green-1 9/17/2015 green band discussion satellite instrument synergy working group september 2003

23
Green-1 07/04/22 Green Band Discussion Satellite Instrument Synergy Working Group September 2003

Upload: moses-horton

Post on 20-Jan-2016

214 views

Category:

Documents


1 download

TRANSCRIPT

Page 1: Green-1 9/17/2015 Green Band Discussion Satellite Instrument Synergy Working Group September 2003

Green-104/21/23

Green Band Discussion

Satellite Instrument Synergy Working Group

September 2003

Page 2: Green-1 9/17/2015 Green Band Discussion Satellite Instrument Synergy Working Group September 2003

Green-204/21/23

‘Green’ on Satellite Systems

• Research Instruments w/ Green Band– MODIS*– MISR– SeaWiFS– POLDER

• Operational Environmental Satellites– Current

• Meteosat MSG does not • Other GEO’s do not

– Future• NPOESS VIIRS will have Green• Meteosat MTG calls for Green

• Operational Land Resources Satellite– LANDSAT has Green

Note: One of the most requested products from the 36-band MODIS is the RGB images

Page 3: Green-1 9/17/2015 Green Band Discussion Satellite Instrument Synergy Working Group September 2003

Green-304/21/23

‘Green’ – Phenomenology

• Maximum in human eye’s spectral response

• Peak in solar spectral irradiance arriving at Earth’s surface

• A primary color in (R-G-B) ‘true-color’ composite imagery

Page 4: Green-1 9/17/2015 Green Band Discussion Satellite Instrument Synergy Working Group September 2003

Green-404/21/23

The Human Vision(peaks at green)

Page 5: Green-1 9/17/2015 Green Band Discussion Satellite Instrument Synergy Working Group September 2003

Green-504/21/23

Solar Irradiance from and through the Top of the Atmosphere

Effects of the AtmosphereUS Atmosphere, 5, 23 and 50 km Visibility

0.00

0.05

0.10

0.15

0.20

0.25

0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4

Wavelength (m)

Irra

dia

nce

(W

/cm2 -

m)

TOA

No Aerosols

Vis 50 km

Vis 23 km

Vis 5 km

Page 6: Green-1 9/17/2015 Green Band Discussion Satellite Instrument Synergy Working Group September 2003

Green-604/21/23

Effects of the AtmosphereUS Atmosphere, 5 and 23 km Visibility

0.00

0.05

0.10

0.15

0.20

0.25

0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4

Wavelength (m)

Irra

dia

nce (W

/cm2 -

m

)

TOA

Sfc No Aerosols

Sfc Vis 23 km

Sfc Vis 5 km

Solar Irradiance from and through the Top of the Atmosphere

Page 7: Green-1 9/17/2015 Green Band Discussion Satellite Instrument Synergy Working Group September 2003

Green-704/21/23

ABI ‘Green’ Band Arguments

• ‘True-Color’ Imagery– (For) Readily incorporated into R-G-B imagery– (Against) Can approximate G from R, B, and NIR

• Ocean ‘Color’– (For) Cannot use other-color surrogate– (Against) HES and VIIRS will provide quantitative product

• Aerosol Properties– (For) Standard reference wavelength is green (0.55 m)– (Against) Most heritage retrievals do not use green

• Smoke Depiction– (For) Quantitative smoke product possible – (Against) Red/blue provides better contrast with green vegetation

• Fire Product: Quick look imagery– (For) ‘True-color‘ readily understood– (Against) Might be good enough synthesized via LUT, etc. method

Active, smoldering, burn-scar better with NIR/SWIR

• Snow & Ice Extent– (For) Brightest surface– (Against) Contrast better with NIR

Page 8: Green-1 9/17/2015 Green Band Discussion Satellite Instrument Synergy Working Group September 2003

Green-804/21/23

RGB: True and Synthetic Color(It is all about visual)

• Justification for replacing G with 860 nm – NDVI uses 860 and 680 nm (chlorophyll based), not 550 nm (relatively

low reflection)

• Success criteria for “Visual” – Green vegetation– White clouds, snow– Bright whitish sand, desert– Aqua (blue/green) ocean– Correct representation for smoke, fire– Grey culture features

• Additionally, the color combination (weights) has to be “universally set”, continuous re-adjusting is not possible in real applications

Page 9: Green-1 9/17/2015 Green Band Discussion Satellite Instrument Synergy Working Group September 2003

