Aoxue XU

I am an environmental policy researcher pursuing a Master of Environmental Policy (iMEP) at Duke University, with study at both the Durham and Kunshan campuses. My work sits at the intersection of environmental and land-use economics, energy and climate policy, and urban environmental measurement. I use causal inference and spatial methods to study how economic shocks, regulation, and policy interventions reshape land use and environmental outcomes.

My master's thesis, supervised by Prof. Binbin Li and developed in collaboration with research partners at IPAM and ICV in Brazil, examines how trade and policy shocks reshape deforestation and land-use decisions. The empirical setting is a multi-year panel of Brazilian municipalities; methods combine difference-in-differences with attention to biome heterogeneity across the Amazon and Cerrado.

My undergraduate training in environmental chemistry gave me a mechanistic understanding of pollutants and biogeochemistry; during my master's I pivoted to environmental economics, where I apply causal-inference and spatial methods to study how trade and policy shocks reshape environmental outcomes. As first author, I published in Acta Scientiae Circumstantiae (北大核心, CSCD) on biochar-derived DOM characterization, and co-authored work in Environmental Pollution (SCI Q1, IF: 7.2, JCR 2024) and Environmental Science (北大核心, CSCD). I supported applied-economics research at the Duke Kunshan Environmental Research Center and as a research assistant to Prof. Wumeng He.

Beyond environmental research, I have developed a strong interest in applying AI to the research workflow. To that end, I designed and built a full-stack literature research workbench at workbench.hazalxu.click that integrates multi-source search, screening, retrieval, and synthesis support for end-to-end literature work.

"How do economic shocks, environmental regulation, and spatial measurement jointly shape land-use and climate outcomes?"

My quantitative work uses causal inference (DiD, event studies) and remote sensing (PRODES, Sentinel-2, nighttime lights) to identify policy effects on environmental, land-use, and ecological outcomes, with placebo and permutation tests as robustness checks. My qualitative work complements this with semi-structured interviews and grounded-theory coding, as in a recent field study of mountain tea villages in Xishuangbanna.