Kw_492 = Kw_490 + ((492 - 490) * ((Kw_490 - Kw_495) / (490 - 495)))
Rrs = data_obj.Rrs ## Reflectância SR superfície
## Separação dos comprimentos de onda de referência para os cálculos em um
## dicionário de parâmetros
wvl_ref = {
'referencia': list(range(400, 751, 1)), ## Bandas PAR
'lb0': lb0 ## Banda inicial (lambda zero)
}
## Variáveis para chamada nos cálculos
wl = pd.Series(wvl_ref['referencia'])
## Variáveis de reflectâncias de SR para chamada nos cálculos
r_ref = filter_wvl(rrs, wvl_ref)
R_ref = filter_wvl(Rrs, wvl_ref)
###############################################################################
############################### OPERACOES ###################################
###############################################################################
## Importando pacotes para realização do teste - Empirical Models
from qaapy_kdpy.empirical_calc import *
from qaapy_kdpy.utils.plot_Kd_campo_vs_qaa import plot_ajuste
from scipy.optimize import curve_fit
## Razão de três bandas Rrs para cálculo de a(lb0)
Rrs_ratio = calc_aRrs_ratio(R_ref, lb0, 665, 704)
Rrs_ratio = Rrs_ratio.rename('Rrs_ratio', axis = 'columns')
## Função de ajuste para o cálculo dos índices da função a(lb0) [multiplicador e expoente]
def Qaa_v5_model_RUN(data_obj, ano, lb0):
"""
TODO: WRITE DESCRIPTION AND CREATE A PATERN TO THIS function
"""
###############################################################################
########################### QAA SETTINGS - GENERAL ############################
###############################################################################
data_obj.Rrs = data_obj.Rrs.drop([833, 865, 945, 1374, 1614, 2202])
data_obj.aw = data_obj.aw.drop([833, 865, 945, 1374, 1614, 2202])
data_obj.bbw = data_obj.bbw.drop([833, 865, 945, 1374, 1614, 2202])
# Transformação do Rrs
rrs = Rrs_to_rrs(data_obj.Rrs)
Rrs = data_obj.Rrs
# Cálculando o U
g0 = 0.089
g1 = 0.1245
u = calc_u(g0, g1, rrs)
###############################################################################
#ToDo: checar se isso é necessário aqui, talvez criar uma função que padronize isso fora dos OLCI_models.
# Filtrando os dados
wvl_ref = {
'referencia': [443, 492, 560, 665, 704, 741,
783], # Bandas do sensor Sentinel-2A
# 'referencia': [400,413,443,490,510,560,620,665,674,681,709,754,761,764,768,779], # Bandas do sensor OLCI
# 'referencia': [400,413,443,490,510,560,620,665,674,681,709,754,761,764,768,779,865,885,900], # Bandas do sensor OLCI
# 'referencia': [400,411,443,490,560,620,667,674,681,709,754,761,779,865,555],# Bandas v5
'lb0': lb0
}
wl = pd.Series(wvl_ref['referencia'])
r_ref = filter_wvl(rrs, wvl_ref)
R_ref = filter_wvl(Rrs, wvl_ref)
###############################################################################
############################ FASE 1 - QAA.V5 - at #############################
###############################################################################
p0 = [-1.146, -1.336, -0.469] # Valores iniciais
# eq_lee_v5_OLCI = np.log10( (r_ref[443] + r_ref[490]) / ( r_ref['r_lb0'] + 5 * (r_ref[665]/r_ref[510] * r_ref[665]))); lago = 'QAA.v5_OLCI'
eq_lee_v5_s2a = np.log10(
(r_ref[443] + r_ref[492]) / (r_ref['lb0'] + 5 *
(r_ref[665] / r_ref[492] * r_ref[665])))
lago = 'QAA.v5_S2a'
#### ATENÇÃO : a agua é adicionada ao acs dentro da função "optimizer_aLb0" ##########
a_Lb0_optimized = optimizer_aLb0(eq_lee_v5_s2a,
wvl_ref['lb0'],
p0,
data_obj.aw,
data_obj.acs,
NOToptimize=True)
opt = '_OptN'
# a_Lb0_optimized = optimizer_aLb0(eq_lee_v5_s2a, wvl_ref['lb0'], p0, data_obj.aw, data_obj.acs, NOToptimize = False); opt = '_OptY'
#### bbp (lambda_0)
bbp_lb0 = calc_bbp_lb0(u, data_obj.bbw, 560, wvl_ref['lb0'],
a_Lb0_optimized['aref'])
## Cálculando N
n = calc_n(r_ref[443], r_ref[560])
# Realizando cálculo do BBP para todos os comprimentos de onda
bbp = calc_bbp(wl, n, bbp_lb0, wvl_ref['lb0'])
# Realizando cálculo do BB para todos comprimentos de onda
bb = calc_bb(bbp, data_obj.bbw)
# Realizando cálculo do atotal para todos os comprimentos de onda
at = calc_a_total_nw(bbp, data_obj.bbw, u)
