432
P. Schliissel
sheds light on the retrieval accuracy and on the differences between various methods in use.
For single match-ups the standard deviation between ship and satellite measurements is about
30 W/m 2 leading to monthly mean errors less than 10 W/m2.
16.1 Introduction
Ocean and atmosphere are coupled by fluxes of energy, matter, and momentum through the
air-sea interface. The evaporation over sea accounts for two of these fluxes, for the transfer of
fresh water and for the related latent heat flux. Besides these fluxes the evaporation affects the
salinity budget of the ocean. Passive infrared and microwave remote sensing bear the potential
for improving the evaporation estimates in data-sparse areas of the oceans, thus supplying
to the overall goal of GEWEX, the improved understanding of the role of water vapour in
meteorological, hydrological and climatological processes. Direct observation of the evaporation
is only possible by means of eddy correlation measurements which are not accessible by satellite
methods. The determination of the vertical water-vapour flux density E, or water-vapour flux
for short, requires to determine the correlation product
E = pw'q'
(16.1 )
where w' and q' are the fluctuations of the vertical wind velocity and the specific air humidity,
respectively, and p is the air density. Fluctuations of at least 10 Hz have to be resolved over
averaging periods of about an hour in order to get a reliable measurement of E. Equation
(16.1) is usually parameterised by the so-called bulk aerodynamic formula that allows the use
of readily accessible surface measurements from the routine meteorological observation network:
E = -CEPu(qo - q)
(16.2)
where u is the average horizontal surface wind speed, qo and q are the average specific humidities
of the surface and the air, respectively, and CE is the bulk transfer coefficient for water vapour,
also known as Dalton number. The transfer coefficient is normally referred to measurements
at 10 m and averaging periods of 10 minutes (Hasse, 1993). The key parameters to be entered
in (16.2) are surface temperature, To, which determines qo, surface wind speed, u, and surface
air humidity, q. The latent heat flux related to E is QE = LE where L is the latent heat of
evaporation.
The satellite instruments being employed for the retrieval of latent heat fluxes are mainly
infrared and microwave radiometers, but also scatterometer winds have been used so far. While
microwave radiometry is used for the measurement of surface wind speed and atmospheric water
vapour infrared radiometers are needed for an accurate measurement of sea surface temperature.
Instruments that are currently operational and utilized for the retrieval of the evaporation over
sea are the Special Sensor Microwave/Imager (SSM/I), described by Simmer (this issue), and
the infrared imagers AVHRR (Advanced Very High Resolution Radiometer, flown on the NOAA
satellites) and the ATSR (Along Track Scanning Radiometer, that is part of the ERS-l satellite).
Both infrared radiometers are designed to measure radiation in atmospheric window regions.
Their main characteristics are listed in tables 16.1 and 16.2. The spectral response functions
of their thermal channels are shown together with the atmospheric transmission spectra for
a tropical and a subarctic atmosphere in figure 16.1. The main innovation of ATSR is the
active detector cooling which leads to a reduced noise behaviour when compared to the AVHRR
instrument that allows for a more accurate retrieval of sea surface temperatures. While A VHRR
has a radiometric noise equivalent to a temperature difference NEt:.T = 0.12 K the ATSR
values are 0.02, 0.04, and 0.08 K for the 11, 12, and 3.7 11m channels, respectively (Schliissel
P. Schliissel
sheds light on the retrieval accuracy and on the differences between various methods in use.
For single match-ups the standard deviation between ship and satellite measurements is about
30 W/m 2 leading to monthly mean errors less than 10 W/m2.
16.1 Introduction
Ocean and atmosphere are coupled by fluxes of energy, matter, and momentum through the
air-sea interface. The evaporation over sea accounts for two of these fluxes, for the transfer of
fresh water and for the related latent heat flux. Besides these fluxes the evaporation affects the
salinity budget of the ocean. Passive infrared and microwave remote sensing bear the potential
for improving the evaporation estimates in data-sparse areas of the oceans, thus supplying
to the overall goal of GEWEX, the improved understanding of the role of water vapour in
meteorological, hydrological and climatological processes. Direct observation of the evaporation
is only possible by means of eddy correlation measurements which are not accessible by satellite
methods. The determination of the vertical water-vapour flux density E, or water-vapour flux
for short, requires to determine the correlation product
E = pw'q'
(16.1 )
where w' and q' are the fluctuations of the vertical wind velocity and the specific air humidity,
respectively, and p is the air density. Fluctuations of at least 10 Hz have to be resolved over
averaging periods of about an hour in order to get a reliable measurement of E. Equation
(16.1) is usually parameterised by the so-called bulk aerodynamic formula that allows the use
of readily accessible surface measurements from the routine meteorological observation network:
E = -CEPu(qo - q)
(16.2)
where u is the average horizontal surface wind speed, qo and q are the average specific humidities
of the surface and the air, respectively, and CE is the bulk transfer coefficient for water vapour,
also known as Dalton number. The transfer coefficient is normally referred to measurements
at 10 m and averaging periods of 10 minutes (Hasse, 1993). The key parameters to be entered
in (16.2) are surface temperature, To, which determines qo, surface wind speed, u, and surface
air humidity, q. The latent heat flux related to E is QE = LE where L is the latent heat of
evaporation.
The satellite instruments being employed for the retrieval of latent heat fluxes are mainly
infrared and microwave radiometers, but also scatterometer winds have been used so far. While
microwave radiometry is used for the measurement of surface wind speed and atmospheric water
vapour infrared radiometers are needed for an accurate measurement of sea surface temperature.
Instruments that are currently operational and utilized for the retrieval of the evaporation over
sea are the Special Sensor Microwave/Imager (SSM/I), described by Simmer (this issue), and
the infrared imagers AVHRR (Advanced Very High Resolution Radiometer, flown on the NOAA
satellites) and the ATSR (Along Track Scanning Radiometer, that is part of the ERS-l satellite).
Both infrared radiometers are designed to measure radiation in atmospheric window regions.
Their main characteristics are listed in tables 16.1 and 16.2. The spectral response functions
of their thermal channels are shown together with the atmospheric transmission spectra for
a tropical and a subarctic atmosphere in figure 16.1. The main innovation of ATSR is the
active detector cooling which leads to a reduced noise behaviour when compared to the AVHRR
instrument that allows for a more accurate retrieval of sea surface temperatures. While A VHRR
has a radiometric noise equivalent to a temperature difference NEt:.T = 0.12 K the ATSR
values are 0.02, 0.04, and 0.08 K for the 11, 12, and 3.7 11m channels, respectively (Schliissel