An
understanding of the mechanisms underlying tropical precipitation
fluctuations over a range of space and time scales is critical to interpreting past and present climate variability and projecting
the impact of future climate change scenarios on global climate. Furthermore, tropical precipitation variability is a useful metric for constructing,
refining, and constraining climate models.
Perhaps more importantly, the variability of tropical precipitation has direct
and profound consequences for human well-being. Indeed, many inhabitants of tropical land regions are directly dependent
on rainfall for their livelihood. A lack or
overabundance of rainfall may lead to famine, pestilence, and the disruption of economic systems. Nowhere are the implications of tropical precipitation
fluctuations more
profound than in those regions which are, in a climatological sense, neither "too wet" nor "too dry". Perhaps the most well-known of such "marginal" environments
is the African Sahel, located between the arid Sahara Desert to the north and the
humid rainforests to the south.

November climatological-mean
rainfal distribution from the CMAP dataset.
The figure above illustrates a convective margin environment over tropical South America. The interior of the continent is characterized by precipitation values of up
to 10 mm/day (green shading); near the Atlantic coastal margins, precipitation values
are of order a few mm/day or less (brown shading). Between these extremes is
a region of intermediate climatological rainfall values for which the year-to-year flucutations
are rather sizeable: the blue contours demarcate the area for which the
standard deviation of November rainfall values is between 40%-60% of the
climatological mean for 1979-2005. Clearly, the occurrence of large
interannual variations of precipitation in the convective transition zone, or
along the convective margin, has significant implications for natural
ecosystems, agricultural practices, and land management.
General circulation model (GCM) studies of El Niño/Southern Oscillation (ENSO) teleconnections (e.g., Chiang and Sobel 2002; Chiang and Lintner, 2005; Lintner and Chiang 2005; Joseph and Nigam, 2006) or global warming (e.g., Neelin et al., 2006) indicate significant sensitivity
in how tropical precipitation varies with the forcing. The figure below, taken
from Neelin et al. (2006), shows sizeable differences in both the amplitude and geographic distribution of simulated
precipitation changes produced by global warming forcing scenarios in different model frameworks.
Lintner and Neelin (2007), hereafter LN07, outlines an analytic model of the convective margin for the specific case of inflow
into a tropical land region from a neighboring ocean region, as over the
northeastern corner of South America seen above. Starting from a set of simplified, vertically-averaged, steady-state tropospheric temperature and moisture
fields, a simple expression is derived for the convective margin location
xc:
Despite the intermodel disagreement, there appears to be some tendency for the
largest drying signatures to occur on the margins of the strongly convecting
regions of the Tropics. A detailed, mechanistic understanding of this
structure is
crucial to predicting how global warming will ultimately impact tropical climates.
An inflow convective margin prototype
defines a characteristic length-scale as a
function of the inflow wind projected onto the vertical structure of humidity
(uq) and the top-of-the-atmosphere radiative heating
(FTOA), which is related to the vertical motion. The
sense of the relationship is such that if ocean-to-land inflow increases,
the margin is displaced further away from the interface.
Consider what happens to the margin location under a T increase,
as during El Niño or under greenhouse gas warming. With an anomalously warm
troposphere, the convective threshold humidity value increases via the
"upped-ante mechanism" of Neelin and Chou (2004). All other factors being
equal, the margin should retreat further from the land-ocean interface.
Precisely how far the margin retreats depends on how far q0 is from qc(T) as well as the thermodynamic factors in &lambda. The simple prototype appears to provide a template for the
observed tendency for larger El Niño events (which warm the tropical troposphere more) to produce more spatially-extended droughts (Lyon, 2004; Lyon and
Barnston, 2005).
Current-generation GCM simulations of the climate in the vicinity of the SPCZ
exhibit many well-known biases, including excessively zonal convection and/or
deep convection extending too far eastward. As a first step in addressing the
role that (low-level) trade wind inflow may play in producing such biases, we
have investigated the high-frequency variability along the eastern margin of the
SPCZ (Lintner and Neelin, 2008). Using a compositing index based on a spatial average of
low-level zonal winds near the mean edge of the SPCZ, we computed anomalous
westerly/easterly (weaker/stronger trade wind inflow) composites of low-level
winds, moisture, and precipitation from observed and reanalysis daily data
(see figure below).
This analysis demonstrates the pronounced high-frequency sensitivity of
moisture and precipitation along the eastern edge of the SPCZ to inflow wind
variability, with anomalous westerlies/weaker trades associated with enhanced
tropospheric moisture and a more eastward position of the 4 mm day
precipitation contour. Our hypothesis for this behavior, formulated in terms
of convective margins, is that trade wind reduction leads to enhanced moisture
along the eastern SPCZ margin—because the "dry descent region" to the east of
the SPCZ is characterized by relatively low tropospheric moisture values,
reduced trades produce less dry air inflow—thereby supporting the occurrence of convection further to the east. A 2D generalization of the LN07 convective
margins prototype supports the plausibility of this interpretation, with the
prototype further demonstrating how changes in basic control parameters (such
as the convective moisture threshold) alter expression of the variability
(e.g., the amount by which the convection shifts).
Current work is underway both to test the usefulness of the simple prototype and to extend its applicability to more realistic situations.
Applicability to the South Pacific Convergence Zone (SPCZ)

Composite differences of daily mean (a) SSM/I satellite column-integrated water vapor (shading) and 925 mb NCEP Reanalysis winds (vectors) and (b) SSM/I precipitation, using daily NCEP Reanalysis 925 mb zonal wind averaged over the yellow box as the compositing index. Also shown in (b) are the 4 mm/day contours for anomalous westerlies (green) and easterlies (brown).
Current and Future Directions
back to home