Preparing for the Future:
Building a Foundation of Science to Support
Policy to Solve Problems
Air Quality Applied Sciences Team
10 th Semi-Annual (AQAST 10)
Dan Costa, Sc.D., DABT
National Program Director
Air, Climate, and Energy Research Program
Office of Research and Development
January 5, 2016

The Many Dimensions of ACE
2

Goal 1: Addressing Climate Change and Improving Air Quality
•
Conduct integrated science assessments of criteria air pollutants and provide new data and approaches for improving these assessments
•
Develop credible models and tools to inform sustainable policies, decisions, and responses to a changing climate by EPA national and regional offices, state, tribal, and local governments, and others
•
Conduct research to change the paradigm for air pollution monitoring, with a focus on lower cost measurements
•
Develop and evaluate models and decision support tools to integrate multi-media processes and systems
•
Develop approaches to assess multi-pollutant exposures and the resulting human and ecological effects of air pollutant mixtures
•
Conduct research to inform policies protecting human and ecosystem health in an evolving energy landscape, including impacts of unconventional oil and gas and lowcarbon energy sources http://www2.epa.gov/sites/production/files/2014-09/documents/epa_strategic_plan_fy14-18.pdf
3

2012 Heat Anomaly
EPA Partner-Stakeholder Priorities
Drought
•
Implementation Sciences - new measurement technologies (e.g., sensors, satellites); air quality models/tools
•
Emissions Science - oil & gas priority; updating multiple inventories (e.g., CAFO’s)
•
Public Health/Welfare multipollutant issues
(e.g., NOx/SOx, near-source risks, exposure science); translational science for public health
•
Climate Change Preparedness – assessments and impacts relevant to adaptation are important
(e.g., cross-media models); mitigation
4

5

Near Road
Objective 1: Assess Impacts
Assess human and ecosystem exposures and effects associated with air pollutants and climate change at individual, community, regional, and global scales
Objective 2: Prevent and Reduce Emissions
Provide data & tools to develop and evaluate approaches to prevent and reduce emissions of pollutants to the atmosphere, particularly environmentally sustainable, cost effective, and innovative multipollutant and sector-based approaches
Objective 3: Prepare for and Respond to Changes in
Climate & Air Quality
Provide human exposure and environmental modeling, monitoring, metrics and information needed by individuals, communities, and governmental agencies to adapt to the impacts of climate change and make informed public health decisions regarding air quality
6

Climate Impacts
Vulnerability and
Adaptation
Assess the impacts of climate change on the environment and public health to inform the development of sustainable approaches to prepare for climate change
Emissions and
Measurements
Atmospheric and
Integrated
Modeling Systems
Protecting
Environmental
Public Health and
Wellbeing
Sustainable Energy and Mitigation
Develop innovative technologies and approaches to characterize source emissions and ambient air pollutants
Develop and apply air quality and cross-media models to support regulatory and community-based decisions
Develop solutions-oriented approaches to assess multipollutant exposures and resulting human and ecological effects of air pollutant mixtures to inform policy and public health practices
Assess the environmental impacts and those factors affecting energy sectors choices from extraction to end-use
7

• New technology revolutionizing regional, community, fenceline, personal monitoring - 1 st 3 Google listings (note EPA)
• EPA’s Air Sensor Toolbox on the web along with a prototype testing platform
• Promoting community science, outreach and education
• ACE Clean Air Res Ctrs working to link satellite data with health outcomes
• Working with NOAA, NASA, NSF to relate satellite-based air quality data
Mobile monitoring for geospatial mapping of pollutants (GMAP)
“Village Green” park bench monitors air quality
Jointly funded Innovation Project with NIEHS
8

A New Air Pollution Monitoring and
Modeling Paradigm
Furthering development, evaluation, and assimilation of new technologies to complement and enhance air quality assessment and forecasting
9

Systematic and concurrent observation of columnintegrated, surface, and verticallyresolved distributions of aerosols and trace gases relevant to air quality as they evolve throughout the day
Ground sites/measurements
• Ambient trace gases and aerosols, primarily based on EPA FRM/FEM
•
Remote sensing of trace gas and aerosol columns
•
Aerosol and Ozone profiles
NASA P-3B (in situ meas.)
In situ profiling of aerosols and trace gases over surface measurement sites
NASA King Air (Remote sensing) Continuous mapping of aerosols with HSRL and trace gas columns with ACAM
10

