Observation of Smoke and Dust Plume Transport and Impact on the Air Quality Remote Sensing in New York City Yonghua Wu*, Chowdhury Nazmi, Cuiya Li, Daniel Hoyos, Barry Gross, Fred Moshary NOAA-CREST and Optical Remote Sensing Lab, City College of New York (CCNY), NY 10031, USA. *Email: yhwu@ccny.cuny.edu

1. Motivation Wild fires and dust storms inject large amount of aerosols that can be

transported in the long distance. --Influences on air quality in the local, regional and continental scales; challenge for modeling and local emission management by EPA. Potential effects on cloud formation, radiation and climate. --Dust as ice nuclei (IN), biomass burning aerosols as IN or CCN of clouds.

An issue for satellite remote sensing of air quality, e.g. using aerosol

optical depth (AOD) to estimate ground PM2.5. --Aloft plumes or plumes mixing down to PBL and surface.

This study focuses on: -- Cases: smoke & dust vertical distribution, optical properties and types. -- Climatology: monthly occurrence and transport paths. Quantify the influences on local aerosol optical properties, PM2.5, and correlation of column AOD-PM2.5.

2

2. Ground-based aerosol/cloud remote sensing testbed at CCNY a. A multi-wavelength Mie-Raman Lidar

1064,532,355,387, 407-nm, 3-elastic & 2-Raman channels Vertical distribution of aerosol, cloud and water vapor; Aerosol optical properties: extinction/backscatter,

Angstrom exponent, lidar-ratio, PBL-top; 2-3 day/week daytime run (10:00 ~ 18:00 EDT) b. A Ceilometer (Vaisala CL31/51) near surface aerosol and cloud height up to 7.5km; 24-hr/7-day automatic run. c. A CIMEL sunphotometer (SP) (AERONET-CCNY) AOD at 8-wavelength 340~1064nm, Angstrom exponent; Inversion data (volume size distribution, refractive index,

single-scattering-albedo). d: Air quality monitoring station NYDEC: PM2.5, O3, CO e. Multi-filter shadow band radiometer (MFR-07) f. Microwave radiometer (MWR-3000a , T, RH, liquid water) g. Wind Doppler lidar

3

Radar antenna

Lidar hatch Ceilometer

MFR-07 Microwave radiometer Cimel SP

3. Results: Snapshots of aloft aerosol plumes in 2012

Spring

Fall

Summer

Aloft aerosol plumes (yellow) on the different days and seasons

Aloft plumes Aloft plumes Aloft plumes

Cloud

PBL

Aloft plumes Cloud

5 5

(1) Case: Summer haze related to wildfire smoke transport

NYC urban

NYC up- & downwind nonurban

EPA-PM2.5 stations

Aloft aerosols plume

Aerosol plume

Angstrom exp=~2, fine mode smoke particles

NAAQS 35μg/m3

AOD= 0.4 at 532-nm

6 6

Transport path and source by model and satellite

Mean AOD from MODIS/Aqua June 7-12

NYC

http://www.ssd.noaa.gov/PS/FIRE/GASP/gasp.html

NOAA GASP-AOD on June 12

7 7

Case: Asian dust (coarse-mode dominated) long-range transport

Aloft plume: 70~80% to total AOD (0.3~0.5) Angstrom exp. ~1.0 (coarse mode)

PBL aerosol

7-day duration

OMPS AI over VIIRS RGB image

ozoneaq.gsfc.nasa.gov/omps/blog/2015/04/

Asian dust

Siberian fire smoke

8

Depolarization ratio and color-ratio in the East US by CALIPSO ---Coarse mode and nonspheric particles dominated (Dust-like)

Dust/polluted dust

(2) Climatology analysis : Range-resolved monthly occurrence

2006-2013

9

Define an aloft-aerosol-layer event: height> PBL-top, geometric depth ∆z>300-m Duration time T>=3-hr Occurrence fre. during 2006-2013: Aloft-plume-day/tot-lidar-day Total days count per month > 20; plume days >10 per month Main occurrence: March to Sep., Z< 8-km, low in summer, high in Mar-Apr

(3) Climatology analysis : Cluster of transport paths

10

Six clusters of transport paths at 4-km level and 72-hr long duration. Cluster-2: higher AE and SSA, but smaller Reff. Fine mode and non-absorbing aerosol: industrial aerosols such as sulfate, nitrate. Cluster-6: Smaller AE and large Reff , but higher SSA. Coarse-mode Asian dust. Cluster 1, 3, 4: Smaller SSA and Reff (fine-mode absorbing aerosols: smoke) Cluster-5: mixture of smoke and dust.

