Simon Oliver

1
MODIS Development Activities at ACRES
Simon Oliver Remote Sensing Science and Strategy
2
Overview

• MOD09 Surface Reflectance Direct Broadcast
  implementation
• SEADAS L1 processor
• Toward SMAC (Simplified Method for Atmospheric
  Correction)
• Vegetation Indices (NDVI / EVI)
• Compositing (1day / n-day)

3
MOD09 DB implementation

• Download from NASA Direct Readout Lab
• note not latest institutional code
• Install SDP Toolkit (provides routines to MOD09)
• GBLAV (surface pressure and temperature) and AVN
  ozone data sets in lieu of GDAS and OZ_Daily
  products for NRT
• MOD021KM, HKM, QKM and MOD03 geolocation - MOD035
  cloud mask is optional
• Requires ozone estimate and forecast data
  preceding and following pass time linear
  interpolation
• Routines for ftp retrieval of these included in
  DB port

4
MOD09 DB implementation Hints

• Direct Broadcast Test data not on DRL
• Implement SEADAS first (easier to integrate than
  IMAPP outputs)
• Integration with SEADAS requires minor software
  patch to metadata check remake
• PCF can be picky and difficult to configure so
  look here first

5
MOD09 DB implementation Progress

• ACRES product run successfully (SEADAS inputs)
• Swath to Grid for re-projection
• Go to production once test processing complete

6
MOD09 Surface Reflectance
7
MOD09 DB implementation Comparison

• MODIS DB Surface Reflectance comparison with
  LPDAAC products

8
SEADAS

• SEADAS4.8 not configured for current ACRES MODIS
  L1B output file content i.e. single L1B file
  output.
• Modl1db version enables 250m, 500m and 1000m
  product separation.
• Installation simple substitution in ACRES IMAPP
  workflow straightforward
• IMAPP Terrain correction bug fixed in SEADAS

9
PDS
SEADAS
CLOUDMASK
Ldope (Mask Creator)
MOD09
NDVI EVI
MRTSwath (project)
Surface Reflectance

• Composite Tool requirements
• Maximum value from stack
• Exclude cloud (Mask1)
• Exclude swath angle gt 45 from nadir
• Include closest to nadir

NDVI EVI
Cloudmask Mask1
Composite Tool 1 day or n days
Swath angle Mask2
Land/Sea Mask3
10
Toward SMAC

• In ACRES MODIS development environment
• University of Wisconsin
• MOD07 Profiles implemented
• MOD04 Aerosol implemented
• MOD05 WVNIR (Water vapour ) implemented
• Integration with DLIs SMAC routine not attempted
  (pre-processing requirements)
• MOD09 development to parallel SMAC (potential for
  application to other ACRES products)

11
Sentinel Developments

• Adding supplementary datasets to current Sentinel
  website
• NDVI and EVI from Rapid Response DB software
  (NASA DRL)
• L1B / MOD09 mosaicking/compositing (currently
  with HDFLook)
• Compositing mechanisms to be developed in-house
  to constrain view angle, remove cloud pixels and
  determine maximum pixel value

12
EVI composite
13
Sentinel Developments - others

• DLI MODIS Night Fire Implementation
• AVHRR / CSIRO hotspot algorithm implementation
  IDL and C