with an allegory: We are like a man sitting in
a cave looking at the wall, a fire burning behind him. If objects (i.e., ideas) are carried before the fire, the man sees only their enlarged
shadows on the wall, and he will take them
as the real world. Today’s neuroscientist also
can view “shadows” of the real world. The enlarged shadows of neurons, dendritic spines,
and synapses can now be seen through expansion microscopy and attempts can be
made to put together a super-resolved model
of the brain (see the figure).
Expanded brains have the great advantage
that the heavy absorption of water makes
them completely transparent. Thus, they are
ideal objects for light sheet microscopy, which
requires transparent specimens. In light sheet
microscopy, an object is illuminated from the
side with a thin sheet of light to create optical sectioning, whereas in traditional microscopy, the light comes from below or above.
Since the demonstration 8 years ago (4) that
neuronal networks in cleared mouse brains
can be visualized with light sheet microscopy,
the application of this approach has grown
rapidly. In retrospect, it is puzzling that light
sheet microscopy was not taken up sooner to
study the brain, as the technique is 100 years
old. Technically, it is easier to implement than
confocal microscopy, in which a light beam
is used to take images through serial planes
of a sample and their subsequent stacking
creates a three-dimensional (3D) image. But
light sheet microscopy still needed clearing
technology to have sufficient impact. Today’s
clearing technology not only aims at perfect
transparency but also now provides, with the
method of Chen et al., a way to circumvent
Abbe’s resolution limit.
One might imagine that in the future, a
whole mouse brain could be expanded 10
times to the size of a small orange, with all
neurons and synapses of interest labeled in
various colors. Light sheet microscopy may
well adopt superresolving stimulated emission depletion (STED) technology to scan
such a sample with extremely thin sheets of
light in all three dimensions. For the thousands of exabytes then generated to create
3D images, supercomputers for “the human
brain project” will be desperately needed. ■
1. T. A. Klar, S. Jakobs, M. Dyba, A. Egner, S. W. Hell, Proc.Natl.
Acad. Sci. U.S.A. 97, 8206 (2000).
2. E. Betzig et al ., Science 313, 1642 (2006).
3. F. Chen et al., Science 347, 543 (2015).
4. H.-U. Dodt et al ., Nat. Methods 4, 331 (2007).
5. H. Hama et al ., Nat. Neurosci. 14, 1481 (2011).
6. K. Becker et al., PLOS ONE 7, e33916 (2012).
7. A. Ertürk et al., Nat. Protoc. 7, 1983 (2012).
8. T.Kuwajima et al., Development 140,1364(2013).
9. M. T. Ke, S. Fujimoto, T. Imai, Nat.Neurosci. 16, 1154 (2013).
10. K. Chung et al ., Nature 497, 332 (2013).
It has been widely accepted since Carnot’s eminal work (1) that the atmosphere acts as a thermodynamic heat engine: Air motions redistribute the energy gained from the Sun in the warm part of the globe to colder regions where it
is lost through the emission of infrared radiation to space. Through this process, some
internal energy is converted into the kinetic
energy needed to maintain the atmospheric
circulation against dissipation. The analogy
to a heat engine has been applied to explain
various atmospheric phenomena, such as
the global circulation (2), hurricanes (3), and
dust devils (4). On page 540 of this issue,
Laliberté et al. (5) show that the hydrological cycle reduces the efficiency of the global
atmospheric heat engine.
Although the atmosphere acts overall as
a heat engine, it may not be a very efficient
one. The distribution of energy sources and
sinks around Earth implies a maximum
work of ~10 W·m−2 when averaged glob-
ally, yet estimates for the rate of kinetic
energy production by atmospheric motions
are about half this figure. The difference is
very likely due to Earth’s hydrological cycle,
which reduces the production of kinetic en-
ergy in two ways. First, as rain droplets, hail
pellets, or snowflakes fall through the atmo-
sphere, they generate microphysical shear
zones in which a substantial amount of dis-
sipation occurs. Satellite data yield an esti-
mated dissipation rate from precipitation of
~1.2 W·m−2 in the tropics (6). Second, water
mostly evaporates in unsaturated air, and
this evaporation is thus thermodynamically
irreversible. Such irreversibility reduces the
work produced by a heat engine compared
to a Carnot cycle. In particular, when the en-
ergy source of the Carnot cycle is replaced
by evaporation, the mechanical work of the
cycle is reduced and strongly depends on the
relative humidity (7).
Idealized cycles such as the Carnot cycle
are useful theoretical models for understanding how much wind can be sustained around
the globe. However, the atmosphere is highly
turbulent, and extracting a single cycle from
the complex motions that make up the global
circulation is not straightforward. Thermodynamic studies of the climate system have
typically required detailed knowledge of all
thermodynamic transformations and have
involved a complex analysis of numerical
simulations (8–10). These intensive data
requirements explain why there have been
relatively few thermodynamic studies of the
Laliberté et al. offer an elegant way to address this problem. Meteorologists have studied air flow on surfaces of constant entropy
since the 1930s. Atmospheric scientists and
oceanographers have extended this approach
to study how air and water move around the
planet on various thermodynamic surfaces,
such as surfaces of constant temperature
or salinity (11, 12). Laliberté et al. push this
idea even further by averaging the circulation in a three-dimensional thermodynamic
space (pressure, entropy, and water content).
Dalton’s law, however, states that all thermodynamic properties of moist air can be determined from the knowledge of these three
state variables. This means that the circula-
The global engine that could
Fine weather machine. Nice, sunny weather occurs when dry air from the upper troposphere sinks and is mixed with
moist air near Earth’s surface. Laliberté et al. show that such mixing weakens the atmospheric heat engine.
How do hydrological processes affect Earth’s heat engine?
Center for Atmosphere Ocean Science, New York University,
New York, NY 10012, USA. E-mail: email@example.com
By Olivier M. Pauluis