil pH, with species-specific thresholds determining whether soil pH increases or decreases, directly influencing carbon sequestration potential (Nature Communications, 9, 520). In another study, we found that SOC responses to afforestation are highly context-dependent, increasing in carbon-poor soils while decreasing in carbon-rich soils, particularly in deeper soil layers. These findings challenge the fixed biomass-to-soil carbon ratios commonly assumed in carbon sequestration models and highlight the need to incorporate pre-afforestation soil properties for accurate carbon storage projections (Nature Sustainability, 3(9), 694-700). 
Our work also examines carbon emissions from ecosystem disturbances. By integrating satellite and model data, we quantified fire-induced carbon emissions across China over two decades, demonstrating that fires significantly contribute to regional carbon fluxes and are driven by both climatic and anthropogenic factors. This work highlights the need for improved fire management to mitigate emissions (Geography and Sustainability, 1(1), 47-58). Together, these studies provide critical insights into the interplay between land use, climate, and carbon processes, which can inform policies to enhance carbon sequestration and reduce emissions in diverse ecosystems.

Hong et al. (2020) Nature sustainability

• Hong, S., & Chen, A. (2022). Contrasting Responses of Soil Inorganic Carbon to Afforestation in Acidic Versus Alkaline Soils. Global Biogeochemical Cycles, 36(1), e2021GB007038. https://doi.org/10.1029/2021GB007038
• Hong, S., Piao, S., Chen, A., Liu, Y., Liu, L., Peng, S., Sardans, J., Sun, Y., Peñuelas, J., & Zeng, H. (2018). Afforestation neutralizes soil pH. Nature Communications, 9(1), 520. https://doi.org/10.1038/s41467-018-02970-1
• Hong, S., Yin, G., Piao, S., Dybzinski, R., Cong, N., Li, X., Wang, K., Peñuelas, J., Zeng, H., & Chen, A. (2020). Divergent responses of soil organic carbon to afforestation. Nature Sustainability, 3(9), 694–700. https://doi.org/10.1038/s41893-020-0557-y
• Chen, A., Tang, R., Mao, J., Yue, C., Li, X., Gao, M., Shi, X., Jin, M., Ricciuto, D., Rabin, S., Ciais, P., & Piao, S. (2020). Spatiotemporal dynamics of ecosystem fires and biomass burning-induced carbon emissions in China over the past two decades. Geography and Sustainability, 1(1), 47–58. https://doi.org/10.1016/j.geosus.2020.03.002

Modeling and Technological Integration in Ecosystem Research
Advancing ecosystem research requires leveraging cutting-edge modeling approaches and technological innovations to enhance predictions and inform management strategies. Our research focuses on leveraging satellite data and machine learning to enhance the precision of ecosystem process models. By integrating data from multiple sources, we aim to improve predictions of vegetation productivity and carbon cycling, advancing our ability to understand and manage ecosystem dynamics.
We utilized satellite-derived SIF to improve the parameterization of photosynthesis in land surface models, particularly the DOE’s E3SM Land Model. By integrating SIF data with machine learning techniques, we demonstrated that SIF can effectively constrain model uncertainty and improve predictions of GPP at both local and global scales. Our findings also revealed substantial spatial and seasonal variations in the relationship between SIF and GPP, underscoring the need to incorporate climate-driven variations into productivity models (Global Change Biology, 27(6), 1144-1156; Journal of Advances in Modeling Earth Systems, 15(4), e2022MS003135).
To address gaps in herbaceous ecosystem modeling, we explored the limitations of process-based dynamic vegetation models, which have traditionally underrepresented grasslands and other herbaceous ecosystems. Through simulation experiments and empirical data synthesis, we identified key ecological processes and traits that need to be incorporated to improve the representation of these ecosystems in Earth system models (Global Change Biology, 29, 6453–6477). 