###############################################################################
############################ FASE 2 - Kd ######################################
###############################################################################
m = [0.005, 4.26, 0.52, 10.54]
chi = 0.265
theta_10 = 10
theta_15 = 15
raz_bbw_bb = calc_raz_bbw_bb(data_obj.bbw, bb)
Kd_10 = calc_kd_lee_2013(m, theta_10, at, bb, chi, raz_bbw_bb)
Kd_15 = calc_kd_lee_2013(m, theta_15, at, bb, chi, raz_bbw_bb)
###############################################################################
########################### Data Extraction ###################################
###############################################################################
result_dict = {}
result_dict['dados'] = {
'lb0': wvl_ref['lb0'],
'r_ref': list(r_ref.keys()),
'campanhas': ano,
'equacao': lago,
'chute_inicial': a_Lb0_optimized['chute_inicial'],
'h0h1h2': a_Lb0_optimized['h_otimizado'],
'g0': g0,
'g1': g1
}
result_dict['a_Lb'] = a_Lb0_optimized['aref']
result_dict['bb'] = bb
result_dict['at'] = at
result_dict['n'] = n
result_dict['bbp_lb0'] = bbp_lb0
result_dict['bbp'] = bbp
result_dict['Kd_10'] = Kd_10
result_dict['Kd_15'] = Kd_15
result_dict['Kd_ZEU'] = data_obj.Kd_ZEU
result_dict['Kd_3m'] = data_obj.Kd_3m
result_dict['Kd_6m'] = data_obj.Kd_6m
os.makedirs('./Results_S2A', exist_ok=True)
export_result(
result_dict, './Results_S2A/QAAv5_' + ano + '_S2a_' +
str(wvl_ref['lb0']) + opt + '.pickle')
return result_dict
def Qaa_v6_model_RUN_original(data_obj, ano, lb0):
'''
Função que aplica o modelo QAA v6 dado um objeto de Data Frames contendo todos
os dados necessários, uma string com o nome/ano da campanha para salvar o arquivo
dos resultados e um valor de comprimento de onda inicial (Lambda Zero).
----------
Parameters
----------
data_obj [Object]
Objeto com Data Frames contendo dados usados nas funções do modelo QAA/Kd
alocados pela função read_files através dos argumentos contendo as strings
do caminho do diretório para cada planilha '.xlsx' de dados.
O objeto de Data Frames deve conter no mínimo:
[acs] Absorção total na coluna d'água;
[Rrs] Reflectância de sensoriamento remoto superficial;
[aw] Coeficiente de absorção espectral da água;
[bbw] Coeficiente de retroespalhamento espectral da água;
[coleta] Planilha contendo Coordenadas, datas e horários das
coletas por estação;
[Kd_ZEU] Kd medido com dados até a zona de atenação da luz até 1%;
[Kd_3m] Kd medido com dados até 3 metros de profundidade;
[Kd_6m] Kd medido com dados até 6 metros de profundidade.
* Os dados devem concordar em faixa espectral, bandas e nome das estações.
ano [String]
String contendo o nome ou ano da campanha, usado para salvar arquivo
dicionário contendo todos os dados produtos no formato '.pickle'.
Ex.: 'TRM2019', '2013', 'PACO_2011', etc...
lb0 [Value]
Comprimento de onda incial usado no modelo QAA (Lambda Zero).
Ex.: 665, 560, 490, etc...
-------
Returns
-------
result_dict [Dictionary]
O resultado desta função é um dicionário contendo todos os resultados e
dados de entrada relevantes.
Além disso o modelo salva um arquivo contendo todos os dados em formato
'.pickle' na pasta '.\Results_Espectral'.
Dados salvos:
[dados/lb0] Comprimento de onda incial (Lambda Zero);
[dados/wl] Comprimentos de onda utilizados;
[dados/r_ref] Dados de rrs;
[dados/campanha] Nome/ano da campanha;
[dados/equação] Equação utilizada no modelo em questão;
[dados/g0] Coeficiente g1 - Passo 1 QAA;
[dados/g1] Coeficiente g0 - Passo 1 QAA;
[a_Lb] Absorção em Lambda Zero;
[bb] Retroespalhamento total;
[at] Absorção total estimada pelo modelo QAA;
[n] Decaimento do retroespalhamento do particulado;
[bbp_lb0] Retroespalhamento do particulado em Lambda Zero;
[bbp] Retroespalhamento do particulado;
[Kd_QAA] Kd estimado pelo modelo QAA / Kd;
[Kd_ZEU] Kd medido com dados até a zona de atenação da luz
até 1%;
[Kd_3m] Kd medido com dados até 3 metros de profundidade;
[Kd_6m] Kd medido com dados até 6 metros de profundidade.
