Title: Satellite observations of the atmosphere and the ocean surface Heraeus Summer School
1Satellite observations of the atmosphere and the
ocean surface Heraeus Summer School Physics of
the Environment
- Andreas Richter
- Institute of Environmental Physics
- University of Bremen
- tel. 49 421 218 4474
- e-mail richter_at_iup.physik.uni-bremen.de
- http//www.iup.physik.uni-bremen.de/doas
2Lecture Contents
- What is Remote Sensing?
- Which Quantities can be Measured?
- What are the Underlying Physical Principles?
- Examples
- Tropospheric Aerosols
- Stratospheric Ozone
- Tropospheric NO2
- Stratospheric Aerosols
- Temperature Profiles
- Wind Speed and Direction
- Sea Surface Temperature
- Sea Ice
- Summary
3What is Remote Sensing?
- Remote sensing is the science and art of
obtaining information about an object, area, or
phenomenon through the analysis of data acquired
by a device that is not in contact with the
object, area, or phenomenon under investigation
(Lillesand and Kiefer 1987) - The art of dividing up the world into little
multi-coloured squares and then playing computer
games with them to release unbelievable potential
that's always just out of reach. (Jon
Huntington, Commonwealth Scientific and
Industrial Research
Organisation Exploration,
Geoscience, Australia)
4The Eye as a Remote Sensing Instrument
- eye remote sensing instrument in the visible
wavelength region (350 - 750 nm) - signal processing in the eye and in the brain
- colour (RGB) and relative intensity are used to
identify surface types - large data base and neuronal network used to
derive object properties
5The Eye as a Remote Sensing Instrument
- eyes are scanning the environment with up to 60
frames per second - 170 field of view, 30 focus
6The Eye as a Remote Sensing Instrument
- stereographic view, image processing, and a large
data base enables detection of size, distance,
and movement
7The Eye as a Remote Sensing Instrument
- passive remote sensing instrument, relies on
(sun) light scattered from the object - no sensitivity to thermal emission of objects
?
8-14 microns image of a cat
8The Eye as a Remote Sensing Instrument
- active remote sensing by use of artificial light
sources
?
9Why should we use Remote Sensing?
- not all measurement locations are accessible
(atmosphere, ice, ocean) - remote sensing facilitates creation of long time
series and extended measurement areas - for many phenomena, global measurements are
needed - remote sensing measurements usually can be
automated - often, several parameters can be measured at the
same time - on a per measurement basis, remote sensing
measurements usually are less expensive than
in-situ measurements
10Why NOT to use Remote Sensing
- remote sensing measurements are always indirect
measurements - the electromagnetic signal is often affected by
more things than just the quantity to be measured - usually, additional assumptions and models are
needed for the interpretation of the measurements - usually, the measurement area / volume is
relatively large - validation of remote sensing measurements is a
major task and often not possible in a strict
sense - estimation of the errors of a remote sensing
measurement often is difficult
11Schematic of Remote Sensing Observation
12Classification of Remote Sensing Techniques
- active / passive
- platform
- wavelength range
- spectral resolution
- low / medium / high
- spatial resolution
- low / high
- detection technique
- absorption, emission or extinction spectroscopy
- spectral reflectance
13Active vs. Passive Remote Sensing
- Active Remote Sensing
- Artificial source of radiation, the reflected or
scattered signal is analysed - sound SONAR
- radio waves RADAR (RAdio Detection And Ranging)
- laser light LIDAR (LIght Detection And Ranging)
- white light long path DOAS (Differential Optical
Absorption Spectroscopy) - Passive Remote Sensing
- Natural sources of radiation, the attenuated,
reflected, scattered, or emitted radiation is
analysed - solar light
- lunar light
- stellar light
- thermal emission
14Remote Sensing Platforms
- ground-based measurements
- continuous, high accuracy, easy accessibility
- local measurement
- air-borne measurements (up to 15 km)
- fast moving, long distance
- expensive, sporadic
- sonde / balloon measurements (up to 30 km)
- high altitude
- logistically difficult, often expensive
- rocket measurements (up to 80 km)
- very high altitude
- expensive, sporadic
- Space Shuttle / Space Station measurements
- global coverage, limited time coverage, good
accessibility - satellite measurements
- global coverage
- poor accessibility, expensive
15Wavelength Ranges in Remote Sensing
- UV some absorptions profile information
- aerosols
- vis surface information (vegetation)
- some absorptions
- aerosol information
-
- IR temperature information
- cloud information
- water / ice distinction
- many absorptions / emissions
- profile information
- MW no problems with clouds
- ice / water contrast
- surfaces
- some emissions profile information
16Which Quantities are Measured?