DISCOVER-AQ: AOD Column
Measurements Used to Predict Daily PM
2.5
•
2012: Collaborative work with NASA and Martin’s
Dalhousie University
– Daily predictions of PM
2.5
by scaling satellite AOD and surface based bias adjustment
– Model captured spatial varying relation of AOD and PM
2.5
but not necessarily magnitude of AOD
, η,
– Method implemented in satellite data stream to AIRNow from
NOAA (van Donkelaar et al., 2012, ES&T)
•
2014: mid-Atlantic Region (Baltimore)
– Purely measurement-based approach showed normalizing AOD with haze layer height (i.e., AOD/HLH) can product reasonable PM
2.5
predictions from AOD over a regional (100 km); results similar to models for scaling
– Characterization of mixed layer height shown to be a key variable in AOD/PM
2.5
•
2015 (Denver) predictions (Chu et al., 2014, AE)
– Spatial, autoregressive model predicts daily PM
2.5
using
VIIRS AOD data, meteorological variables, and previous day PM
2.5
over the conterminous U.S.
– In sample and out-of-sample model validation results indicated marginal predictive improvement with AOD in model
Bias Corrected for Jun 27, 2005
11

Column
Density and
DISCOVER-AQ: NO
2
Column Measurements Provide
Unique Ability to Assess Urban Scale Variability
Mixing Ratio
Mixing
Height
 Assume NO
2
VCD confined to mixing layer
(ML) and well mixed

Use surface and P-3 data to calc. air number density profiles throughout flight day.

Derive ML air density profile along flight track using HRSL ML heights.

Estimate ML NO
2 mixing ratio by normalizing ACAM NO
2
VCD with HSRL ML air number density
•
Scaling NO
2 tropospheric column densities by a modeled boundary layer height accounts for up to
≈75% of the variability between the column and in-situ NO
2
(Knepp et al., J Atmos. Chem. 2013)
•
Derived mixed layer concentrations from ACAM vertical column density NO
2 data at ~1km 2
(Baltimore)
– NO
2 gradients are strongest in A.M. with spatial correlation of 20 km or less
– NO
2 hot spots associated with major point sources and mobile sources and are lower in concentration and more confined in space versus high resolution ( 1 km and 4 km) WRF-CMAQ D-AQ results.
– Mixing layer heights and the vertical distribution of NO2 are likely two of the most controlling variables
12 in translating column-to-surface mixing ratios

•
2.5
2
•
– $32 million investment to four centers - thirteen institutions - over five years
13

• Satellite-retrieved cloud fraction (CF) and cloud optical thickness (COT) could be used to help fill the PM2.5 prediction gap (Yu et al., 2015)
• Statistical downscaling using
PM2.5 LUR models with AOD and techniques for combining data with different spatial resolutions increases accuracy compared to ground measurements (Chang et al.,
2014)
14

• The new Multi-Angle Implementation of Atmospheric
Correction (MAIAC) algorithm for AOD at 1km resolution shows improved correlation to PM
(Chudnovsky et al. 2014)
2.5
measured by 27 EPA ground monitoring stations as
• Introduced a model to expand limited spatial coverage of MODIS daily air temperature to high resolution in large geographical areas (Kloog et al. 2014)
• Used satellite AOD to determine location-specific
PM2.5 trends over 2000-2008 and suggest that primary particles have decreased more than secondary particles (Lee et al. 2014a)
• Developed spatially and temporally resolved exposure assessments of NO
2014b)
2 from a combination of satellite remote sensing and land use regression (Lee et al.,
• Examined chronic and acute exposure to air pollution using fine-scale satellite data to find biases from predicted air pollution exposures (Alexeeff et al. 2014)
15

2.5
(HSPH / SCAPE)
• Objective: to better understand various satellite
PM2.5 models re exposure assessment and monitoring network design
• Models under evaluation:
• SCAPE:
NC domain ~600K km 2, 126 EPA stations. Study period: 2006 – 2008
(1) Chang's spatial downscaler
(2) Liu group’s LME-GWR/GAM model
(3) Russell group’s CMAQ PM2.5 simulation
• Harvard: Koutrakis group’s mixed effects model
• Satellite data under evaluation:
Aqua MODIS C6 10 km dark target AOD
16

•
•
2.5
2
•
17

Air Research: http://www2.epa.gov/air-research
Climate Research: http://www2.epa.gov/climate-research
•
StRAPs for EPA six national research programs, including ACE, available at: http://www2.epa.gov/research/strategic-researchaction-plans-2016-2019
•
Next Generation Air Measuring Research: http://www.epa.gov/air-research/next-generation-air-measuringresearch
•
Discover-AQ: http://www.epa.gov/sciencematters/discover-aq
18