INTRODUCTION
1
1 Introduction
The growing interest in climate and climate change has renewed the demand for climate
descriptions of the world oceans. Fields of climate variables are used as boundary conditions
or sea truth for atmospheric and oceanic models and coupled models of the atmosphereocean system. The fast improvements that modeling has made in recent years also call for
more detailed descriptions of processes at the air-sea interface, especially of the air-sea fluxes
of momentum, heat and matter. The ocean heat transport is understood as a major branch
of the global climate system. The large scale ocean circulations depend on wind stress and
thermohaline forcing, that both would likely be influenced by climate changes.
Today marine weather observations are still the major source for the study of longterm
climate at the ocean. Since area covering data of direct flux measurements are not available, air-sea fluxes have to be derived by parameterisations using the basic meteorological
observations.
The climate atlas presented here is restricted to the Atlantic Ocean. The Atlantic is of special interest for two reasons. The cross-equatorial ocean heat transport of about 1 petawatt is
an outstanding climate feature, that contributes to the relatively high sea surface temperatures
in the Northeastern North Atlantic and the mild climate of Europe. From a more technical
view, the data coverage of the Atlantic is much denser than for other parts of the world oceans.
H~nce the Atlantic can well serve as a reliable test area of sea truth for both modelling and
satellite remote sensing.
2 Data and Data Treatment
Individual ship reports from the Comprehensive Ocean-Atmosphere Data Set (COADS)
of the Atlantic Ocean from 1940 to 1979 are used for this climatological study. Maps of
monthly mean distribution are calculated not only for the basic ship observations as wind
speed or air pressure but also for the indirectly derived climate variables such as evaporation or wind stress, which are obtained by parameterisations. For the derived variables the
individual method is used that conserves the effects of cross-correlations. The classical climatological method, that uses averaged values of the meteorological parameters, cannot account
for these contributions. For the North East Atlantic JOSEY & al. (1995) showed that the
latent heat flux is overestimated by the crude climatological method. The negative correlation between wind speed and sea-air humidity difference causes an overestimation from 1-2
Wm- 2 in winter to 7 Wm- 2 in summer.
The distribution of ship observations in the Atlantic Ocean is very inhomogeneous. Especially in the South Atlantic reports are concentrated along narrow shipping routes, whereas
broad areas remain nearly unsampled (fig. 1). However, coastal regions, where we suppose
the strongest gradients, are in general well frequented by ships. In the open ocean the relatively small number of available observations is tolerable, becauce strong gradients are not
1
1 Introduction
The growing interest in climate and climate change has renewed the demand for climate
descriptions of the world oceans. Fields of climate variables are used as boundary conditions
or sea truth for atmospheric and oceanic models and coupled models of the atmosphereocean system. The fast improvements that modeling has made in recent years also call for
more detailed descriptions of processes at the air-sea interface, especially of the air-sea fluxes
of momentum, heat and matter. The ocean heat transport is understood as a major branch
of the global climate system. The large scale ocean circulations depend on wind stress and
thermohaline forcing, that both would likely be influenced by climate changes.
Today marine weather observations are still the major source for the study of longterm
climate at the ocean. Since area covering data of direct flux measurements are not available, air-sea fluxes have to be derived by parameterisations using the basic meteorological
observations.
The climate atlas presented here is restricted to the Atlantic Ocean. The Atlantic is of special interest for two reasons. The cross-equatorial ocean heat transport of about 1 petawatt is
an outstanding climate feature, that contributes to the relatively high sea surface temperatures
in the Northeastern North Atlantic and the mild climate of Europe. From a more technical
view, the data coverage of the Atlantic is much denser than for other parts of the world oceans.
H~nce the Atlantic can well serve as a reliable test area of sea truth for both modelling and
satellite remote sensing.
2 Data and Data Treatment
Individual ship reports from the Comprehensive Ocean-Atmosphere Data Set (COADS)
of the Atlantic Ocean from 1940 to 1979 are used for this climatological study. Maps of
monthly mean distribution are calculated not only for the basic ship observations as wind
speed or air pressure but also for the indirectly derived climate variables such as evaporation or wind stress, which are obtained by parameterisations. For the derived variables the
individual method is used that conserves the effects of cross-correlations. The classical climatological method, that uses averaged values of the meteorological parameters, cannot account
for these contributions. For the North East Atlantic JOSEY & al. (1995) showed that the
latent heat flux is overestimated by the crude climatological method. The negative correlation between wind speed and sea-air humidity difference causes an overestimation from 1-2
Wm- 2 in winter to 7 Wm- 2 in summer.
The distribution of ship observations in the Atlantic Ocean is very inhomogeneous. Especially in the South Atlantic reports are concentrated along narrow shipping routes, whereas
broad areas remain nearly unsampled (fig. 1). However, coastal regions, where we suppose
the strongest gradients, are in general well frequented by ships. In the open ocean the relatively small number of available observations is tolerable, becauce strong gradients are not