fig9.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Figure 9: ECS calculation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "from matplotlib.pyplot import *\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "from netCDF4 import Dataset\n",
    "from scipy import stats\n",
    "import ols,glob\n",
    "from collections import namedtuple"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "printf=False # set to true to write out plot files"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "def anomaly(data):\n",
    "\t\"\"\"calculate anomaly of a 1D array (assumes no missing time slots)\"\"\"\n",
    "\td1=np.ma.copy(data)\n",
    "\tind=np.ma.array(range(len(data))) % 12\n",
    "\tfor num in range(12):\n",
    "\t\td1[num == ind]=data[num == ind]-np.ma.average(data[num == ind])\n",
    "\treturn d1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "## make plots pretty\n",
    "def colorplot(ax):\n",
    "    \n",
    "    almost_black = '#262626' # tweek from Thorsten\n",
    "    \n",
    "    for spx in ['top','right']:\n",
    "            ax.spines[spx].set_visible(False)\n",
    "\n",
    "    # Only show ticks on the left and bottom spines\n",
    "    ax.yaxis.set_ticks_position('left')\n",
    "    ax.xaxis.set_ticks_position('bottom')\n",
    "        \n",
    "    for spx in ['bottom','left']:\n",
    "        ax.spines[spx].set_linewidth(0.5)\n",
    "        ax.spines[spx].set_color(almost_black)\n",
    "    \n",
    "    for ticks in ax.xaxis.get_ticklines() + ax.yaxis.get_ticklines():\n",
    "        ticks.set_color(almost_black)\n",
    "\n",
    "def colorplotRight(ax):\n",
    "    \n",
    "    almost_black = '#262626' # tweek from Thorsten\n",
    "    \n",
    "    for spx in ['top','left']:\n",
    "            ax.spines[spx].set_visible(False)\n",
    "\n",
    "    # Only show ticks on the left and bottom spines\n",
    "    ax.yaxis.set_ticks_position('right')\n",
    "    ax.xaxis.set_ticks_position('bottom')\n",
    "    \n",
    "    # set axis positions\n",
    "#     ax.yaxis.set_label_position(pspines[0]) \n",
    "#     ax.xaxis.set_label_position(pspines[1]) \n",
    "    \n",
    "    for spx in ['bottom','right']:\n",
    "        ax.spines[spx].set_linewidth(0.5)\n",
    "        ax.spines[spx].set_color(almost_black)\n",
    "    \n",
    "    for ticks in ax.xaxis.get_ticklines() + ax.yaxis.get_ticklines():\n",
    "        ticks.set_color(almost_black)\n",
    "    \n",
    "    ax.yaxis.set_label_position('right')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "number of runs in Monte Carlo calculations: 25,000\n"
     ]
    }
   ],
   "source": [
    "## monte carlo calculations\n",
    "numRuns=25000 # number of runs of the monte carlo calculation\n",
    "print('number of runs in Monte Carlo calculations: {:,}'.format(numRuns))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "def monteCarlo1(x1,y1,runs=10000,autocorr=False):\n",
    "    \"\"\"x1 and y1 are the input arrays, runs is the number of times to do the calculation,\n",
    "    returns list of slopes\n",
    "    if autocorr is True, then the number of samples is reduced; it is set equal to the \n",
    "    effective number of degrees of freedom, per Santer et al. 2000\n",
    "    \"\"\"\n",
    "\n",
    "    # calculate neff; effective number of independent samples\n",
    "    # follows Santer et al. 2000\n",
    "    lag1cov = lambda a1: (np.average(a1[1:]*a1[:-1])/np.std(a1)**2)\n",
    "    sl0=stats.linregress(x1,y1)\n",
    "    yhat=y1-(sl0[0]*x1+sl0[1])\n",
    "    lag1=lag1cov(yhat)\n",
    "    if not autocorr: lag1=0\n",
    "    neff=len(x1)*(1-lag1)/(1+lag1)\n",
    "\n",
    "    slope=[];intercept=[];newy=[]\n",
    "    for ii in range(runs):\n",
    "        \n",
    "        ## regress p1 vs. r2\n",
    "        rind=np.random.randint(0,high=len(x1),size=int(neff))\n",
    "        fit1=stats.linregress(x1[rind],y1[rind])\n",
    "        slope.append(fit1[0]);intercept.append(fit1[1])\n",
    "    \n",
    "    return np.array(slope),np.array(intercept)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## read in CMIP5 data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ACCESS1-0 ACCESS1-3 BCC-CSM1-1 BCC-CSM1-1-M CCSM4 CNRM-CM5 CNRM-CM5-2 CanESM2 FGOALS-g2 FGOALS-s2 GFDL-CM3 GFDL-ESM2G GFDL-ESM2M GISS-E2-H GISS-E2-R INMCM4 IPSL-CM5A-LR IPSL-CM5A-MR IPSL-CM5B-LR MIROC5 MPI-ESM-LR MPI-ESM-MR MPI-ESM-P MRI-CGCM3 NorESM1-M \n"