- absolute intensities in dedicated wavelength
intervals - intensities relative to the intensity of a
reference source - ratios of intensities at different wavelengths
- variations of intensities
- degree of polarisation
- phase and delay of signal
17Which Quantities can be Determined?
- Surface
- height
- albedo
- vegetation type
- surface (water) temperature
- fires
- surface roughness
- wind speed
- water turbidity / chlorophyll concentrations
- ice cover
- ice type
- Meteorology
- pressure
- temperature
- cloud cover
- cloud top height
- cloud type
- lightning frequency
- Chemical constitution of the atmosphere
- aerosol burden
- aerosol type
- trace species
18The Electromagnetic Spectrum
- nearly all energy on Earth is supplied by the sun
through radiation - wavelengths from many meters (radio waves) to nm
(X-ray) - small wavelength high energy
- radiation interacts with atmosphere and surface
- absorption (heating, shielding)
- excitation (energy input, chemical reactions)
- re-emission (energy balance)
19Radiative Transfer in the Atmosphere
- Contributions
- Direct Solar Ray
- Reflection on the Surface
- Reflection from Clouds
- Scattering in the Atmosphere
- Rayleigh Scattering
- Mie Scattering
- Raman Scattering
- Absorption in the Atmosphere
- Emission in the Atmosphere
- Emission from the Surface
- Emission from Clouds
20Radiative Transfer in the Atmosphere
Atmosphere
Absorption
Scattering
Emission
Aerosol / Molecules
21Scattering in the Atmosphere
- Depending on the ratio of the size of the
scattering particle (r) to the wavelength (?) of
the radiation - Mie parameter ? 2? r / ?,
- different regimes of atmospheric scattering can
be distinguished.
gt different wavelengths probe different parts of
the atmosphere / surface
22What is the Optimal Instrument?
- A compromise must be found to get the optimum
amount of information out of the limited number
of photons available under the given boundary
conditions
- instrument size and price
- satellite orbit
- measurement quantity
- data rate
- measurement error
spatial coverage
spatial resolution
time resolution
vertical resolution
time coverage
spectral resolution
spectral coverage
23Satellite Orbits
- (Near) Polar Orbit
- orbits cross close to the pole
- global measurements are possible
- low earth orbit LEO (several 100 km)
- ascending and descending branch
- special case sun-synchronous orbit
- overpass over given latitude always at the same
local time, providing similar illumination - for sun-synchronous orbits day and night
branches - Geostationary Orbit
- satellite has fixed position relative to the
Earth - parallel measurements in a limited area from low
to middle latitudes - 36 000 km flight altitude, equatorial orbit
http//www2.jpl.nasa.gov/basics/bsf5-1.htm
http//www.ccrs.nrcan.gc.ca/ccrs/learn/tutorials/f
undam/chapter2/chapter2_2_e.html
24How can Vertical Information be Derived?
- In many atmospheric application, vertical
profiles of quantities are needed. - Approaches
- Vertical Scanningsequential of parallel
measurements at different altitudesgt e.g.
