1

Shobha KondraguntaNOAA/NESDIS Center for Satellite

Applications and Research

Introducing VIIRS Aerosol Products for Global and

Regional Model Applications

VIIRS Aerosol Cal/Val Team2

Name Organization Major TaskKurt F. Brueske IIS/Raytheon Code testing support within IDPS

Ashley N. Griffin PRAXIS, INC/NASA JAM

Brent Holben NASA/GSFC AERONET observations for validation work

Robert Holz UW/CIMSS Product validation and science team support

Nai-Yung C. Hsu NASA/GSFC Deep-blue algorithm development

Ho-Chun Huang UMD/CICS SM algorithm development and validation

Jingfeng Huang UMD/CICS AOT Algorithm development and product validation

Edward J. Hyer NRL Product validation, assimilation activities

John M. Jackson NGAS VIIRS cal/val activities, liaison to SDR team

Shobha Kondragunta NOAA/NESDIS Co-lead

Istvan Laszlo NOAA/NESDIS Co-lead

Hongqing Liu IMSG/NOAA Visualization, algorithm development, validation

Min M. Oo UW/CIMSS Cal/Val with collocated MODIS data

Lorraine A. Remer UMBC Algorithm development, ATBD, liason to VCM team

Andrew M. Sayer NASA/GESTAR Deep-blue algorithm development

Hai Zhang IMSG/NOAA Algorithm coding, validation within IDEA

NOAA Team

Navy Global Aerosol Forecasting

• Assimilation system is “2D-VAR” for total column AOD based on NAVDAS 3DVAR system

• Large development effort required for producing DA-quality products from off-the-shelf MODIS data

• Operational AOD assimilation reduces RMS error in analyzed AOD by >50%

• Operational at FNMOC from September 2009 (over ocean)

• Land and ocean MODIS assimilated in operations since February 2012

• Publications about aerosol DA

• Zhang, J. L., et al.: Evaluating the impact of assimilating caliop-derived aerosol extinction profiles on a global mass transport model, Geophys. Res. Lett., 38, L14801, doi:/10.1029/2011gl047737, 2011.

• Zhang, J. L., et al.: A system for operational aerosol optical depth data assimilation over global oceans, J. Geophys. Res.-Atmos., 113, D10208, D10208, doi:/10.1029/2007jd009065, 2008.

GEOS-5/GOCART Forecasts

CO

Smoke

SO4

http://gmao.gsfc.nasa.gov/forecasts/

Global 5-day chemical forecasts customized for each campaign O3, aerosols, CO, CO2, SO2

Resolution: Nomally 25 km

Driven by real-time biomass emissions from MODIS

Assimilated aerosols interacts with circulation through radiation

NOAA GFS Aerosol Component (NGAC)

5

Model Configuration:

Forecast model: Global Forecast System (GFS) based on NOAA Environmental Modeling System (NEMS), NEMS-GFS

Aerosol model: NASA Goddard Chemistry Aerosol Radiation and Transport Model, GOCART

Phased Implementation:

Dust-only guidance is established in Q4FY12

Dust, sea-salt, OC/BC, and sulfate aerosol forecast once real-time global smoke emissions are developed and tested (NWS/NCEP-NESDIS/STAR-NASA/GSFC collaboration)

Near-Real-Time Dust Forecasts

5-day dust forecast once per day (at 00Z), output every 3 hour, at T126 L64 resolution

ICs: Aerosols from previous day forecast and meteorology from operational GDAS

6

Need to know: IP is a pixel level AOT retrieval; EDR is an aggregate of about 8 X 8 IP pixels; IP product has Navy Aerosol Analysis and Prediction System AOT filled in for pixels with no retrievals but this is not included in EDR; no AOT over inland water bodies; no negative retrievals allowed; no AOTs over 2.0 reported; VIIRS products come with several quality flags that are useful to screen the data.

VIIRS Granule: 86 seconds; 48 scan lines; 3040 km swath width; 16 M-bands (412 nm to 12 µm); 768 X 3200 fixed array size per granule.

Outputs (HDF5): (1) AOT Intermediate Product, IP; (2) AOT Environmental Data Record, EDR; (3) Suspended Matter; (4) Aerosol Model Information; (5) Geolocation file for IP; (6) Geolocation file for EDR.

VIIRS vs. MODIS EOS7

MODIS VIIRS

Orbit altitude

690 km 824 km

Equator crossing time

13:30 LT

13:30 LT

Granule size

5 min 86 sec

swath 2330 km

3040 km

Pixel nadir

0.5 km 0.75 km

Pixel edge

2 km 1.5 km

EDR AOT Product nadir

10 km 6 km

EDR AOT Product EOS

40 km 12 km

IP AOT to EDR AOT8

IP (750m) EDR (6 km)

• 8 x 8 750m IP pixels aggregated to 6 km EDRs;

• For EDR to be of “best quality”, at least a minimum of 16 out of 64 IP pixels should have “best quality” AOT;

• During the aggregation process, top 40% and bottom 20% “best quality” IP pixels are discarded.