'''
###############################################################################
########################### QAA SETTINGS - GERAL ##############################
###############################################################################
# data_obj.Rrs = data_obj.Rrs.drop([833, 865, 945, 1374, 1614, 2202])
# data_obj.aw = data_obj.aw.drop([833, 865, 945, 1374, 1614, 2202])
# data_obj.bbw = data_obj.bbw.drop([833, 865, 945, 1374, 1614, 2202])
# Transformação do Rrs
rrs = Rrs_to_rrs(data_obj.Rrs)
Rrs = data_obj.Rrs
# Cálculando o U
g0 = 0.089
g1 = 0.1245
u = calc_u(g0, g1, rrs)
##############################################################################################################
# Filtrando os dados
wvl_ref = {
# 'referencia': [443, 492, 560, 665, 704, 741, 783], # Bandas do sensor Sentinel-2A
# 'referencia': [400,413,443,490,510,560,620,665,674,681,709,754,761,764,768,779], # Bandas do sensor OLCI
# 'referencia': [400,413,443,490,510,560,620,665,674,681,709,754,761,764,768,779,865,885,900], # Bandas do sensor OLCI
# 'referencia': [400,411,443,490,560,620,667,674,681,709,754,761,779,865,555], # Bandas v6
# 'referencia': list(range(400, 951, 1)), # Bandas v6
'referencia': list(range(400, 751, 1)), # Bandas PAR v6
'lb0': lb0
}
wl = pd.Series(wvl_ref['referencia'])
r_ref = filter_wvl(rrs, wvl_ref)
R_ref = filter_wvl(Rrs, wvl_ref)
###############################################################################
############################ FASE 1 - QAA.V6 ##################################
###############################################################################
# QAA v6 considera at_lb0 = a(670)
aw = data_obj.aw
# a_lb0_v6_492 = calc_alb0_v6_492(492, R_ref, 665, 704, data_obj.aw); lago = 'Eq. QAA.v6 AJUSTE 492 nm'
# a_lb0_v6_560 = calc_alb0_v6_560(560, R_ref, 665, 704, data_obj.aw); lago = 'Eq. QAA.v6 AJUSTE 560 nm'
# a_lb0_v6 = calc_alb0_v6(wvl_ref['lb0'], R_ref, 443, 490, data_obj.aw); lago = 'Eq. QAA.v6'
# a_lb0_v6 = aw.loc[lb0].values; lago = 'Eq. QAA.v6 aw'
# a_lb0_v6_olci = calc_alb0_v6(wvl_ref['lb0'], R_ref, 443, 490, data_obj.aw); lago = 'QAA.v6_OLCI'
a_lb0_v6_s2a = calc_alb0_v6(wvl_ref['lb0'], R_ref, 443, 492, data_obj.aw)
lago = 'QAA.v6_S2a'
## bbp (lambda_0)
# bbp_lb0_492 = calc_bbp_lb0_v6(u, data_obj.bbw, 492, a_lb0_v6_492)
# bbp_lb0_560 = calc_bbp_lb0_v6(u, data_obj.bbw, 560, a_lb0_v6_560)
bbp_lb0_665 = calc_bbp_lb0_v6(u, data_obj.bbw, 665, a_lb0_v6_s2a)
## Cálculando N
# n_acs_492 = calc_n_acs_492(r_ref[665], r_ref[704])
# n_acs_560 = calc_n_acs_560(r_ref[665], r_ref[704])
# n_hydro_560 = calc_n_hydro_560(r_ref[443], r_ref[492])
# n_hydro_704 = calc_n_hydro_704(r_ref[443], r_ref[492])
n_slope = calc_n(r_ref[443], r_ref[560])
# Realizando cálculo do BBP para todos os comprimentos de onda
# bbp_492_acs_492 = calc_bbp(wl, n_acs_492, bbp_lb0_492, 492)
# bbp_492_acs_560 = calc_bbp(wl, n_acs_560, bbp_lb0_492, 492)
# bbp_492_hydro_560 = calc_bbp(wl, n_hydro_560, bbp_lb0_492, 492)
# bbp_492_hydro_704 = calc_bbp(wl, n_hydro_704, bbp_lb0_492, 492)
# bbp_560_acs_492 = calc_bbp(wl, n_acs_492, bbp_lb0_560, 560)
# bbp_560_acs_560 = calc_bbp(wl, n_acs_560, bbp_lb0_560, 560)
# bbp_560_hydro_560 = calc_bbp(wl, n_hydro_560, bbp_lb0_560, 560)
# bbp_560_hydro_704 = calc_bbp(wl, n_hydro_704, bbp_lb0_560, 560)
bbp_665 = calc_bbp(wl, n_slope, bbp_lb0_665, 665)
