Assessing the NPOESS Preparatory Project NPP VIIRS Fire Product as a Climate Data Record'

1
 Assessing the NPOESS Preparatory Project (NPP)
VIIRS Fire Product as a Climate Data Record.  

• Chris Justice,David Roy, Louis Giglio, Ivan
  Csiszar
• University of Maryland/SSAI
• Nov 4-6, 2003 

2
EDR / CDR

• Fire EDRs
• Fire is an Application
• Fire is a Subset of the Surface Type EDR
• Three types of EDR information
• per pixel detection of a fire or fires
• subpixel average temperature of the fire or
  fires detected
• subpixel area of the fire or fires detected
• Climate Data Record (CDR) current concerns re.
  meeting science and applications needs
• For active fire detection product continuity
  VIIRS EDR needs to be evaluated together with
  heritage products to account for sensor
  differences
• No operational heritage for active fire
  area/temperature and burned area products
• IDPS/ADS interfaces and RDR/SDR/TDR/EDR data
  feeds required to generate CDRs in SDS

3
Approach

• Help make the VIIRS EDR for Active Fire
  acceptable for the science and applications
  community
• Ensure that Fire CDRs (e.g., burned area) will
  be able to be generated from the VIIRS data
• Build on Teams Experience with MODIS and
  precursor fire systems (TRMM, AVHRR)
• Instrument
• Algorithm
• Data System (production/archive/distribution)
• Validation
• Extend MODIS fire simulation modeling to VIIRS
• Use of real data to evaluate EDR algorithm
• MODIS
• ASTER
• BIRD
• Aircraft data as available

4
BIRD vs. MODIS active fire
July 24 2002 1515 UTC 50.82N 93.69E (Russia)
Wm-2µm-1str-1
Radiance
Band 21 (3.929-3.989 µm)
HSRS MIR (3.4-4.2 µm)
5
BIRD / MODIS MIR Band Comparison
6
BIRD Radiative Energy
D. Oertel, DLR
7
MODIS BIRD Fire Radiative Energy Comparison
MODIS BIRD

M. Wooster, ( Dept. Geography, Kings College,
London)
8
MODIS Active Fire Validation Collocating ASTER
and MODIS data
Aug 17 2001 0908 UTC 18.8S 19.9 E (NE Namibia)
White squares MODIS fire pixels
Burn scar
Fire fronts
R 2.16 µm G 1.65 µm B 0.56 µm
Smoke
9
Near-coincident BIRD low gain ASTER fire
observations
Santarém/Tapajos, Brazil October 8 2003 1405 UTC
Terra/ASTER
BIRD/HSRS
10
Fire Product Science and Applications Needs

• Use MODIS products as a prototype of what will be
  needed from and provided by VIIRS
• Standard
• Browse
• Climate Modeling Grid
• Rapid Response (applications community)

11
MODIS Level 3 8-Day Daily Composite 1km Active
Fire Standard Product MOD14A1 (Terra) and MYD14A1
(Aqua)
Cape York, Australia (H31V10)

• The MODIS daily Level 3 fire product is
  tile-based, with each product file spanning one
  of the 460 MODIS tiles, of which 326 contain land
  pixels.
• The product is a 1-km gridded composite of fire
  pixels detected in each grid cell over each daily
  (24-hour) compositing period. For convenience,
  eight days of data are packaged into a single
  file.
• Fire white, H20 blue, clouds violet,
  non-fire land grey, "unknown" yellow.

Fire mask for 9 October 2001 from the 8-15
October 2001 daily Level 3 fire product (version
3).
12
MODIS Aqua Daily Global Browse
13
MODIS CMG Product - monthly, 25km, July 2001
14
(No Transcript)
15
Research Tasks for the Proposal Team

• Reviewing the EDR Fire Product Specifications
  from a Climate Research Perspective NASAs ESE
  Fire Needs from NPP VIIRS
• Evaluating the instrument specification
  assessing instrument performance from the
  perspective of fire monitoring
• Assessing the EDR Fire ATBD
• Active Fire Detection
• Active Fire Characterization
• Burned Area Mapping
• Assessing EDR Calibration Data, Quality
  Assessment and Validation plans
• Evaluating the Data Systems for Fire EDR
  generation (IDPS) and dissemination (ADS) and
  documenting suggested improvements
• Identifying improvements needed for a CDR and
  contribute to the design of the CDR Ops Plan
• Advising on Direct Broadcast capabilities for
  fire monitoring

16
NPP Active Fires EDR Spec Aug 2002

• Active surface fires are natural or anthropogenic
  fires. This EDR provides (a) geolocation of the
  pixels in which active fires are detected, (b)
  the sub-pixel average temperature of each active
  fire, and (c) the sub-pixel area of each active
  fire.
• A global, binary "fire/no fire" map is neither
  required nor desired. The products for this
  application are desired during both day and night
  time for clear-sky conditions and within clear
  areas under conditions of broken clouds.
• Units Degrees latitude and longitude for
  geolocation, K for sub-pixel average temperature,
  m2 for active fire area.
• Paragraph Subject
  