Chen et al. (2021) Global Change Biology

• Chen, A., Mao, J., Ricciuto, D., Xiao, J., Frankenberg, C., Li, X., Thornton, P. E., Gu, L., & Knapp, A. K. (2021). Moisture availability mediates the relationship between terrestrial gross primary production and solar-induced chlorophyll fluorescence: Insights from global-scale variations. Global Change Biology, 27(6), 1144–1156. https://doi.org/10.1111/gcb.15373
• Chen, A., Ricciuto, D., Mao, J., Wang, J., Lu, D., & Meng, F. (2023). Improving E3SM Land Model Photosynthesis Parameterization via Satellite SIF, Machine Learning, and Surrogate Modeling. Journal of Advances in Modeling Earth Systems, 15(4), e2022MS003135. https://doi.org/10.1029/2022MS003135
• Wilcox, K. R., Chen, A., Avolio, M. L., … Komatsu, K. J. (2023). Accounting for herbaceous communities in process-based models will advance our understanding of “grassy” ecosystems. Global Change Biology, 29(23), 6453–6477. https://doi.org/10.1111/gcb.16950

Current funded projects:
 
Quantifying Boreal and Temperate Ecosystem Vulnerabilities and Their Model Uncertainties (PI Anping Chen). Funded by the DOE Oak Ridge National Lab, we integrate multiple ecosystem functions (e.g., primary productivity, soil carbon storage, water yield) and stressors (e.g., heatwaves, droughts, floods) to create a robust framework for assessing ecosystem resilience under different stresses. The project uses historical ground and satellite data to estimate ecosystem exposure, sensitivity, and adaptability, generating assessment maps of vulnerability across regions and enabling data-model comparisons to improve predictive accuracy for ecosystem vulnerability.
 
Enhancing Drought Monitoring, Prediction, and Impact Assessment to Inform Rangeland Management in the Western Great Plains (PI Anping Chen). Funded by the USDA National Institute of Food and Agriculture (NIFA), the project aims to develop a high-resolution (30-m), near real-time monitoring and forecasting system for drought and forage production, utilizing soil moisture and aboveground net primary productivity (ANPP) as key indicators. By integrating satellite and ground data, applying machine learning, and co-developing tools with ranchers, this system will improve predictions of drought impacts on forage availability. The system will be incorporated into GrassCast, offering weekly updates to support adaptive livestock management.
 
Quantifying and Understanding Dryland Carbon Cycle Pulse Dynamics Using SMAP Soil Moisture and Carbon Flux Monitoring (PI Anping Chen). This NASA-funded project aims to quantify carbon cycle pulses in drylands and identify the processes driving these pulses by leveraging SMAP soil moisture and carbon flux data. By integrating SMAP data with other satellite and ground observations, the study will identify and analyze “hot moments” and “hot spots” of soil moisture and carbon cycling in selected dryland regions. The project will assess how these events contribute to variability and trends in carbon cycling, particularly under water-limited conditions. The findings will enhance understanding of dryland biogeochemical cycles and support improved modeling of carbon fluxes in these ecosystems.
 
Droughts and deluges in semi-arid grassland ecosystems: Implications of co-occurring extremes for C cycling (Co-PI Anping Chen, PI Melinda Smith). Supported by the DOE, this research aims to understand how compound climate extremes, such as co-occurring droughts and deluges, impact carbon (C) cycling in the semi-arid shortgrass steppe of the US Great Plains. We hypothesize that these combined events may trigger intense "hot moments" or "hot spots" of C cycling due to favorable conditions like warm temperatures, soil moisture, and nitrogen availability. We conduct a three-year field experiment to measure the responses of above- and belowground C cycling processes to individual and combined drought-deluge events, while also using historical climate data and remote sensing to expand findings regionally. The study will also integrate these observations into the DOE’s E3SM Land Model to improve Earth System projections, addressing a critical gap in understanding the ecological consequences of increasing climate extremes.

In addition, we are also working with field ecologists in assessing terrestrial ecosystem productivity responses to drought by integrating field experiments, remote sensing, and ­ ecosystem modeling. Key to this research is a globally distributed network of drought experiments across >100 sites. Our complementary research approaches aims to provide a rigorous and comprehensive assessment of ecosystem vulnerability to drought intensification as well as identifying where there is the greatest uncertainty in drought responses.