# Realizando cálculo do BB para todos comprimentos de onda
# bb_492_acs_492 = calc_bb(bbp_492_acs_492, data_obj.bbw)
# bb_492_acs_560 = calc_bb(bbp_492_acs_560, data_obj.bbw)
# bb_492_hydro_560 = calc_bb(bbp_492_hydro_560, data_obj.bbw)
# bb_492_hydro_704 = calc_bb(bbp_492_hydro_704, data_obj.bbw)
# bb_560_acs_492 = calc_bb(bbp_560_acs_492, data_obj.bbw)
# bb_560_acs_560 = calc_bb(bbp_560_acs_560, data_obj.bbw)
# bb_560_hydro_560 = calc_bb(bbp_560_hydro_560, data_obj.bbw)
# bb_560_hydro_704 = calc_bb(bbp_560_hydro_704, data_obj.bbw)
bb_665 = calc_bb(bbp_665, data_obj.bbw)
# Realizando cálculo do atotal para todos os comprimentos de onda
# at_492_acs_492 = calc_a_total(bbp_492_acs_492, data_obj.bbw, u)
# at_492_acs_560 = calc_a_total(bbp_492_acs_560, data_obj.bbw, u)
# at_492_hydro_560 = calc_a_total(bbp_492_hydro_560, data_obj.bbw, u)
# at_492_hydro_704 = calc_a_total(bbp_492_hydro_704, data_obj.bbw, u)
# at_560_acs_492 = calc_a_total(bbp_560_acs_492, data_obj.bbw, u)
# at_560_acs_560 = calc_a_total(bbp_560_acs_560, data_obj.bbw, u)
# at_560_hydro_560 = calc_a_total(bbp_560_hydro_560, data_obj.bbw, u)
# at_560_hydro_704 = calc_a_total(bbp_560_hydro_704, data_obj.bbw, u)
at_665 = calc_a_total(bbp_665, data_obj.bbw, u)
###############################################################################
############################ FASE 2 - Kd ######################################
###############################################################################
m = [0.005, 4.26, 0.52, 10.54]
chi = 0.265
theta_s = calc_theta_s(data_obj.coleta)
'''
Renomear tudo de acordo!!!!!!
Reescrever o modelo e automatizar etapa dos ajustes
'''
# raz_bbw_bb_492_acs_492 = calc_raz_bbw_bb(data_obj.bbw, bb_492_acs_492)
# raz_bbw_bb_492_acs_560 = calc_raz_bbw_bb(data_obj.bbw, bb_492_acs_560)
# raz_bbw_bb_492_hydro_560 = calc_raz_bbw_bb(data_obj.bbw, bb_492_hydro_560)
# raz_bbw_bb_492_hydro_704 = calc_raz_bbw_bb(data_obj.bbw, bb_492_hydro_704)
# raz_bbw_bb_560_acs_492 = calc_raz_bbw_bb(data_obj.bbw, bb_560_acs_492)
# raz_bbw_bb_560_acs_560 = calc_raz_bbw_bb(data_obj.bbw, bb_560_acs_560)
# raz_bbw_bb_560_hydro_560 = calc_raz_bbw_bb(data_obj.bbw, bb_560_hydro_560)
# raz_bbw_bb_560_hydro_704 = calc_raz_bbw_bb(data_obj.bbw, bb_560_hydro_704)
raz_bbw_bb_665 = calc_raz_bbw_bb(data_obj.bbw, bb_665)
# kd_qaa_492_acs_492 = calc_kd_lee_2013(m, theta_s, at_492_acs_492, bb_492_acs_492, chi, raz_bbw_bb_492_acs_492)
# kd_qaa_492_acs_560 = calc_kd_lee_2013(m, theta_s, at_492_acs_560, bb_492_acs_560, chi, raz_bbw_bb_492_acs_560)
# kd_qaa_492_hydro_560 = calc_kd_lee_2013(m, theta_s, at_492_hydro_560, bb_492_hydro_560, chi, raz_bbw_bb_492_hydro_560)
# kd_qaa_492_hydro_704 = calc_kd_lee_2013(m, theta_s, at_492_hydro_704, bb_492_hydro_704, chi, raz_bbw_bb_492_hydro_704)
# kd_qaa_560_acs_492 = calc_kd_lee_2013(m, theta_s, at_560_acs_492, bb_560_acs_492, chi, raz_bbw_bb_560_acs_492)
# kd_qaa_560_acs_560 = calc_kd_lee_2013(m, theta_s, at_560_acs_560, bb_560_acs_560, chi, raz_bbw_bb_560_acs_560)
# kd_qaa_560_hydro_560 = calc_kd_lee_2013(m, theta_s, at_560_hydro_560, bb_560_hydro_560, chi, raz_bbw_bb_560_hydro_560)
# kd_qaa_560_hydro_704 = calc_kd_lee_2013(m, theta_s, at_560_hydro_704, bb_560_hydro_704, chi, raz_bbw_bb_560_hydro_704)
kd_qaa_665 = calc_kd_lee_2013(m, theta_s, at_665, bb_665, chi,
raz_bbw_bb_665)
###############################################################################
########################### Teste Kd - absorção insitu ########################
###############################################################################
'''
at_acs = data_obj.acs + aw.values
bb_acs = calc_bb_ref(u, wl, at_acs)
bb_bb = data_obj.bb
dif = set(theta_s).difference(at_acs)
theta_s_2 = theta_s.drop(dif, axis = 1)
raz_bbw_bb_acs = calc_raz_bbw_bb(data_obj.bbw, bb_acs)
raz_bbw_bb_bb = calc_raz_bbw_bb(data_obj.bbw, bb_bb)
