@PhDThesis{Nunes:2008:EsAnNu,
author = "Nunes, Andr{\'e} Becker",
title = "Crescimento da camada limite convectiva: estudo anal{\'{\i}}tico
e num{\'e}rico",
school = "Instituto Nacional de Pesquisas Espaciais (INPE)",
year = "2008",
address = "S{\~a}o Jos{\'e} dos Campos",
month = "2008-10-09",
keywords = "camada limite planet{\'a}ria, modelagem anal{\'{\i}}tica,
simula{\c{c}}{\~a}o de grandes v{\'o}rtices, crescimento da
convec{\c{c}}{\~a}o, resolu{\c{c}}{\~a}o de grade. planetary
boundary layer, analytical modeling, large eddy simullation,
growth of convection, grid resolution.",
abstract = "Com o avan{\c{c}}o computacional, a modelagem num{\'e}rica da
turbul{\^e}ncia atmosf{\'e}rica tem sido de valiosa
import{\^a}ncia na an{\'a}lise da f{\'{\i}}sica da Camada
Limite Planet{\'a}ria (CLP). Neste {\^a}mbito, destaca-se a
Simula{\c{c}}{\~a}o de Grandes V{\'o}rtices (LES). O modelo
LES, vastamente empregado em estudos micrometeorol{\'o}gicos, tem
como objetivo a simula{\c{c}}{\~a}o direta dos grandes
v{\'o}rtices, fen{\^o}menos da escala resolvida, e
parametriza{\c{c}}{\~a}o dos pequenos v{\'o}rtices, de escala
de subgrade. Contudo, al{\'e}m do desenvolvimento de
processamentos cada vez mais r{\'a}pidos e robustos, a modelagem
num{\'e}rica depende, essencialmente, da elabora{\c{c}}{\~a}o
de modelos anal{\'{\i}}ticos. Por outro lado, a modelagem
num{\'e}rica auxilia na comprova{\c{c}}{\~a}o de modelos
te{\'o}ricos, principalmente quando consegue substituir dados
observacionais dif{\'{\i}}ceis de serem obtidos. Desta forma, o
objetivo desta tese {\'e} a verifica{\c{c}}{\~a}o de duas
metodologias te{\'o}ricas por meio do modelo LES. A primeira
{\'e} baseada na escolha de grade a ser selecionada em uma
simula{\c{c}}{\~a}o num{\'e}rica. Aqui, o modelo LES comprova a
coer{\^e}ncia da teoria que sugere uma restri{\c{c}}{\~a}o
f{\'{\i}}sica, a altura do topo da camada convectiva, para a
determina{\c{c}}{\~a}o do espa{\c{c}}amento de grade. A escolha
da resolu{\c{c}}{\~a}o a ser empregada em um modelo
num{\'e}rico era, at{\'e} ent{\~a}o, subjetiva, pois baseava-se
no n{\'u}mero m{\'a}ximo de pontos que a estrutura computacional
dispon{\'{\i}}vel conseguisse calcular em um tempo
razo{\'a}vel. Tal escolha poderia gerar um gasto computacional
excessivo ou perda de informa{\c{c}}{\~a}o. A segunda teoria
desenvolvida nesta tese {\'e} a modelagem anal{\'{\i}}tica da
fase de transi{\c{c}}{\~a}o matutina, baseada na
equa{\c{c}}{\~a}o de espectro tridimensional do crescimento da
convec{\c{c}}{\~a}o - quest{\~a}o ainda em aberto na literatura
micrometeorol{\'o}gica e, portanto, principal
contribui{\c{c}}{\~a}o deste trabalho. O pleno entendimento da
CLP depende da an{\'a}lise das fases de transi{\c{c}}{\~a}o.