SCIAMACHY limb profiles - Pressure / Temperature dependence of signal (e.g.
line shape)inversion of signal using a priori
information on e.g. vertical p-profilegt e.g.
microwave sounding - Saturation Effects at different wavelengths
(frequencies)using spectral regions with
different penetration depthsgt e.g. SBUV ozone
profile measurements - Time Resolved measurementsusing pulsed signals
and photon flight time informationgt e.g. LIDAR - Combination of different types of measurements,
instruments or modelsgt e.g. GOME tropospheric
NO2 measurements
25How can Vertical Information be Derived?
- In many atmospheric application, vertical
profiles of quantities are needed. - Approaches
- Vertical Scanningsequential of parallel
measurements at different altitudes
Nadir observation of scattered and reflected
light, total column determination (and O3
profile), good spatial resolution, global
coverage, good SNR Limb observation of
scattered light, stratospheric and upper
atmosphere profiles, poor spatial resolution,
near global coverage, SNR decreases with
altitude Occultation direct observation of sun
or moon at horizon, stratospheric profiles, poor
spatial resolution, limited coverage (close to
terminator), high SNR but low UV
sensitivity Limb Nadir Matching combination of
nadir and limb measurements to estimate the
tropospheric column of a trace gas
http//www.sciamachy.de
26How can Vertical Information be Derived?
- In many atmospheric application, vertical
profiles of quantities are needed. - Approaches
- Pressure / Temperature dependence of signal (e.g.
line shape)
pressure broadening
inversion
http//www.ram.uni-bremen.de/index_ram.html
27How can Vertical Information be Derived?
- In many atmospheric application, vertical
profiles of quantities are needed. - Approaches
- Saturation Effects at different wavelengths
(frequencies)Example ozone profiling in the UV
(e.g. SBUV, GOME)Ozone absorption is increasing
by orders of magnitude over 50 nm in the UV, and
virtually no photons reach the surface below 300
nm. By measuring ozone at different wavelengths,
different sub-columns are determined gt profile
28How can the desired signal be isolated?
- In most measurements, several effects on the
signal interfere and need to be corrected. - Example retrieval of NO2 by UV/vis absorption
spectroscopy of scattered sun light - NO2 absorption
- absorption by other species (O3, O4, H2O, ...)gt
use of measurements at many wavelengths and
characteristic absorption spectrum for correction - colour of the surface (e.g. ocean colour)gt use
of measurements at many wavelengths and
characteristic absorption spectrum for
correction - scattering by aerosols
- gt fit of broad band contribution
- elastic scattering by air moleculesgt fit of
broad band contribution - inelastic scattering by air molecules
- gt explicit correction by modelling the effect
- gt in many cases, measurements at several
wavelengths / frequencies help
29Validation of Remote Sensing Measurements
- Remote Sensing measurements are indirect
measurements, and need validation! - The perfect validation measurements should
- measure the same quantity
- integrate over the same volume
- measure at the same time
- use an independent technique
- have higher accuracy and precision than the
measurement to be validated - cover a large range of geophysical conditions
- have no location bias such as measurements
- only over land,
- only during clear weather or
- mostly in the Northern Hemisphere
- not be too expensive
- gt such measurements do usually not exist!
30Problems for Validation
- Example Stratospheric NO2 measurements from
SCIAMACHY - Amount of data SCIAMACHY provides about 150 000
NO2 measurements per day or more than 50 000 000
measurements per year. To validate even a small
part of these data necessitates a large number of
validation measurements - Global coverage hardly any validation
measurements are truly global in coverage but
usually biased over land in NH mid-latitudes - Averaging volume even a small SCIAMACHY ground
pixel is 30 x 60 km2 large and at high sun
vertically integrated over the whole atmosphere.
Sampling this volume at 3 km resolution
horizontally and vertically (up to 20 km) would
take many hours in an aircraft. - Inhomogeneity in time and space many validation
measurements do not coincide exactly in time and
space with the remote sensing measurement.