kd_acs = calc_kd_lee_2013(m, theta_s_2, at_acs, bb_acs, chi, raz_bbw_bb_acs)
kd_bb = calc_kd_lee_2013(m, theta_s_2, at_acs, bb_bb, chi, raz_bbw_bb_bb)
'''
###############################################################################
########################### Data Extraction ###################################
###############################################################################
result_dict = {}
result_dict['dados'] = {
'lb0': wvl_ref['lb0'],
'wl': list(r_ref.keys()),
'r_ref': r_ref,
'campanhas': ano,
'equacao': lago,
'chute_inicial': 'não otimizado',
'h0h1h2': 'não otimizado',
'g0': g0,
'g1': g1
}
# result_dict['a_lb0_492'] = a_lb0_v6_492
# result_dict['a_lb0_560'] = a_lb0_v6_560
result_dict['a_lb0_665'] = a_lb0_v6_s2a
# result_dict['bb_492_acs_492'] = bb_492_acs_492
# result_dict['bb_492_acs_560'] = bb_492_acs_560
# result_dict['bb_492_hydro_560'] = bb_492_hydro_560
# result_dict['bb_492_hydro_704'] = bb_492_hydro_704
# result_dict['bb_560_acs_492'] = bb_560_acs_492
# result_dict['bb_560_acs_560'] = bb_560_acs_560
# result_dict['bb_560_hydro_560'] = bb_560_hydro_560
# result_dict['bb_560_hydro_704'] = bb_560_hydro_704
result_dict['bb_665'] = bb_665
# result_dict['at_492_acs_492'] = at_492_acs_492
# result_dict['at_492_acs_560'] = at_492_acs_560
# result_dict['at_492_hydro_560'] = at_492_hydro_560
# result_dict['at_492_hydro_704'] = at_492_hydro_704
# result_dict['at_560_acs_492'] = at_560_acs_492
# result_dict['at_560_acs_560'] = at_560_acs_560
# result_dict['at_560_hydro_560'] = at_560_hydro_560
# result_dict['at_560_hydro_704'] = at_560_hydro_704
result_dict['at_665'] = at_665
# result_dict['n_acs_492'] = n_acs_492
# result_dict['n_acs_560'] = n_acs_560
# result_dict['n_hydro_560'] = n_hydro_560
# result_dict['n_hydro_704'] = n_hydro_704
result_dict['n_665'] = n_slope
# result_dict['bbp_lb0_492'] = bbp_lb0_492
# result_dict['bbp_lb0_560'] = bbp_lb0_560
result_dict['bbp_lb0_665'] = bbp_lb0_665
# result_dict['bbp_492_acs_492'] = bbp_492_acs_492
# result_dict['bbp_492_acs_560'] = bbp_492_acs_560
# result_dict['bbp_492_hydro_560'] = bbp_492_hydro_560
# result_dict['bbp_492_hydro_704'] = bbp_492_hydro_704
# result_dict['bbp_560_acs_492'] = bbp_560_acs_492
# result_dict['bbp_560_acs_560'] = bbp_560_acs_560
# result_dict['bbp_560_hydro_560'] = bbp_560_hydro_560
# result_dict['bbp_560_hydro_704'] = bbp_560_hydro_704
result_dict['bbp_665'] = bbp_665
# result_dict['kd_qaa_492_acs_492'] = kd_qaa_492_acs_492
# result_dict['kd_qaa_492_acs_560'] = kd_qaa_492_acs_560
# result_dict['kd_qaa_492_hydro_560'] = kd_qaa_492_hydro_560
# result_dict['kd_qaa_492_hydro_704'] = kd_qaa_492_hydro_704
# result_dict['kd_qaa_560_acs_492'] = kd_qaa_560_acs_492
# result_dict['kd_qaa_560_acs_560'] = kd_qaa_560_acs_560
# result_dict['kd_qaa_560_hydro_560'] = kd_qaa_560_hydro_560
# result_dict['kd_qaa_560_hydro_704'] = kd_qaa_560_hydro_704
result_dict['kd_qaa_665'] = kd_qaa_665
# result_dict['kd_acs'] = kd_acs
# result_dict['kd_bb'] = kd_bb
result_dict['Kd_ZEU'] = data_obj.kd_zeu
result_dict['Kd_3m'] = data_obj.kd_3m
result_dict['Kd_6m'] = data_obj.kd_6m
os.makedirs('./Results_Espectral', exist_ok=True)
export_result(
result_dict, './Results_Espectral/QAAv6_' + ano + '_' +
str(wvl_ref['lb0']) + '_Ajustado.pickle')
return result_dict
def Qaa_v5_model_RUN(data_obj, ano, lb0):
'''
Função que aplica o modelo QAA v5 dado um objeto de Data Frames contendo todos
os dados necessários, uma string com o nome/ano da campanha para salvar o arquivo
dos resultados e um valor de comprimento de onda inicial (Lambda Zero).
