2007 March 5 1 EUVI Early Observations Jean-Pierre Wuelser James Lemen Markus Aschwanden Nariaki Nitta
Jan 05, 2016
2007 March 5 1
EUVI Early Observations
Jean-Pierre Wuelser
James Lemen
Markus Aschwanden
Nariaki Nitta
2007 March 5 2
Lunar Transit 2007 February 25
2007 March 5 3
2007 March 5 4
2007 March 5 5
Image Compression - ICER
• Almost all EUVI images are ICER compressed on-board– ICER was developed at JPL and used on the Mars Rovers
• The main parameter for ICER is the desired size of the compressed image (in Byte)– SECCHI uses 12 sets of compression parameters, including
• ICER0 : 2 MByte (usually lossless for EUVI images)• ICER4 : 400 kByte• ICER5 : 300 kByte• ICER6 : 200 kByte• ICER7 : 100 kByte
• Choosing compression level essentially involves trading image quality versus image cadence– ICER4 instead of ICER6 means half the image cadence– There is an ongoing debate on how much the EUVI images
should be compressed
2007 March 5 6
ICER Performance - EUVI 195
• Analysis of an EUVI image with various levels of ICER
• Plot shows average compression error as a function of intensity
• Horizontal axis is square root scaled– Poisson noise limit
is a straight line (dotted)
– Top: ICER7– Bottom: ICER4
• All curves are below the single pixel Poisson noise limit
2007 March 5 7
ICER Performance - EUVI 171, 284, 304
• 284 compresses very well• 304 compresses the least, with
ICER7 exceeding the Poisson limit• All wavelengths compress below
the Poisson limit for up to ICER6• Most EUVI observations to date
use ICER6
2007 March 5 8
ICER
Faint areas:
• ICER adjusts spatial resolution to match the noise level by summing faint areas into superpixels
• Areas at the noise level show “blocky” appearance
• Loss of true information is small
• Left to right:– ICER6– ICER4– Lossless
2007 March 5 9
First 3D in 171
2007 March 5 10
2007 March 5 11
Observing Strategies - General Concept
• Synoptic program– Continuous coverage to catch all events– 85 - 90 % of available telemetry– Moderate cadence
• Event buffer program– Observes into ring buffer (“SSR2”)– On-board event detection algorithm on Cor2 images stops
observations when triggered by CME– Ring buffer has 3-4 hour capacity of high cadence observations– Allows for some “well observed” events
• Ongoing discussion on cadence versus compression ratio
2007 March 5 12
Observing Strategies - Sequences (1)
• LMSAL proposed synoptic program:– 171 @ 2.5 min cadence, ICER6– 195, 284, 304 @ 10 min cadence, ICER6– 171, 195, 304 @ ICER4, once every hour– Rationale:
• 171 shows the most structures suitable for 3D reconstruction
• Many transients last less than 10 min, requiring a cadence of at least 2.5 min
• NRL proposed synoptic program:– 195, 304 @ 10 min cadence, ICER4– 171, 284 @ 10 min cadence, ICER5– Rationale:
• ICER4/5 for observing very faint structures far off-limb
2007 March 5 13
Observing Strategies - Sequences (2)
• Event program:– LMSAL:
• 171 @ 50 sec cadence, ICER6• 195, 284, 304 @ 2.5 min cadence, ICER6
– NRL:• 195, 304 @ 2.5 min cadence, ICER4
• Approach for choosing best program:– Take observations with both proposed observing programs– For a limited time (one week?):
• Run NRL program on Ahead• Run LMSAL program on Behind
– Re-evaluate
2007 March 5 14
2007 March 5 15