  Specified Value
• 40.6.4.1-1 1. At Nadir VIIRS Guarantee 0.75
  km
• 40.6.4.1-2 2. Worst Case VIIRS
  Guarantee 1.6 km
• 40.6.4.1-3 b. Horizontal Reporting Interval
  VIIRS Guarantee HCS
• 40.6.4.1-4 c. Horizontal Coverage VIIRS
  Guarantee Land
• d. Measurement Range
• 40.6.4.1-5 1. Sub-pixel Average Temperature of
  Active Fire 800 K 1200 K
• VIIRS Guarantee
• 40.6.4.1-6 2. Sub-pixel Area of Active Fire
• VIIRS Guarantee From 1000 m2 to 50 m
  Times Ground Sample Distance in Scan
  Direction
• e. Measurement Uncertainty
• 40.6.4.1-7 1. Sub-pixel Average Temperature of
  Active Fire
• VIIRS Guarantee 50 K
• 40.6.4.1-8 2. Sub-pixel Area of Active Fire
  VIIRS Guarantee 30
• 40.6.4.1-9f. Mapping Uncertainty, 3 Sigma VIIRS
  Guarantee 1.5 km

17
Instrument Sensor Performance Issuesrelated to
fire that will be given attention

• Geolocation
• Band to band registration
• High gain band calibration

18
Anticipated Algorithm Issues

• Fire Size and Area most significantly, need to
  review the established uncertainty estimates or
  sensitivity analysis, for the temperature and
  area retrieval.
• Pixel Aggregation could present problems if the
  saturation of the pre-aggregated footprints are
  not properly handled.
• Saturation the proposal team lobbied the IPO to
  increase the saturation level of M13 and M15. The
  failure to increase the saturation level results
  in dependency on different bands for detection
  than are used by MODIS.
• Cloud Mask, Land Water Mask and Sun Glint will
  play an important role in NPP fire detection and
  are needed for meaningful CDR generation.

19
Team Primary Responsibilities

• Justice Team lead and management, CDR
  definition, fire applications community liaison
  (GOFC/GOLD)
• Roy Ops Concept, Data System, QA, Burned Area
• Giglio Instrument performance, simulation,
  active fire algorithm, validation, fire
  characterization
• Csiszar Fire characterization, validation,
  GOFC/GOLD Liaison, NOAA NESDIS liaison

20
Strategic Collaborations

• NASA Collaborations
• Ed Masuoka, MODAPS
• Jacques Descloitres, MODIS Land Rapid Response
• Jeff Morisette, CEOS LPV
• Simon Hook, ASTER JPL
• National Collaborations
• Tom Bobbe USFS RSAC, Utah
• Donna Mcnamara NOAA NESDIS
• Chris Elvidge NOAA NGDC
• International Collaborations (in the framework of
  GTOS GOFC/GOLD-Fire and the CEOS LPV WG ) - ITAR
  issues TBD
• Martin Wooster, Kings College, London
• Dieter Oertel, BIRD, DLR
• MODIS Fire Ground Stations CSIRO, CSIR, RAS,
  IBAMA/INPE, CONABIO
• GOFC/GOLD-Fire Networks (SAFNet, SEAFire,
  Australia)

21
Priority Issues

• Algorithm
• Make sure we have access to the latest version of
  the EDR algorithm
• Early discussion with algorithm developer on
  planned changes
• MODIS lessons learned
• Make sure lessons learned from MODIS have been
  taken into account in the development of VIIRS
• QA
• Operational Production Reprocessing
• Rapid Response System
• Validation
• Archive and Distribution
• Validation
• Early discussion with developer on plans for
  pre-launch algorithm evaluation and post-launch
  product validation
• Direct Broadcast (DB) Stations
• Engage DB-Fire stations early in the discussions
  as to VIIRS capabilities and plans
• Data System

22
Resources Required

• Timely disbursement of ST funds!
• High resolution data from ASTER / BIRD
• Support for a small community outreach workshop
  on VIIRS-Fire (2004)

23
Concerns

• Although fire characterization is the focus of
  the EDR - we DO need global operational active
  fire detection
• Contractor fire area/temperature algorithm
  apparently is non-performing
• MODIS and NPP should be used to prototype the
  data flows for these products
• Value added system capabilities will be needed to
  continue operational data flows developed by
  MODIS e.g., building on the MODIS Rapid Response
  system and the link to the operational users
• How will the NPP active fire EDR be validated ?
• No ASTER-type high resolution sensor on NPP
• Will the data flows to the Science Team be
  adequate to meet our needs ?
• We will be encouraging significant foreign
  participation within the bounds of ITAR
• Burned area algorithm for MODIS is being
  validated regionally we will be advocating
  burned area as a CDR