TRMM: The Tropical Rainfall Measuring Mission
561
that matches the preferred locations of deep convection such as the ITCZ and the summer
midlatitude continents (Orville and Henderson, 1986). But while convection occurs frequently
over both land and ocean, lightning occurrence is an order of magnitude more frequent over
continental convection regions than over maritime locations of persistent convection.
The most plausible explanation for this discrepancy is the different strength of convection over
land vs. ocean. The microphysical factors leading to lightning production in thunderstorms are
still not totally understood, but appear to be dependent on aspects of mixed phase microphysics
(Williams et al., 1991). This in turn implies that lightning is favored by vertical advection of
liquid above the freezing level into the ice-forming regions of a convective storm. Thus, the
more vigorous the cumulus updrafts, the more likely the occurrence of lightning.
Observations support the idea of systematic differences between continental and maritime convection. Jorgensen and LeMone (1989) show that updraft strengths are significantly weaker in
oceanic deep convection than in midlatitude continental thunderstorms. The likely reason is
that the ocean is a consistent source of low-level moisture that allows the tropical atmosphere to
convect in response to even a small amount of conditional instability, thereby remaining close to
neutral stability. Over land, on the other hand, boundary layer relative humidity may at times
be too low to trigger convection despite conditionally unstable lapse rates in the free troposphere, allowing large amounts of convective available potential energy (CAPE) to build up until
sufficient low-level moistening occurs. Since in an undilute buoyant rising parcel, vertical velocity is proportional to (CAPE)1/2, oceanic convection should be characterized by small CAPE,
weak updrafts, and infrequent lightning, while continental convection should be more vigorous
and thus more prone to lightning formation. TRMM LIS observations may thus provide us
with an indirect near-global climatology of CAPE, a fundamental quantity in theories of deep
convection. Because of the implications for upward advection of liquid water, LIS data may
also serve as an indirect source of validation for the vertical structure of hydrometeors derived
by other TRMM instruments and assumed in TRMM physically-based retrieval algorithms.
20.5.2 Clouds and the Earth's Radiant Energy System (CERES)
TRMM is the first of several platforms on which the CERES instrument will fly to provide
sufficient geographical, diurnal and temporal sampling to permit accurate estimates of the
shortwave and longwave components of the top-of-the-atmosphere (TOA) radiation budget.
The science goals for CERES are more general than simply studies of tropical precipitating
systems, encompassing the radiative effects of all clouds as well as the Earth's surface.
But CERES plays a central role in the connection between TRMM science and questions about
global climate change. The radiation budget signature of tropical convective systems is large
and unique, with simultaneous large negative TOA shortwave cloud forcing and large positive
TOA longwave cloud forcing. These are correlated so as to produce small net TOA cloud forcing
(cf. Figure 3 of Del Genio, 1996, chapter 2), but this does not imply a small net radiative effect
because the vertical profiles of the shortwave and longwave effects differ dramatically. In the
GISS GCM, the cloud feedback and climate sensitivity in a climate change appear to depend
more on changes in the radiative properties of tropical anvils than on any other aspect of clouds
in any part of the globe (Del Genio et al., 1996).
The radiative properties in turn depend on the sources and sinks of anvil water, e.g., the
condensate detrained from cumulus updrafts and the stratiform precipitation produced by the
anvil itself. The former depends on the strength of updrafts, an index of which should be
provided by LIS, while the latter will be measured by the TRMM rain package instruments.
By combining these data with CERES information on anvil reflectance variations, it should be
561
that matches the preferred locations of deep convection such as the ITCZ and the summer
midlatitude continents (Orville and Henderson, 1986). But while convection occurs frequently
over both land and ocean, lightning occurrence is an order of magnitude more frequent over
continental convection regions than over maritime locations of persistent convection.
The most plausible explanation for this discrepancy is the different strength of convection over
land vs. ocean. The microphysical factors leading to lightning production in thunderstorms are
still not totally understood, but appear to be dependent on aspects of mixed phase microphysics
(Williams et al., 1991). This in turn implies that lightning is favored by vertical advection of
liquid above the freezing level into the ice-forming regions of a convective storm. Thus, the
more vigorous the cumulus updrafts, the more likely the occurrence of lightning.
Observations support the idea of systematic differences between continental and maritime convection. Jorgensen and LeMone (1989) show that updraft strengths are significantly weaker in
oceanic deep convection than in midlatitude continental thunderstorms. The likely reason is
that the ocean is a consistent source of low-level moisture that allows the tropical atmosphere to
convect in response to even a small amount of conditional instability, thereby remaining close to
neutral stability. Over land, on the other hand, boundary layer relative humidity may at times
be too low to trigger convection despite conditionally unstable lapse rates in the free troposphere, allowing large amounts of convective available potential energy (CAPE) to build up until
sufficient low-level moistening occurs. Since in an undilute buoyant rising parcel, vertical velocity is proportional to (CAPE)1/2, oceanic convection should be characterized by small CAPE,
weak updrafts, and infrequent lightning, while continental convection should be more vigorous
and thus more prone to lightning formation. TRMM LIS observations may thus provide us
with an indirect near-global climatology of CAPE, a fundamental quantity in theories of deep
convection. Because of the implications for upward advection of liquid water, LIS data may
also serve as an indirect source of validation for the vertical structure of hydrometeors derived
by other TRMM instruments and assumed in TRMM physically-based retrieval algorithms.
20.5.2 Clouds and the Earth's Radiant Energy System (CERES)
TRMM is the first of several platforms on which the CERES instrument will fly to provide
sufficient geographical, diurnal and temporal sampling to permit accurate estimates of the
shortwave and longwave components of the top-of-the-atmosphere (TOA) radiation budget.
The science goals for CERES are more general than simply studies of tropical precipitating
systems, encompassing the radiative effects of all clouds as well as the Earth's surface.
But CERES plays a central role in the connection between TRMM science and questions about
global climate change. The radiation budget signature of tropical convective systems is large
and unique, with simultaneous large negative TOA shortwave cloud forcing and large positive
TOA longwave cloud forcing. These are correlated so as to produce small net TOA cloud forcing
(cf. Figure 3 of Del Genio, 1996, chapter 2), but this does not imply a small net radiative effect
because the vertical profiles of the shortwave and longwave effects differ dramatically. In the
GISS GCM, the cloud feedback and climate sensitivity in a climate change appear to depend
more on changes in the radiative properties of tropical anvils than on any other aspect of clouds
in any part of the globe (Del Genio et al., 1996).
The radiative properties in turn depend on the sources and sinks of anvil water, e.g., the
condensate detrained from cumulus updrafts and the stratiform precipitation produced by the
anvil itself. The former depends on the strength of updrafts, an index of which should be
provided by LIS, while the latter will be measured by the TRMM rain package instruments.
By combining these data with CERES information on anvil reflectance variations, it should be