BIOGEOPHYSICAL CONSEQUENCES OF A TROPICAL DEFORESTATION SCENARIO - A GCM SIMULATION STUDY

Citation
Yc. Sud et al., BIOGEOPHYSICAL CONSEQUENCES OF A TROPICAL DEFORESTATION SCENARIO - A GCM SIMULATION STUDY, Journal of climate, 9(12), 1996, pp. 3225-3247
Citations number
57
Categorie Soggetti
Metereology & Atmospheric Sciences
Journal title
ISSN journal
08948755
Volume
9
Issue
12
Year of publication
1996
Part
2
Pages
3225 - 3247
Database
ISI
SICI code
0894-8755(1996)9:12<3225:BCOATD>2.0.ZU;2-L
Abstract
Two 3-year (1979-1982) integrations were carried out with a version of the GLA GCM that contains the Simple Biosphere Model (SiB) for simula ting land-atmosphere interactions. The control case used the usual SIE vegetation cover (comprising 12 vegetation types), while its twin, th e deforestation case, imposed a scenario in which all tropical rainfor ests were entirely replaced by grassland. Except for this difference, all other initial and prescribed boundary conditions were kept identic al in both integrations. An intercomparison of the integrations shows that tropical deforestation decreases evapotranspiration and increases land surface outgoing longwave radiation and sensible heat flux, ther eby warming and drying the planetary boundary layer. This happens desp ite the reduced absorption of solar radiation due to higher surface al bedo of the deforested land. produces significant and robust local as well as global climate changes. The local effect includes significant changes (mostly reductions) in precipitation and diabatic heating, whi le the large-scale effect is to weaken the Hadley circulation but invi gorate the southern Ferrel cell, drawing larger air mass from the indi rect polar cells. decreases the surface stress (drag force) owing to r educed surface roughness of deforested land, which in turn intensifies winds in the planetary boundary layer, thereby affecting the dynamic structure of moisture convergence. The simulated surface winds are abo ut 70% stronger and are accompanied by significant changes in the powe r spectrum of the annual cycle of surface and PBL winds and precipitat ion. Our results broadly confirm several findings of recent tropical d eforestation simulation experiments. In addition, some global-scale cl imatic influences of deforestation not identified in earlier studies a re delineated.