ESM 266: Multispectral remote sensing

1
ESM 266 Multispectral remote sensing
2
Radiation and remote sensing

• For example, if energy being sensed comes from
  the Sun, it
• is radiated by atomic particles at the source
  (the Sun)
• propagates through the vacuum of space at the
  speed of light
• interacts with Earth's atmosphere
• interacts with Earth's surface
• interacts with Earth's atmosphere once again, and
  
• finally reaches our sensing system
• where it interacts with various optical systems,
  filters, emulsions, or detectors

3
Atmospheric absorptionand scattering
emission
absorption
scattering
4
Relation between frequency and wavelength
5
(No Transcript)
6
Sun glint and wildfires near Carpenteria Bay,
Queensland, Australia
7
Terminology

• Radiant flux F, units W
• Irradiance (flux density) E, units Wm2
• (called Exitance M when away from surface)
• Radiance L, units Wm2sr1
• Note All can be functions of wavelength and have
  units µm1

8
Reflectance terminology
9
Relation between flux density and intensity
If L is isotropic (same for all ?r??? r), then
? r
? r
10
Steradian (Solid Angle)

• The steradian (sr) is the unit of solid angle
• defined as W A/r2

Spherical cap area, A
Radius, r
W
Diameter of flat cone
Sphere
11
Overview of typical system
12
Format of a multispectral image
13
Detector configurations breaking up the spectrum
14
Characteristics of several sensors
15
Spatial, spectral characteristics of some
multispectral sensors
16
Suns radiant energy distribution
17
Landsat ETM spectral bands
18
Landsat Thematic Mapper bands and their uses

• Band 1 (Blue 0.45 - 0.52 mm)
• good water penetration
• differentiating soil and rock surfaces from
  vegsmoke plumes
• most sensitive to atmospheric haze
• Band 2 (Green 0.52 - 0.60 mm)
• water turbidity differences
• sediment and pollution plumes
• discrimination of broad classes of vegetation
• Band 3 (Red 0.63 - 0.69 mm)
• strong chlorophyll absorption (veg. vs. soil)
• urban vs. rural areas

19
Landsat Thematic Mapper bands and their uses

• Band 4 (NIR1 0.76 - 0.90 mm)
• different vegetation varieties and conditions
• dry vs. moist soil
• coastal wetland, swamps, flooded areas
• Band 5 (NIR2 1.55 - 1.75 mm)
• leaf-tissue water content
• soil moisture
• snow vs cloud discrimination
• Band 6 (Thermal 10.4 - 12.5 mm)
• heat mapping applications (coarse resolution)
• radiant surface temperature range -100oC to
  150oC
• Band 7 (NIR3 2.08 - 2.35 mm)
• absorption band by hydrous minerals (clay, mica)
• lithologic mapping (clay zones)

20
Landsat 7 orbits and acquisitions, 4/30/2005
Converter between path/row and lat/lon
21
ASTER Advanced Spaceborne Thermal Emission and
Reflection Radiometer

• Launched on Terra 12/18/1999
• 4 VNIR bands
• 15m resolution, 24º crosstrack pointing
• 6 SWIR bands
• 30m resolution, 8.5º crosstrack pointing
• 5 thermal bands
• 60m resolution, 8.5º crosstrack pointing
• See ASTER website for details

22
(No Transcript)
23
MODIS Moderate Resolution Imaging
Spectroradiometer

• Launched on Terra 12/18/1999, on Aqua 3/24/2002
• Terra crosses Equator 1030 am, Aqua 130 pm
  (orbit time is 98 min)
• 36 spectral bands with resolutions of 250m, 500m,
  and 1km
• http//modis.gsfc.nasa.gov/about/specifications.ph
  p
• Global coverage 1-2 days depending on latitude
• Pointing, 55 from nadir

24
Spectral signatures

• The spectral signature of an object in an image
  involves
• spectral BRDF of the object, i.e. fr(?, angles)
• (and other objects in the IFOV)
• (may change w time, e.g. vegetation in summer vs
  winter)
• the sensors bands, wavelengths, spectral and
  radiometric resolution, dynamic range
• scattering and absorption by the atmosphere in
  those bands
• amount of energy available (Sun or emitted) in
  those bands
• (A combination of physical and sensor
  characteristics)