Horizontal variability as well as changes over
time often are the largest uncertainty in
validation - Errors of validation measurements validation
measurements often have themselves relatively
large random and systematic errors, in particular
if they are remote sensing measurements (example
neglect of temperature dependence of ozone
cross-section in Brewer measurements,
interference by PAN and other compounds with
in-situ NO2 measurements, pump rate problems at
high altitudes in ozone-sonde measurements, ...)
31Validation Example
- Example
- Validation of SCIAMACHY NO2 total columns with
ground-based DOAS zenith-sky measurements - Results
- validation at several stations (latitudes)
- validation of complete seasonal cycle
- comparable measurement volume
- good agreement
- Problems
- ground-based measurements AM / PM twilight,
SCIAMACHY at 1000 LT - zenith-sky measurements not sensitive to
tropospheric pollution - zenith-sky measurement is also remote sensing
measurement, not truly independent technique
32LIDAR Measurements of tropospheric aerosols
- Target Quantity Tropospheric aerosol
concentrations - Measurement Quantity Backscatter ratio at 532 nm
and time lag - Instrument type LIDAR
- Instrument LITE on Space Shuttle, September 1994
33LIDAR (LIght Detection And Ranging)
- Idea
- Use of an active system that emits light pulses
and measures the intensity of the backscattered
light (from air molecules, aerosols, thin clouds)
as a function of time (optical Radar) - Instrument
- a strong laser with short pulses
- possibly several wavelengths emitted
- a large telescope to collect the weak signal
- Measurement quantity
- time lag gives altitude information
- signal intensity gives information on
backscattering at given altitude and extinction
along the light path - measurements at different wavelengths provide
information on absorbers and aerosol types - polarisation measurements provide information on
phase of scatterers - gt Very good vertical resolution can be achieved!
34Lidar In-space Technology Experiment (LITE)
- Instrument
- flashlamp-pumped NdYAG laser
- 1064 nm, 532 nm, and 355 nm
- 1-meter diameter lightweight telescope
- PMT for 355 nm and 532 nm avalanche photodiode
(APD) for 1064 nm - Mission Aims
- test and demonstrate lidar measurements from
space - collect measurements on
- clouds
- aerosols (stratospheric tropospheric)
- surface reflectance
- Operation
- on Discovery in September 1994 as part of the
STS-64 mission - 53 hours operation
http//www-lite.larc.nasa.gov/index.html
35LITE Example of Aerosol Measurements
Clouds (ITCZ)
Atlas mountains
complex aerosol layer
maritime aerosol layer
- 5 minutes of LITE data over the Sahara
- low maritime aerosol layer
- high complex aerosol layer over Sahara
- Atlas Mountains separate two regimes
- clouds close to the ITCZ
http//www-lite.larc.nasa.gov/index.html
36UV absorption measurements of stratospheric O3
- Target Quantity Stratospheric Ozone columns
- Measurement Quantity Differential absorption of
backscattered UV radiation - Instrument type low resolution nadir viewing UV
spectrometer - Instrument TOMS (Total Ozone Mapping
Spectrometer )
37Total Ozone Mapping Spectrometer TOMS
- Idea
- global measurements of ozone columns using
differential measurements in the UV - good spatial resolution through fast measurements
- additional products (SO2, aerosols) by clever
selection of wavelengths - continuous measurements, long time series, high
consistency, little changes in instrumentation gt