Cluster # AOD500 AE SSA670 Reff (µm)

1 0.19 1.42 0.90 0.28 2 0.31 1.58 0.92 0.24 3 0.22 1.59 0.81 0.21 4 0.25 1.37 0.90 0.23 5 0.16 1.47 0.94 0.20

6 0.15 1.32 0.93 0.31

AE: Angstrom exponent at 440-870 SSA670: single-scattering albedo at 670-nm Reff: effective radius of aerosol

Total 283 trajectories/HYSPLIT model

Dust-like

smoke

sulfate

(4) Transport influence on local PM2.5 (mixing-down vs. clear skies)

0 5 10 15 200

5

10

15

20

Time (EST, hr)

PM

2.5 c

once

ntra

tion

(µg/

m3 )

PM2.5

average at Newburgh, Apr-Sep 2007-2013

lidar-mixingdown,N=109lidar-clear/nonmixing,N=238EPA-all days, N=1260

11

0 5 10 15 200

5

10

15

20

Time (EST, hr)

PM2.

5 con

cent

ratio

n ( µ

g/m

3 )

PM2.5

average at CCNY-site, Apr-Sep 2007-2013

lidar-mixingdown (N=109)lidar-clear/nonmixing (N=238)EPA-all days (N=1260)difference

CCNY-site/NYC Newburgh, NYS Upwind nonurban

Mean PM2.5 over the days of plume mixing down into PBL or clear/nonmixing down skies. Increase by 3~5 µg/m3 on average during the mixing down days (2007-2013, summer). Similar trends at other upwind sites.

0 0.5 1 1.50

10

20

30

40

50

60

70 Y=40.86X+1.83

R2= 0.72

PM2.5

vs. MODIS-AOD, lidar-clear-day/summer, 6-site/NYC

MODIS-AOD

PM

2.5 (µ g

/m3 )

Clear sky filtered out aloft plume days

12

Identify aerosol-transport events from the lidar profiles. Evaluate the AOD-PM2.5 correlation in NYC during the clear sky and aloft-plume days

during summer, 2006-2013. Linear correlation R2 increases from 0.39 (aloft plume) to 0.72 (clear skies). Regression linear slope increases from 29.6 (aloft plume) to 40.9 (clear skies).

(5) Aloft plume influence on MODIS-AOD and PM2.5 correlation

0 0.5 1 1.50

10

20

30

40

50

60

70 Y= 29.55X+2.72R2= 0.39

PM2.5

vs. MODIS-AOD, Lidar-plume-day/summer, 6-site/NYC

MODIS-AOD

PM

2.5 (µg

/m3 )

Aloft plume days

Summer, 2006-2013

Lidar-ceilometer-sunphotometer measurements: time-height distribution and evolution of aerosol, optical properties and type classification.

Climatology analysis of monthly occurrence of aloft aerosol plumes, transport path and mean optical properties.

6 clusters of transport paths at 4-km altitude and 72hr transport. Cluster-2: fine-mode non-absorbing industrial aerosols. Cluster-6: coarse mode dust-dominated. Cluster-3 and other: fine-mode absorbing smoke aerosols.

Increase of the ground PM2.5 on those plume mixing-down into PBL. Enhance correlation and linear regression slope between the AOD-PM2.5

when filtering out the aloft aerosols.

Near-term plans: add a depolarized channel for discriminating dust/smoke aerosol; combine lidar-SP-LiRIC retrieving aerosol mass profiles.

4. SUMMARY

13

ACKNOWLEDGEMENTS. This study was supported by National Oceanic and Atmospheric Administration (NOAA) under the CREST Grant # NA11SEC4810004. Authors greatly appreciate the data from NASA-AERONET, MODIS, EPA, and NOAA-HMS, HYSPLIT and GASP products, and the partners of NOAA-CREST Lidar Network for sharing aerosol transport information.

Thanks for your attention ! Comments and Questions?

14