Green-904/21/23

Three examples follow

1. Florida Coast• Land, ocean, clouds

2. Narragansat Bay• Vegetation, culture feature, water

3. Middle East/Persian Gulf• Desert, dust storm, water

Page 10: Green-1 9/17/2015 Green Band Discussion Satellite Instrument Synergy Working Group September 2003

Green-1004/21/23

R-G-B

R-(R+B)/2-B

Replacing Green with Average Red and Blue(Tim Schmit)

Page 11: Green-1 9/17/2015 Green Band Discussion Satellite Instrument Synergy Working Group September 2003

Green-1104/21/23

This is as good as can be reproduced, since the image used the 0.55 um from the same image to build the LUT.

Page 12: Green-1 9/17/2015 Green Band Discussion Satellite Instrument Synergy Working Group September 2003

Green-1204/21/23

Page 13: Green-1 9/17/2015 Green Band Discussion Satellite Instrument Synergy Working Group September 2003

Green-1304/21/23

RGB True Color Image(Narragansat Bay, AVIRIS Data)

Page 14: Green-1 9/17/2015 Green Band Discussion Satellite Instrument Synergy Working Group September 2003

Green-1404/21/23

G= (R+B)/2

RGB Image without the Green Band

Page 15: Green-1 9/17/2015 Green Band Discussion Satellite Instrument Synergy Working Group September 2003

Green-1504/21/23

G= NIR

RGB Image without the Green Band

Page 16: Green-1 9/17/2015 Green Band Discussion Satellite Instrument Synergy Working Group September 2003

Green-1604/21/23

G= NIR/5, if NIR/5> (R+B)/2 and G=(R+B)/2, if NIR/5<(R+B)/2

RGB Image without the Green Band(A more sophisticated color combination)

Page 17: Green-1 9/17/2015 Green Band Discussion Satellite Instrument Synergy Working Group September 2003

Green-1704/21/23

RGB Image with and without the Green Band(Middle East/Persian Gulf, MODIS Data)

G=G G=(R+B)/2

Page 18: Green-1 9/17/2015 Green Band Discussion Satellite Instrument Synergy Working Group September 2003

Green-1804/21/23

RGB Image without the Green Band

G=NIR G=NIR/5 if NIR/5>(R+B)/2, else G=(R+B)/2

Page 19: Green-1 9/17/2015 Green Band Discussion Satellite Instrument Synergy Working Group September 2003

Green-1904/21/23

Observations

• Green band substitution requires some level of algorithm development – Simple substitution with NIR, or average of red and blue does

not work for all cases

• Combined use of NIR, blue and red shown to work for a variety of backgrounds

Page 20: Green-1 9/17/2015 Green Band Discussion Satellite Instrument Synergy Working Group September 2003

Green-2004/21/23

Dust Enhancement Technique

Reference:

Steve Miller, NRL-Monterey, AMS Conference 2003

Challenge:

Detection of dust storm over desert

Approach:

Utilize normalized difference between NIR and blue

Red Gun = (RNIR-RBLUE)/ (RNIR+RBLUE)

Data: MODIS

Page 21: Green-1 9/17/2015 Green Band Discussion Satellite Instrument Synergy Working Group September 2003

Green-2104/21/23

Dust Enhancement Product Red Gun = (RNIR-RBLUE)/ (RNIR+RBLUE)

Green gun = Green Green gun =(R+B)/2

Page 22: Green-1 9/17/2015 Green Band Discussion Satellite Instrument Synergy Working Group September 2003

Green-2204/21/23

Dust Enhancement ProductG=(R+B)/2

Red Gun = (RNIR-RBLUE)/ (RNIR+RBLUE)Red Gun = Red

Page 23: Green-1 9/17/2015 Green Band Discussion Satellite Instrument Synergy Working Group September 2003

Green-2304/21/23

Summary

• Green is a highly desirable band– High demand for RGB imagery key reason

• However, it is not crucial for ABI because 1. For some applications, another band or band combinations will do

(algorithm development required)2. For some applications, other sensors either from HES or VIIRS will do  

• Looking at satellite instrument synergy opportunities, if ABI had green band, it could be more effective in helping to task HES-CW

• Looking at the trend of future remote sensing systems, it is strongly recommend that the green band be considered– as “goal” for ABI– for systems beyond GOES-R