----------
Parameters
----------
data_obj [Object]
Objeto com Data Frames contendo dados usados nas funções do modelo QAA/Kd
alocados pela função read_files através dos argumentos contendo as strings
do caminho do diretório para cada planilha '.xlsx' de dados.
O objeto de Data Frames deve conter no mínimo:
[acs] Absorção total na coluna d'água;
[Rrs] Reflectância de sensoriamento remoto superficial;
[aw] Coeficiente de absorção espectral da água;
[bbw] Coeficiente de retroespalhamento espectral da água;
[coleta] Planilha contendo Coordenadas, datas e horários das
coletas por estação;
[Kd_ZEU] Kd medido com dados até a zona de atenação da luz até 1%;
[Kd_3m] Kd medido com dados até 3 metros de profundidade;
[Kd_6m] Kd medido com dados até 6 metros de profundidade.
* Os dados devem concordar em faixa espectral, bandas e nome das estações.
ano [String]
String contendo o nome ou ano da campanha, usado para salvar arquivo
dicionário contendo todos os dados produtos no formato '.pickle'.
Ex.: 'TRM2019', '2013', 'PACO_2011', etc...
lb0 [Value]
Comprimento de onda incial usado no modelo QAA (Lambda Zero).
Ex.: 665, 560, 490, etc...
-------
Returns
-------
result_dict [Dictionary]
O resultado desta função é um dicionário contendo todos os resultados e
dados de entrada relevantes.
Além disso o modelo salva um arquivo contendo todos os dados em formato
'.pickle' na pasta '.\Results_Espectral'.
Dados salvos:
[dados/lb0] Comprimento de onda incial (Lambda Zero);
[dados/wl] Comprimentos de onda utilizados;
[dados/r_ref] Dados de rrs;
[dados/campanha] Nome/ano da campanha;
[dados/equação] Equação utilizada no modelo em questão;
[dados/g0] Coeficiente g1 - Passo 1 QAA;
[dados/g1] Coeficiente g0 - Passo 1 QAA;
[a_Lb] Absorção em Lambda Zero;
[bb] Retroespalhamento total;
[at] Absorção total estimada pelo modelo QAA;
[n] Decaimento do retroespalhamento do particulado;
[bbp_lb0] Retroespalhamento do particulado em Lambda Zero;
[bbp] Retroespalhamento do particulado;
[Kd_QAA] Kd estimado pelo modelo QAA / Kd;
[Kd_ZEU] Kd medido com dados até a zona de atenação da luz
até 1%;
[Kd_3m] Kd medido com dados até 3 metros de profundidade;
[Kd_6m] Kd medido com dados até 6 metros de profundidade.
'''
###############################################################################
########################### QAA SETTINGS - GENERAL ############################
###############################################################################
# data_obj.Rrs = data_obj.Rrs.drop([833, 865, 945, 1374, 1614, 2202])
# data_obj.aw = data_obj.aw.drop([833, 865, 945, 1374, 1614, 2202])
# data_obj.bbw = data_obj.bbw.drop([833, 865, 945, 1374, 1614, 2202])
# Transformação do Rrs
rrs = Rrs_to_rrs(data_obj.Rrs)
Rrs = data_obj.Rrs
# Cálculando o U
g0 = 0.089
g1 = 0.1245
u = calc_u(g0, g1, rrs)
###############################################################################
#ToDo: checar se isso é necessário aqui, talvez criar uma função que padronize isso fora dos OLCI_models.
