Cooperative Networked Built and Natural Environments
Computational infrastructure and environment system sustainability management and optimization for the protection of human health and natural systems; inter-relationship between infrastructure and air quality, multi-scale air quality modeling to link health risks modeling to air quality policy and infrastructure management.
Understanding the complex emissions and atmospheric pathways of air pollutants and quantifying their societal effects plays a crucial role in societal decision-making because air pollution poses one of the most serious environment and health burdens in both developed and underdeveloped countries. It is recognized that air quality management system is quite complex, requiring accounting of a number of competing temporally- and spatially-dependent parameters and constraints. Significant efficiencies in system-level performance can thus be realized by treating this infrastructure-environment-health problem using optimization formalism and related numerical algorithms (as compared with heuristic treatment).
This thrust of CUTES research aims to formulate and conduct research designed to:
- Characterize the relationships between infrastructure, users, air quality, climate change, and health costs
- Model air pollution and health risks of individual major pollutants (such as ozone, and particles) and multipollutant mixtures
- Assess the health implication of pollutants near sources such as roads, ports, particularly for at-risk human populations who are disproportionately impacted by pollution
- Evaluate the air quality and health impacts of environmental policies and regulation, infrastructure policy and operations management and seek management optimization and policy design towards green infrastructure for livable communities. Results of this research are used to inform the public and decision makers at local, national and international levels for infrastructure investment, planning, and operations management, environmental assessments and plan for air quality standards and public health.
Specifically, ground-level ozone is a secondary air pollutant formed in the atmosphere by complex photochemical reactions of precursor volatile organic compounds (VOCs) and nitrogen oxides (NOx). Transportation emissions are a major source of ozone precursors in urban areas. Identification of efficient control strategies for mitigation of contributing transportation emissions plays a crucial role in ozone reduction policies. One indication of the difficulty in assessing potential effectiveness of transportation emission control strategies is the “ozone weekend effect (OWE).”
In many cities, elevated ozone readings are recorded on weekends (especially Sundays), even though there is less traffic. In this situation, a postulated direct relationship (less transportation emission à less ozone) fails. Hypotheses explaining the OWE have been postulated primarily upon day-of-week variations in the timing, magnitude and mix of transportation activities and the associated emissions.
My research has made important contributions to statistically robust understanding of the OWE through:
- Functional data analysis (FDA) of diurnal ozone cycles and their dynamics
- Non-parametric factorial analysis of the spatial and temporal patterns of transportation activities and emissions
For instance, in our FDA analysis comparisons of Sunday ozone formation rate curve to those on weekdays reveal significantly earlier, faster, and longer duration of ozone accumulation on Sunday, which contributes to the OWE. My nonparametric factorial analyses of the longitudinal Weigh-in-Motion (WIM) traffic data in southern California indicate statistically significant differences in daily total as well as period-based traffic volumes by day of week, and between the weekly patterns of light-duty vehicles (LDV) and heavy-duty trucks (HDT) volumes. The weekend increase in the ratio of transportation VOC to NOx emissions, resulting from decreased HDT activity relative to LDV, boosts ozone accumulation rates and constitutes a plausible factor to explain the OWE in the area (known as NOx-reduction hypothesis). My analyses of transportation activity and emission patterns also provide statistically robust supports to other OWE hypothesis including ozone-quenching hypothesis and aerosol hypothesis (See more details in the appendix).
Futher, using a state-of-the-art air quality simulation model or so-called chemical transport model (CTM), we clarified the risk and uncertainties of long-standing concerns over the public health effects of PM2.5 and its precursor gaseous emissions—we have developed a valuable policy research tool called the Estimating Air pollution Social Impact Using Regression (EASIUR) model. Because CTMs are computationally expensive and require expertise in atmospheric science as well as high performance computing systems, they are generally used to simulate a limited number of policy scenarios even by big institutions like U.S. Environmental Protection Agency (EPA). They often remain out of reach to a large part of the policy research community. Derived from regressions on big (∼30 TB) data created by a state-of-the-art CTM and U.S. EPA’s latest valuation methods using epidemiological studies and environmental economics, the EASIUR model estimates the social costs or public health effects of major air pollutants anywhere in the United States at trivial computational costs without compromising the technical rigor and precision of CTMs. Since the EASIUR model can be updated as its parent tool or input data change or improve, our model will continue to assist policy research in incorporating up-to-date atmospheric science. Built upon the dataset created for and by EASIUR, we have also constructed the air Pollution Social Cost Accounting (APSCA) model, which estimates air pollution social costs at any downwind (receptor) location in unprecedented detail compared with so-called receptor models. For example, this model can quantify the sources of emissions responsible for the public health burden of New York City (NYC) and their contributions that are spatially, temporally, and sectorally resolved. Results from APSCA suggest the possibility of developing region- and season-specific air quality management strategies. More importantly, detailed accounting would inform policy makers what emission sources to target to improve air quality from the receptor’s point of view, creating further research opportunities to develop new methods for designing optimal control strategies at various levels (municipal, state, and federal)
The computational efficiency of our models allows employing formal policy analysis methods such as detailed cost benefit analysis, optimization, and uncertainty analysis, which would open up new research opportunities. For example, it enables us to integrate infrastructure analysis with spatially, temporally, and chemically detailed emission inventories, comprehensive fine scale state of-science modeling and health impact modeling for ozone and fine particulate matter (PM2.5) using complex systems analysis and risk. In a technical report on modeling regional and microenvironmental air quality impact of transportation emissions and clean diesel strategies in NYS, we studied two priority challenges at the transportation-air quality nexus that face researchers and policy makers alike: clean diesel strategies and air quality, and modeling microenvironment air quality near roadways.