trends - The TOMS programme
- Satellite Period Orbit
- Nibus 7 Oct 78 May 93 955 km
- Meteor3 Aug 91 Dec 94
- Adeos Aug 96 Jun 97 830 km
- Earth Probe (EP) Jul 96 Dec 97 500 km
- Dec 97 today 740 km
- Wavelengths
- 380.0 339.7 331.0 317.4 312.3 308.6 nm
http//jwocky.gsfc.nasa.gov/
38TOMS Observation of the Ozone Hole
- The Ozone Hole
- forms in the Antarctic winter / spring
- formation of Polar Stratospheric Clouds PSC in
the extremely cold vortex - heterogeneous activation of chlorine reservoirs
on the cold PSC surfaces - rapid ozone destruction by ClO and BrO as the sun
rises - end of ozone destruction after warming when
chlorine is transformed back to its reservoirs
HCl and ClONO2 and vortex air mixes with ozone
rich air
http//jwocky.gsfc.nasa.gov/
39UV/vis absorption measurements of tropospheric NO2
- Target Quantity Tropospheric Nitrogen Dioxide
columns - Measurement Quantity Differential absorption of
backscattered radiation - Instrument type medium resolution nadir viewing
UV/vis spectrometer - Instrument GOME (Global Ozone Monitoring
Experiment) on ERS-2
40Global Ozone Monitoring Experiment (GOME)
- Idea
- simultaneous measurements from the UV to the near
IR - good spectral resolution (0.2 0.4 nm)
- use of DOAS to retrieve columns of several
species (O3, NO2, OClO, BrO, HCHO, SO2, H2O) - use of UV wavelengths to retrieve ozone profiles
- global coverage
- Launch April 1995 on ERS-2 (sun synchronous)
- GOME successor instruments
- Instrument Satellite Launch
- SCIAMACHY ENVISAT March 2002
- OMI EOS-Aura Spring 2004
- GOME-2 Metop-1 .. Metop-3 2006 2020
http//www.iup.physik.uni-bremen.de/gome/
41GOME tropospheric NO2 excess
- NOx plays a key role in the formation of
photochemical ozone smog - sources of NOx are both anthropogenic (combustion
of fossil fuels, biomass burning) and natural
(fires, soil emissions, lightning) - NOx emissions are changing as result of
- changes in land use
- improvements in emission control
- economic development (e.g. China)
- GOME data provided the first global maps of
tropospheric NO2
- Data analysis
- cloud screening
- DOAS retrieval of total slant columns
- subtraction of clean Pacific sector to derive
tropospheric slant columns - application of tropospheric airmass factor to
compute tropospheric vertical column
42UV/vis Measurements of Stratospheric Aerosols
- Target Quantity stratospheric aerosol
concentrations - Measurement Quantity backscattered radiation
- Instrument type solar occultation viewing UV/vis
spectrometer - Instrument SAGE-2 (Stratospheric Aerosol and Gas
Experiment)
43Stratospheric Aerosol and Gas Experiment (SAGE)
- Measurement Geometry solar occultation
- Instrument grating spectrometer with
Si-detectors - Spectral coverage 7 wavelengths between 385
1020 nm - 1020, 940, 600, 525, 453, 448 und 385 nm
- Data analysis onion peeling
- Measurement targets vertical profiles of O3,
NO2, H2O and aerosol extinction (at 385, 453,
525 and 1020 nm) - Measurement range stratosphere, at low
stratospheric aerosol loading and outside the
tropics also the upper troposphere - The SAGE programme
- SAM II 1978
- SAGE I 1979-1981
- SAGE II 1984 - today
- SAGE III 2001 - today
- 280 1030 nm, 1-2 nm spectral
resolution CCD detector, lunar solar
occultation
http//www-sage3.larc.nasa.gov/
44SAGE Stratospheric Aerosols
- Stratospheric aerosols are dominated by volcanic
input (H2SO4). - Large eruptions inject ash and SO2 directly into
the stratosphere. - Transport towards poles within one year.