# Filtrando os dados
wvl_ref = {
# 'referencia': [443, 492, 560, 665, 704, 741, 783], # Bandas do sensor Sentinel-2A
# 'referencia': [400,413,443,490,510,560,620,665,674,681,709,754,761,764,768,779], # Bandas do sensor OLCI
# 'referencia': [400,413,443,490,510,560,620,665,674,681,709,754,761,764,768,779,865,885,900], # Bandas do sensor OLCI
# 'referencia': [400,411,443,490,560,620,667,674,681,709,754,761,779,865,555], # Bandas v5
# 'referencia': list(range(400, 951, 1)), # Bandas v5
'referencia': list(range(400, 751, 1)), # Bandas PAR v5
'lb0': lb0
}
wl = pd.Series(wvl_ref['referencia'])
r_ref = filter_wvl(rrs, wvl_ref)
# R_ref = filter_wvl(Rrs, wvl_ref)
###############################################################################
############################ FASE 1 - QAA.V5 - at #############################
###############################################################################
p0 = [-1.146, -1.336,- 0.469] # Valores iniciais
# eq_lee_v5_OLCI = np.log10( (r_ref[443] + r_ref[490]) / ( r_ref['r_lb0'] + 5 * (r_ref[665]/r_ref[510] * r_ref[665]))); lago = 'QAA.v5_OLCI'
# eq_lee_v5_s2a = np.log10( (r_ref[443] + r_ref[492]) / ( r_ref['r_lb0'] + 5 * (r_ref[665]/r_ref[492] * r_ref[665]))); lago = 'QAA.v5_S2a'
eq_lee_v5 = np.log10( (r_ref[443] + r_ref[490]) / (r_ref['r_lb0'] + 5 * (r_ref[670] / r_ref[490] * r_ref[670]))); lago = 'QAA.v5'
aw = data_obj.aw
#### ATENÇÃO : a agua é adicionada ao acs dentro da função "optimizer_aLb0" ##########
# a_Lb0_optimized = optimizer_aLb0(eq_lee_v5, wvl_ref['lb0'], p0, data_obj.aw, data_obj.acs, NOToptimize = True); opt = '_OptN'
# a_Lb0_optimized = optimizer_aLb0(eq_lee_v5, wvl_ref['lb0'], p0, data_obj.aw, data_obj.acs, NOToptimize = False); opt = '_OptY'
a_Lb0_optimized = aw.loc[lb0].values; opt = '_OptAw'
#### bbp (lambda_0)
# bbp_lb0 = calc_bbp_lb0(u, data_obj.bbw, 555, wvl_ref['lb0'], a_Lb0_optimized['aref'])
bbp_lb0 = calc_bbp_lb0(u, data_obj.bbw, wvl_ref['lb0'], a_Lb0_optimized)
## Cálculando N
n = calc_n(r_ref[443], r_ref[555])
# Realizando cálculo do BBP para todos os comprimentos de onda
bbp = calc_bbp(wl, n, bbp_lb0, wvl_ref['lb0'])
# Realizando cálculo do BB para todos comprimentos de onda
bb = calc_bb(bbp, data_obj.bbw)
# Realizando cálculo do atotal para todos os comprimentos de onda
at = calc_a_total_nw(bbp, data_obj.bbw, u)
###############################################################################
############################ FASE 2 - Kd ######################################
###############################################################################
m = [0.005, 4.26, 0.52, 10.54]
chi = 0.265
theta_s = calc_theta_s(data_obj.coleta)
raz_bbw_bb = calc_raz_bbw_bb(data_obj.bbw, bb)
kd_qaa = calc_kd_lee_2013(m, theta_s, at, bb, chi, raz_bbw_bb)
###############################################################################
########################### Data Extraction ###################################
###############################################################################
result_dict = {}
result_dict['dados'] = {
'lb0': wvl_ref['lb0'],
'wl': list(r_ref.keys()),
'r_ref': r_ref,
'campanhas': ano,
'equacao': lago,
# 'chute_inicial': a_Lb0_optimized['chute_inicial'],
# 'h0h1h2': a_Lb0_optimized['h_otimizado'],
'g0': g0,
'g1': g1
}
# result_dict['a_Lb'] = a_Lb0_optimized['aref']
result_dict['a_Lb'] = a_Lb0_optimized
result_dict['bb'] = bb
result_dict['at'] = at
result_dict['n'] = n
result_dict['bbp_lb0'] = bbp_lb0
result_dict['bbp'] = bbp
result_dict['Kd_QAA'] = kd_qaa
result_dict['Kd_ZEU'] = data_obj.kd_zeu
result_dict['Kd_3m'] = data_obj.kd_3m
result_dict['Kd_6m'] = data_obj.kd_6m
os.makedirs('./Results_Espectral', exist_ok = True)
export_result(result_dict, './Results_Espectral/QAAv5_' + ano + '_' + str(wvl_ref['lb0']) + opt + '.pickle')
return result_dict
def Qaa_v6_model_RUN(data_obj, ano, lb0):
"""
TODO: WRITE DESCRIPTION AND CREATE A PATERN TO THIS function
"""
###############################################################################
########################### QAA SETTINGS - GERAL ##############################
###############################################################################
data_obj.Rrs = data_obj.Rrs.drop([833, 865, 945, 1374, 1614, 2202])
data_obj.aw = data_obj.aw.drop([833, 865, 945, 1374, 1614, 2202])
data_obj.bbw = data_obj.bbw.drop([833, 865, 945, 1374, 1614, 2202])
# Transformação do Rrs
rrs = Rrs_to_rrs(data_obj.Rrs)
Rrs = data_obj.Rrs
# Cálculando o U
g0 = 0.089
g1 = 0.1245
u = calc_u(g0, g1, rrs)
##############################################################################################################
#ToDo: checar se isso é necessário aqui, talvez criar uma função que padronize isso fora dos OLCI_models.