Aqui, discute-se a fase matutina, per{\'{\i}}odo do ciclo diurno
menos estudado no meio cient{\'{\i}}fico. No desenvolvimento do
modelo anal{\'{\i}}tico apresentado neste trabalho emprega-se um
conjunto de formula{\c{c}}{\~o}es (metodologias de
convers{\~a}o de espectro unidimensional em tridimensional,
equa{\c{c}}{\~o}es de espectro unidimensional, vari{\^a}ncias
de velocidade, taxas de dissipa{\c{c}}{\~a}o) que, ao serem
inseridas na equa{\c{c}}{\~a}o de espectro do crescimento, geram
quatro modelos anal{\'{\i}}ticos. A precis{\~a}o destes modelos
{\'e} comprovada por meio da compara{\c{c}}{\~a}o entre a
energia cin{\'e}tica turbulenta (ECT) gerada analiticamente e a
ECT gerada pelo modelo LES. Entretanto, para
verifica{\c{c}}{\~a}o da coer{\^e}ncia dos resultados do modelo
LES, previamente foi necess{\'a}ria a simula{\c{c}}{\~a}o
num{\'e}rica do ciclo diurno (camada convectiva, decaimento,
camada neutra, crescimento e camada convectiva novamente), o que
possibilitou uma proveitosa discuss{\~a}o dos diferentes regimes
turbulentos da CLP. Portanto, conclui-se que os resultados
alcan{\c{c}}ados nesta tese contribuem decisivamente na escolha
da resolu{\c{c}}{\~a}o de grade a ser adotada em
simula{\c{c}}{\~o}es convectivas e, principalmente, na modelagem
e discuss{\~a}o da camada limite matutina. ABSTRACT: With the
computational progress, numerical modeling has been very important
in the physics of Planetary Boundary Layer (PBL) analysis. In this
scope, it emphasizes the Large Eddy Simulation (LES). The LES
model, broadly employed in meteorological studies, aims the direct
simulation of large eddies, resolved scales phenomena, and
parameterizes the small ones, subgrid scale phenomena. However,
beyond the development of processing more and more quick and
massive, the numerical modeling essentially depends on the working
out of analytical modeling. On the other hand, numerical modeling
contributes in the proof of theoretical models, mainly when it is
able to substitute difficult to be obtained observed data. In this
way, the objective of this work is the verifying of two
theoretical methodologies through the employment of LES model. The
first one is based on the grid resolution to be chosen in a
numerical simulation. Here, LES model proofs the coherence of the
theory that suggests a physical criterion, the convective layer
top height, to determinate the grid spacing. The choice of the
resolution to be employed in a numerical model was, so far,
subjective, since it was based on the maximum number of points
that the disposable computational structure was able to computing
under a reasonable time. Such choice could generate an excessive
computing spent or lost of information. The second theory
developed in this thesis is the analytical modeling of morning
transition phase, based on equation of convection growing
three-dimension spectrum still open question in
micrometeorological literature and, hence, the main contribution
of this work. The full understanding of PBL depends on the
transition phases analysis. Here, it is argued about morning
phase, less studied period of diurnal cycle in the scientific
knowledge. In the development of analytical model presented in
this work is employed a set of formulations (conversion of
one-dimension to three-dimension spectrum methodologies,
one-dimension spectrum equations, velocity variances, dissipation
rates) which, inserted in the growing spectrum equation, generates
four analytical models. The accuracy of such models is proved
through a comparison between the turbulent kinetic energy (TKE)
generated analytically and the TKE generated by LES. Nevertheless,
to verifying the LES results coherence, was previously made a
numerical simulation of diurnal cycle (convective layer, decay,
neutral layer, growth and convective layer again), what resulted
in a useful discussion about different turbulent regimes in CLP.
Therefore, it concludes that the results obtained in this thesis
contribute well on the grid resolution choice to be adopted on
convective simulations and, principally, in the modeling and
discussion of morning boundary layer.",
committee = "Fisch, Gilberto Fernando (presidente) and Satyamurty, Prakki
(orientador) and Campos Velho, Haroldo Fraga de (orientador) and
Oliveira, Amauri Pereira de and Acevedo, Ot{\'a}vio Costa",
copyholder = "SID/SCD",
englishtitle = "Convective boundary layer growth: numerical and analytical study",
language = "pt",
pages = "192",
ibi = "8JMKD3MGP8W/34C5U9H",
url = "http://urlib.net/ibi/8JMKD3MGP8W/34C5U9H",
targetfile = "publicacao.pdf",
urlaccessdate = "27 abr. 2024"
}