- Exponential decay over many years
- 1985 Nevado del Ruiz, Columbia
- 1990 Kelut, Indonesia
- 1991 Mt. Pinatubo
http//aerosols.larc.nasa.gov/optical_depth.html
45Radio Occultation Measurements of Temperature
Profiles
- Target Quantity temperature profiles
- Measurement Quantity excess phase of GPS signals
- Instrument type GPS occultation
- Instrument CHAMP (CHAllenging Minisatellite
Payload)
46CHAMP radio occultation
- Principle
- GPS receiver observes GPS satellite during
occultation - high accuracy time information provides excess
phase - this is related to the bending angle profile a
- which depends on refractive index n
- which is a function of p, T and humidity
- good vertical resolution
- large number of measurements
- good sampling
- assumptions on 2 of the three variables necessary
- problems with critical layers
http//www.copernicus.org/EGU/acp/acpd/4/7837/acpd
-4-7837_p.pdf
47QBO Temperature Anomalies from CHAMP Radio
Occultation
- downward propagation of temperature anomalies in
the tropical stratosphere - QBO (Quasi Biannual Oscillation) signal
- maximum amplitude of /- 4.5 K
- impact on stratospheric ozone columns
http//www.copernicus.org/EGU/acp/acpd/4/7837/acpd
-4-7837_p.pdf
48Microwave Measurements of Wind Speed and Direction
- Target Quantity wind speed and direction
- Measurement Quantity reflected microwave
intensity and polarisation - Instrument type active microwave
- Instrument Synthetic Aperture Radar (SAR).
49How to derive wind speed from Radar signals
- Idea Bragg-like resonance of cm-size ocean
waves with Radar signals depends monotonically on
surface wind speed - gt wind speed over oceans can be determined from
scatterometer measurements if wind direction is
known from model or other measurements
Validation
Relationship between radar backscatter and
surface wind speed for C-band (5.3 Hz),vertical
polarization at 45 off nadir angle.
http//fermi.jhuapl.edu/sar/stormwatch/user_guide
/bealguide_072_V3.pdf
50Wind Speed from Radarsat SAR
Polar low of 05 Feb 1998 after application of
wind algorithm, embedded in NOGAPS model wind
field (arrows).
Polar low imaged by 430 km wide swath mode
of Radarsat SAR, before application of wind
algorithm, 0602 GMT 05 Feb 1998.
http//fermi.jhuapl.edu/sar/stormwatch/user_guide/
bealguide_072_V3.pdf
51Passive Microwave Measurements of Sea Ice
- Target Quantity sea ice coverage and type
- Measurement Quantity reflected microwave
intensity and polarisation - Instrument type passive microwave radiometer
- Instrument AMSR-E (Advanced Microwave Scanning
Radiometer - EOS )
- 12 channels and 6 frequencies ranging from 6.9 to
89.0 GHz - two polarisations
52Sea Ice Maps from AMSR-E
- Basic principle
- strong contrast in thermal microwave emission
between ice and open ocean - assumption of linear relationship between
brightness and ice cover - parameters
- sea ice concentration,
- surface ice temperature,
- snow depth on ice
- ice type by frequency dependence of emission
http//www.seaice.de/
53IR Measurements of Sea Surface Temperature
- Target Quantity sea surface temperature
- Measurement Quantity emitted IR radiation
- Instrument type nadir broad band IR measurements
- Instrument AVHRR (Advanced Very High Resolution
Radiometers)
54Reminder El Niño La Niña
- reversal of Walker circulation
- change of direction of Trade Winds
- change of ocean upwelling
- displacement of convection areas
- link to Southern Oscillation (difference of
surface pressure between Tahiti and Darwin)
55Sea Surface Anomaly during El Nino Event
- Sensor AVHRR
- Technique broad band IR measurements
- Quantity sea surface temperature
- Sensor TOPEX
- Technique radar altimeter
- Quantity height
56Summary
- Remote Sensing of atmospheric and surface
parameters from space relies on analysis of
electromagnetic radiation emitted / scattered /
reflected by the atmosphere and surface - The target quantities interact with the radiation
through absorption, emission, scattering,
reflection or by indirectly changing the optical
properties - Remote Sensing measurements provide a large
number of parameters for atmospheric physics and
chemistry on a global scale and often over long
time periods - Remote Sensing measurements are indirect
measurements and need thorough and continuous
validation - Spatial and temporal resolution of the
measurements are limited and not always
appropriate for detailed case studies - Technological improvements and progress in data
algorithms will further improve the usefulness of
satellite measurements in the future - Remote Sensing will always be only one of many
data sources needed to understand the Earth System