VIIRS%20Global%20DNB%20Product%20Status

1
VIIRS Global DNB Product Status

• Chris Elvidge
• Earth Observation Group
• NOAA National Geophysical Data Center
• Kim Baugh, Feng-Chi Hsu, Mikhail Zhizhin
• CIRES, University of Colorado
• October 22, 2012

2
Enabling software developed at NGDC

• Reprojection toolkit
• DNB terrain correction
• Stray light filter
• M10 combustion source detection system
• M15 cloud detection relative to NCEP surface
  temperature grids
• Modification of DMSP nighttime lights compositing
  software

3
Stray Light Filter
Raw
Subtracted component
Filtered
4
The First VIIRS Global Cloud-Free Composite April
18-26, 2012 - Low moon nights Output is a 15 arc
second grid (86401 x 33601)
5
The First VIIRS Global Cloud-Free Composite April
18-26, 2012 Washington, DC
VIIRS
DMSP-OLS
6
The First VIIRS Global Cloud-Free Composite April
18-26, 2012 Denver, Colorado
VIIRS
DMSP-OLS
7
The First VIIRS Global Cloud-Free Composite April
18-26, 2012 Delhi, India
VIIRS
DMSP-OLS
8
Issues
Northward Extent 65 Degrees
Residual Stray Light

1. Difficulty setting noise threshold due to
   residual stray light
2. Current hardware lacks memory to geolocate
   aggregates north of 65 degrees
3. Slight westward bias in DNB geolocation relative
   to M10 gas flare detections
4. Low numbers of cloud-free coverages in numerous
   locations
5. Fuzzy lights can be found in some areas

9
Geolocation Bias
Basra Gas Flares, Iraq Red M10 composite Cyan
DNB composite The DNB geolocation is shifted
slightly to the west.
10
Plan to Address Issues

1. Stray Light Make separate composites for stray
   light and non-stray light data. Stray light is
   present for data observed with ground solar
   zenith angles in the range of 90 to 118 degrees.
   Develop separate noise thresholds, remove
   background noise, recombine.
2. Extending North Move processing to computer with
   more memory. This is happening this week!
3. Geolocation Bias Measure the bias between DNB
   and M10 combustion source detections in the
   global cloud-free composites. Introduce bias
   correction as a variable in the reprojection
   toolkit. Reprocess!
4. Expanding the number of cloud-free coverages Add
   data from additional months.
5. Fuzzy lights Develop fuzzy light detector.
   Screen out fuzzy lights in the compositing
   process.

11
Composite with no stray light contaminated data
12
Composite with stray light contaminated data
Composite with stray light filtered data
13
Effect of Stray Light Filtering
14
Use of April cloud-free composite as a reference
for the detection of power outages on November 1,
2012