# Filtrando os dados
wvl_ref = {
'referencia': [443, 492, 560, 665, 704, 741,
783], # Bandas do sensor Sentinel-2A
# 'referencia': [400,413,443,490,510,560,620,665,674,681,709,754,761,764,768,779], # Bandas do sensor OLCI
# 'referencia': [400,413,443,490,510,560,620,665,674,681,709,754,761,764,768,779,865,885,900], # Bandas do sensor OLCI
# 'referencia': [400,411,443,490,560,620,667,674,681,709,754,761,779,865,555],# Bandas v5
'lb0': lb0
}
wl = pd.Series(wvl_ref['referencia'])
r_ref = filter_wvl(rrs, wvl_ref)
R_ref = filter_wvl(Rrs, wvl_ref)
###############################################################################
############################ FASE 1 - QAA.V6 ##################################
###############################################################################
# QAA v6 considera at_lb0 = a(670)
aw = data_obj.aw
# a_lb0_v6 = calc_alb0_v6(670, R_ref, 443, 490, data_obj.aw); lago = 'Eq. QAA.v6'
# a_lb0_v6_olci = calc_alb0_v6(wvl_ref['lb0'], R_ref, 443, 490, data_obj.aw); lago = 'QAA.v6_OLCI'
a_lb0_v6_s2a = calc_alb0_v6(wvl_ref['lb0'], R_ref, 443, 492, data_obj.aw)
lago = 'QAA.v6_S2a'
#### bbp (lambda_0)
bbp_lb0 = calc_bbp_lb0_v6(u, data_obj.bbw, 665, wvl_ref['lb0'],
a_lb0_v6_s2a)
## Cálculando N
n = calc_n(r_ref[443], r_ref[560])
# Realizando cálculo do BBP para todos os comprimentos de onda
bbp = calc_bbp(wl, n, bbp_lb0, wvl_ref['lb0'])
# Realizando cálculo do BB para todos comprimentos de onda
bb = calc_bb(bbp, data_obj.bbw)
# Realizando cálculo do atotal para todos os comprimentos de onda
at = calc_a_total(bbp, data_obj.bbw, u)
###############################################################################
############################ FASE 2 - Kd ######################################
###############################################################################
m = [0.005, 4.26, 0.52, 10.54]
chi = 0.265
theta_10 = 10
theta_15 = 15
raz_bbw_bb = calc_raz_bbw_bb(data_obj.bbw, bb)
Kd_10 = calc_kd_lee_2013(m, theta_10, at, bb, chi, raz_bbw_bb)
Kd_15 = calc_kd_lee_2013(m, theta_15, at, bb, chi, raz_bbw_bb)
###############################################################################
########################### Data Extraction ###################################
###############################################################################
result_dict = {}
result_dict['dados'] = {
'lb0': wvl_ref['lb0'],
'r_ref': list(r_ref.keys()),
'campanhas': ano,
'equacao': lago,
'chute_inicial': 'não otimizado',
'h0h1h2': 'não otimizado',
'g0': g0,
'g1': g1
}
result_dict['a_Lb'] = a_lb0_v6_s2a
result_dict['bb'] = bb
result_dict['at'] = at
result_dict['n'] = n
result_dict['bbp_lb0'] = bbp_lb0
result_dict['bbp'] = bbp
result_dict['Kd_10'] = Kd_10
result_dict['Kd_15'] = Kd_15
result_dict['Kd_ZEU'] = data_obj.Kd_ZEU
result_dict['Kd_3m'] = data_obj.Kd_3m
result_dict['Kd_6m'] = data_obj.Kd_6m
os.makedirs('./Results_S2A', exist_ok=True)
export_result(
result_dict, './Results_S2A/QAAv6_' + ano + '_S2a_' +
str(wvl_ref['lb0']) + '.pickle')
return result_dict
