#! /usr/bin/env python

#############################################################################

#

# Program : create_HadISST_CMIP5_syn_SSTs.py

# Author : Neil Massey

# Purpose : Create a single synthetic SST from everything that has been

# computed before hand

# Inputs : run_type : rcp4.5 | rc8.5 | histo

# ref_start : year to start reference period, 1850->2005

# ref_end : year to end reference period, 1850->2005

# run_type : historical | rcp45 | rcp85

# Notes : all reference values are calculated from the historical run_type

# CMIP5 ensemble members are only included if their historical run

# includes the reference period

# requires Andrew Dawsons eofs python libraries:

# http://ajdawson.github.io/eofs/

# Output : in the output/ directory filename is:

#

# Date : 18/02/15

#

#############################################################################

import os, sys, getopt

from calc_CMIP5_EOFs import get_cmip5_EOF_filename, get_cmip5_proj_PC_scale_filename

from create_CMIP5_syn_PCs import get_syn_SST_PCs_filename

from create_CMIP5_sst_anoms import get_concat_anom_sst_ens_mean_smooth_fname, get_start_end_periods

from create_HadISST_sst_anoms import get_HadISST_smooth_fname, get_HadISST_monthly_reference_fname, get_HadISST_month_smooth_filename, get_HadISST_monthly_residuals_fname

from cmip5_functions import load_data, reconstruct_field, calc_GMSST, load_sst_data

from calc_HadISST_residual_EOFs import get_HadISST_monthly_residual_EOFs_fname, get_HadISST_monthly_residual_PCs_fname

import numpy

from scipy.io.netcdf import *

from ARN import ARN

from eofs.standard import Eof

import pyximport

pyximport.install(setup_args={'include_dirs':[numpy.get_include()]})

from zonal_smoother import *

#############################################################################

def get_SST_output_directory(run_type, ref_start, ref_end, eof_year):

return out_dir

#############################################################################

def get_syn_sst_filename(run_type, ref_start, ref_end, neofs, eof_year, percentile, intvarmode, monthly):

return spath + out_name

#############################################################################

def create_monthly_intvar(run_type, ref_start, ref_end, n_pcs=22, run_n=400):

return monthly_intvar

#############################################################################

def save_3d_file(out_fname, out_data, out_lon_var, out_lat_var, out_attrs, t_data, t_attrs, out_vname="sst"):

out_fh.close()

#############################################################################

def fit_mean_to_likely(ens_mean, monthly=True):

return ens_mean

#############################################################################

# The synthetic SSTs have two distinct time periods:

# 1. The HadISST2 period (1899->2010)

# 2. The CMIP5 period (2006->2100)

#

# To create the SSTs we create the SSTs for each individual time period and

# then interpolate between the two for the overlapping period (2006->2100)

#

# The methods used to generate each one can be decomposed into:

# (HadISST period) HadISST monthly smoothed data +

# HadISST internal variability (statistically generated)

# (CMIP5 period) CMIP5 monthly smoothed ensemble mean +

# CMIP5 monthly smoothed long term trend (anomalies from the ensemble mean) +

# HadISST 1986->2005 yearly reference pattern +

# HadISST internal variability +

# HadISST annual cycle

#

# To avoid discontinuity of the internal variability, we generate one long

# timeseries of this to add onto the results at the end

# functions follow to generate each individual component

#############################################################################

def create_hadisst_long_term_timeseries(monthly=True, run_n=400):

return hadisst_sst

#############################################################################

def correct_cmip5_long_term_mean_timeseries(cmip5_ts, monthly=True, run_n=400):

return cmip5_ts + cmip5_ens_mean_correction

#############################################################################

def create_cmip5_long_term_mean_timeseries(run_type, ref_start, ref_end, monthly=True, run_n=400):

return cmip5_ens_mean

#############################################################################

def create_cmip5_rcp_anomalies(run_type, ref_start, ref_end, eof_year, percentile, monthly=True):

return syn_sst_rcp

#############################################################################

def create_hadisst_monthly_reference(run_type, ref_start, ref_end, n_repeats, run_n=400):

return mon_smooth_ref_tile

#############################################################################

def create_hadisst_cmip5_long_term_timeseries(hadisst_ts, cmip5_ts, monthly=True):

return out_data.astype(numpy.float32)

#############################################################################

def save_syn_SSTs(out_data, run_type, ref_start, ref_end, neofs, eof_year, sample, intvarmode, monthly):

print out_name

#############################################################################

def create_syn_SSTs(run_type, ref_start, ref_end, neofs, eof_year, sample, intvarmode, var_eofs, monthly):

save_syn_SSTs(out_data, run_type, ref_start, ref_end, neofs, eof_year, sample, intvarmode, monthly)

#############################################################################

if __name__ == "__main__":

# defaults arguments only

intvarmode = 0 # internal variability mode - 0 = none, 1 = yearly, 2 = monthly

monthly = False # use the monthly EOFs / PCs ?

opts, args = getopt.getopt(sys.argv[1:], 'r:s:e:n:f:a:i:v:m',

['run_type=', 'ref_start=', 'ref_end=', 'neofs=', 'eof_year=',

'sample=', 'intvarmode=',

'varneofs=', 'monthly'])

for opt, val in opts:

if opt in ['--run_type', '-r']:

run_type = val

if opt in ['--ref_start', '-s']:

ref_start = int(val)

if opt in ['--ref_end', '-e']:

ref_end = int(val)

if opt in ['--neofs', '-n']:

neofs = int(val)

if opt in ['--eof_year', '-f']:

eof_year = int(val)

if opt in ['--sample', '-a']:

sample = int(val)

if opt in ['--intvar', '-i']:

intvarmode = int(val)

if opt in ['--varneofs', '-v']:

var_eofs = int(val)

if opt in ['--monthly', '-m']:

monthly = True

create_syn_SSTs(run_type, ref_start, ref_end, neofs, eof_year, sample, intvarmode, var_eofs, monthly)