In contrast to VIIRS, MODIS retrieves AOT by averaging 500m reflectances in a 10 km x 10 km grid and discarding the

top 50% (25%) and bottom 20% (25%) pixels over land (ocean)

9

Daily global mean AOT

Production Error

Algorithm Upgrade

10

AOT (VIIRS – MODIS)

Upgrades to the algorithm to reduce the high bias over land are in the plans. NDVI

dependent spectral surface reflectance ratios will be implemented soon.

Data:• February and March 2013• MODIS C5• VIIRS IDPS• AOT data mapped to 0.25o

grids• AOT data not paired; some

sampling differences exist

11

AOT (VIIRS vs. AERONET)

Metric VIIRS MODIS

Accuracy 0.014 0.003

Precision 0.060 0.056

Uncertainty

0.062 0.056

Metric VIIRS MODIS

Accuracy -0.021 -0.017

Precision 0.164 0.115

Uncertainty

0.165 0.116

12

Smoke from fires on March 15, 2013 seen in the visible RGB image (left panel) nicely captured in the quantitative retrieval of aerosol optical

thickness (right panel). These images are generated at the STAR IDEA (Infusing satellite Data into Environmental Applications) website using direct broadcast SDRs, VCM, and fire hot spot data. Data latency is ~2

hrs. The website (www.star.nesdis.noaa.gov/smcd/spb/aq/) is routinely accessed by air quality forecasters for satellite aerosol

imagery that provides the spatial extent of smoke, dust, and haze.

13

VIIRS true color image of blowing dust from different sources in Alaska on

April 28, 2013

VIIRS Pixel Level AOD

VIIRS Dust Flag

14

Important Dates:

28 Oct2011

2 May2012

15 Oct2012

27 Nov2012

22 Jan2013

today

Initial instrument check out.

Tuning cloud mask parameters

Aerosol product at Beta status

Software production

error

Beta status

Aerosol product

candidate provisional

status

Red periods: DO NOT USE the product.Beta: Use with caution. Known biases. Provisional: Use data bearing in mind that further validation is ongiong

15

Data (not SM) are available through CLASS: http://www.class.noaa.gov

A “back door” for data via the IDEA site:https://www.star.nesdis.noaa.gov/smcd/spb/aq/index_viirs.php?product_id=4

Users’ guide available at: http://www.star.nesdis.noaa.gov/jpss/ATBD.php#S126472 README file under VIIRS aerosol (AOT) at:http://www.nsof.class.noaa.gov/saa/products/welcome

Other documents are available at: http://npp.gsfc.nasa.gov/science/documents.html

JCSDA Atmospheric Composition (AC) Working Group Update

WG is made up of scientists working on data assimilation methods for atmospheric composition modeling Mostly there are operational customers

behind these efforts Only one telecon held since the last

annual meeting. All working group members working on AC working group topics with “leveraged” funding

JCSDA Atmospheric Composition Working Group Update

Several Aerosol-related applications are making their way to operations Zhiquan Liu at NCAR has developed and implemented

aerosol optical depth DA in GSI. Air Force Weather Agency intends to operationalize this.

NASA LANCE is now producing DA-ready AOD from MODIS based on Naval Research Lab / University of North Dakota algorithm

Dust aerosol is now operational in GFS NASA GMAO is assimilating MODIS data for aerosols,

and includes aerosol direct effects in GEOS-5 Development of global biomass burning emissions as

input for NGAC is progressing well

JCSDA Atmospheric Composition Working Group Update

CRTM is being used in these efforts, with difficulty CRTM trunk has GOCART speciation and

microphysical properties hard-coded; cannot modify without altering source code;

CRTM for CMAQ exists; this is another fixed, inflexible configuration

CRTM developers could future-proof current systems and speed development of future systems by extracting the aerosol microphysics into a flexible framework

19

Aerosol Algorithm Differences between VIIRS and MODIS

VIIRS MODIS Dark Target

Retrieval every pixel(0.75 km resolution)

Retrieval after grouping pixels(10 km resolution)

Aggregates AFTER retrieval, discarding outliers in group

Aggregates BEFORE retrieval, discarding outliers in group

Robust AOT at standard 6 km resolution at nadir, 12 km at edge

Robust AOT at standard 10 km resolution at nadir, 40 km at edge

LAND: chooses aerosol model LAND: mixes fixed fine and coarse models

LAND: fits to atmospherically corrected surface reflectance ratios

LAND: fits to top of atmosphere reflectances

LAND Spectral AOT at 11 wavelength with 1 primary wavelength

LAND AOT at 3 wavelengths with 1 primary wavelength

OCEAN: Based on Tanré retrieval OCEAN: Based on Tanré retrieval

OCEAN spectral AOT at 11 wavelength with 1 primary wavelength

OCEAN spectral AOT at 7 wavelength with 1 primary wavelength

Heavily dependent on cloud mask from outside

Mostly dependent on internal cloud mask.

NO negative values. Max AOT = 2.0 YES negative values. Max AOT = 5.0

Bands used: 0.411, 0.466, (0.550), 0.554, 0.646, 0.856, 1.242, 1.629, 2.114 µm

Bands used: 0.412, 0.445, 0.488, (0.550), 0.555, 0.672, 0.746, 0.865, 1.24, 1.61, 2.25 µm

No level 3 gridded products yet Level 3 daily, 8-day, monthly mean