Dust Observations and Climatology - Mineral Dust: A Key Player in the Earth System - page 156

related to the breakdown of nocturnal low-level jets (see Chap. 6 ) . However, no
quantitative optical depth or source strength can be derived with this approach.
Banks and Brindley ( 2013 ) have developed a more quantitative approach from
SEVIRI/MSG observations in the 10.8 and 13.4 m channels that allows inferring
DOD at a half-hourly temporal resolution during the daytime. Such an approach
offers a new perspective for quantitative analysis of mineral dust source variability
at high temporal resolution over North Africa and the Middle East.
Global Data from Polar-Orbiting Satellite
The AI derived from TOMS/Nimbus 7 measurements (Herman et al. 1997 )inthe
UV part of the spectrum has been the first satellite dataset applied to identification
and characterization of mineral dust sources at a global scale (Prospero et al. 2002 ).
In contrast to the visible range, surface reflectivity is low and constant in the UV.
Thus, TOMS measurements allowed retrieving the loading of absorbing aerosols
(i.e. mainly mineral dust and carbonaceous particles) at a global scale, over both
oceanic and land surfaces, including arid and semi-arid areas of the globe. Regions
of maximum AI or high frequency of high AI occurrences have been used as a proxy
for dust source identification. Such an analysis has highlighted both the importance
of topographic depressions as the strongest sources and the dominance of dust
sources located in the Northern Hemisphere, mainly in a broad “dust belt” over
North Africa, the Middle East and Asia. Figure 3.2 in Chap. 3 nicely illustrates the
global map of dust source areas generated by analysis of the TOMS AI dataset.
From this approach, the Bodélé Depression in Chad has been identified to be the
most intense source region in the world (Prospero et al. 2002 ; Goudie and Middleton
2006 ). However, since the TOMS AI is known to be less sensitive to aerosols at low
altitude, the application of this index for dust source identification is questionable
in some cases (Mahowald and Dufresne 2004 ). Despite the large progress provided
by global TOMS AI analyses, the general difficulty of determining whether the dust
derived from satellite corresponds to transport or emission has been emphasized.
Thus, a more accurate identification of the sources of mineral dust from satellite
may require more sophisticated approaches. Hsu et al. ( 2004 ) have developed the
Deep Blue algorithm based on MODIS measurements in the 412 nm channel,
which has been applied to global dust source identification (Ginoux et al. 2012 ).
The advantage of Deep Blue compared to UV techniques is that it can detect dust
close to the surface and at a higher spatial resolution (0.1 ı ). Using criteria on size
distribution and optical properties allows distinction of DOD from AOD. However,
as the dust retrieval is limited to bright surfaces, dust sources as those in the Northern
Hemisphere high latitudes (Iceland, Alaska) may be underrepresented (Ginoux et al.
2012 ). The analysis of dust sources based on Deep Blue DOD retrievals tends to
confirm the results of Prospero et al. ( 2002 ), especially the remarkably little dust
activity in the Southern Hemisphere (Ginoux et al. 2012 ;Fig. 7.2 ).
Figure 7.2 highlights that the highest DOD values occur during spring and
summer season over arid areas in the Northern Hemisphere. High-resolution