25
Atmospheric spectrum
26
Snow, vegetation, rock spectra of mixed pixels
27
Ice dam on Russell Fjord by Hubbard Glacier
8/7/1985 Landsat 5 TM bands 4 3 2
9/11/1986
28
Ice shelf disintegration from MODIS

• Direct response to regional warming

2000 km2 of Larsen Ice Shelf disintegrated in 2
weeks.
Glaciers feeding lost ice shelves accelerate (up
to 8X)? ice-shelf buttressing
29
ASTER Erupting Mt Usu, Hokkaido
VNIR bands 4/3/2000
Thermal bands 4/5/2000
http//asterweb.jpl.nasa.gov/gallery/
30
MODIS Fires in southeastern U.S.
Acquired 4/1/2003, MODISs image of the day for
4/7/2003
31
Spectra with 7 MODIS land bands (250-500m
resolution, global daily coverage)
32
Simple (snow/no-snow) algorithm for TM bands on
MODIS

• Normalized difference snow index
• Thresholds for MODIS algorithm (Hall)
• If TMband 4gt11
• and NDSIgt0.4
• Pixel is gt50 snow covered

MODIS, Sierra Nevada, 4/29/2000
33
Example vegetation near Hudson Bay
1996 Landsat TM bands 4 3 2 (RGB)
http//edcwww.cr.usgs.gov/earthshots/slow/tableofc
ontents
34
Vegetation (NDVI) from Landsat image
35
Confluence of Tigris and Euphrates, Iraq
2/21/1997 Landsat TM bands 4 2 1 (RGB)
36
MODIS Snow in northeastern U.S.
2/20/2003
http//visible.earth.nasa.gov/
37
Data rate
38
Data rates of some sensors
39
Dynamic Range
Saturation
Ideal Response (offset for clarity)
Image Brightness
Actual Sensor Response
Dark Current Signal
Scene Brightness
40
MISR Multi-angle Imaging SpectroRadiometer

• Launched on Terra, 12/18/1999
• Exploits angular information in signal
• Surface BRDF
• Path lengths through atmosphere
• 4 spectral bands blue, green, red, near-infrared
• 9 angles 0º, fore and aft 26.1º, 45.6º, 60.0º,
  70.5º
• Global coverage 2-9 days depending on latitude
• http//www-misr.jpl.nasa.gov/mission/minst.html

41
MISR Aerosol optical depth over central/eastern
Europe
February March 2003
42
Spectral solar irradiance
43
QuickBird Commercial imagery at 0.6m resolution
panchromatic, 2.44m multispectral
http//www.digitalglobe.com
44
NASAs Earth Observing System (EOS) Idea of
standard products

• Landsat, AVHRR, most earlier missions provide, at
  best, calibrated radiances at satellite
• Interpretation of these radiances to estimate
  geophysical and biological information requires
  arcane knowledge and difficult processing
• Need to understand relationship between
  electromagnetic properties of surface of interest
  and the properties of real interest
• Need to account for effects from atmosphere and
  sensor
• Need to deal with sensor geometry to provide
  information on a map base, so that you can
  integrate with other spatial information
• Therefore, EOS provides information, in addition
  to data

45
MODIS the canonical (but not only) example

• http//modis.gsfc.nasa.gov/data/dataprod/index.php
  
• Each has a nontechnical description, information
  about the data, and an ATBD (algorithm
  theoretical basis document)
• Levels of processing
• Calibrated radiances and geolocation information
• Geophysical products
• Mapped to a projection

46
MODIS Leaf Area Index
September 2000
(from Boston University)
December 2000
April 2001
0
3
6
LAI is defined as the one sided green leaf area
per unit ground area in broadleaf canopies and as
the projected needle leaf area in coniferous
canopies.
47
MODIS FPAR(Fraction of absorbed
Photosynthetically Active Radiation)
September 2000
December 2000
April 2001
0
0.9
0.5
FPAR is defined as the fraction of incident
photosynthetically active radiation (0.4 - 0.7
?m) absorbed by the vegetation canopy.
48
MODIS ocean chlorophyll, Indian Ocean