     ]
    }
   ],
   "source": [
    "modelRun = namedtuple('modelRun', ('ts','t_trop','flux','lwclr','swclr','lwcre','swcre',\\\n",
    "    'lw','sw','clr','cre','time'))\n",
    "modelsx = namedtuple('modelsx', ('abrupt4xCO2','piControl'))\n",
    "\n",
    "# get names of models\n",
    "fn=glob.glob('cmip5/*.{}.nc'.format('piControl'))\n",
    "models=[ii.split('/')[1] for ii in fn]\n",
    "models=[ii.split('.')[0] for ii in models]\n",
    "models.sort()\n",
    "\n",
    "# these models have screwy ratios of lambda(p)/lambda(a)\n",
    "models.remove('HadGEM2-ES') \n",
    "models.remove('MIROC-ESM')\n",
    "models.remove('CSIRO-Mk3-6-0')\n",
    "\n",
    "modelOutput={}\n",
    "for mn in models:\n",
    "    f1=[]\n",
    "    for modrun in ['abrupt4xCO2','piControl']:\n",
    "        ## read nc file\n",
    "        with Dataset('cmip5/{0}.{1}.nc'.format(mn,modrun)) as fni:\n",
    "            ovar=[]\n",
    "            for ii in ['flux','ta','tas','t_trop','tas_trop','rlutcs','rsutcs','rlut','rsut','rsdt']:\n",
    "                try:\n",
    "                    ovar.append(fni[ii][:].squeeze())\n",
    "                except IndexError:\n",
    "                    ovar.append(0)\n",
    "            \n",
    "        feedbacks=modelRun(flux=ovar[0],ts=ovar[2],t_trop=ovar[3],lwclr=-ovar[5],\\\n",
    "            swclr=ovar[9]-ovar[6],lwcre=-(ovar[7]-ovar[5]),swcre=ovar[9]-ovar[8]-(ovar[9]-ovar[6]),\\\n",
    "            sw=ovar[9]-ovar[8],lw=-ovar[7],time=np.arange(len(ovar[0])),\\\n",
    "            clr=-ovar[5]+ovar[9]-ovar[6],cre=-(ovar[7]-ovar[5])+ovar[9]-ovar[8]-(ovar[9]-ovar[6]))\n",
    "        \n",
    "        f1.append(feedbacks)\n",
    "        \n",
    "    print('{} '.format(mn),end=\"\")\n",
    "    sys.stdout.flush()\n",
    "    modelOutput[mn]=modelsx(abrupt4xCO2=f1[0],piControl=f1[1])\n",
    "\n",
    "print \n",
    "print()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### calculate feedbacks in CMIP5 models"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ACCESS1-0 ACCESS1-3 BCC-CSM1-1 BCC-CSM1-1-M CCSM4 CNRM-CM5 CNRM-CM5-2 CanESM2 FGOALS-g2 FGOALS-s2 GFDL-CM3 GFDL-ESM2G GFDL-ESM2M GISS-E2-H GISS-E2-R INMCM4 IPSL-CM5A-LR IPSL-CM5A-MR IPSL-CM5B-LR MIROC5 MPI-ESM-LR MPI-ESM-MR MPI-ESM-P MRI-CGCM3 NorESM1-M "
     ]
    }
   ],
   "source": [
    "### values are calculated from all 150 years of the run\n",
    "### feedbacks is dictionary containing the values for dR/dT from the models\n",
    "### forcingts is the forcing from regression vs. Ts\n",
    "### forcingta is the forcing from regression vs. Ta\n",
    "\n",
    "feedbacks={};feedbacks_20_150={}\n",
    "forcingts={};forcingta={};forcingts_20_150={}\n",
    "\n",
    "# loop through models\n",
    "for mn in models:\n",
    "    # loop over run types\n",
    "    feedbacks[mn]={};feedbacks_20_150[mn]={}\n",
    "    for runType in ['abrupt4xCO2','piControl']:\n",
    "        # loop through radiation variables\n",
    "        feedbacks[mn][runType]={}\n",
    "        feedbacks_20_150[mn][runType]={}\n",
    "\n",
    "        varx='flux' \n",
    "        \n",
    "        # loop through temperature variables\n",
    "        feedbacks[mn][runType][varx]={}\n",
    "        feedbacks_20_150[mn][runType][varx]={}\n",
    "        for tx in ['ts','t_trop']:\n",
    "            # calculate feedback for model, runtype, flux variable, surface T variable\n",
    "\n",
    "            exec('temp=np.copy(modelOutput[mn].{}.{})'.format(runType,tx))\n",
    "            exec('flux=np.copy(modelOutput[mn].{}.{})'.format(runType,varx))\n",
    "\n",
    "            if runType == 'abrupt4xCO2':\n",
    "\n",
    "                exec('control=modelOutput[mn].{}.{}'.format('piControl',tx))\n",
    "                temp -= np.average(control) # subtract average control temperature\n",
    "                temp=temp[:len(temp)//12*12];temp=temp.reshape(len(temp)//12,12).mean(axis=1)\n",
    "\n",
    "                control=modelOutput[mn].piControl.flux\n",
    "                flux -= np.average(control)\n",
    "                flux=flux[:len(flux)//12*12];flux=flux.reshape(len(flux)//12,12).mean(axis=1)\n",
    "\n",
    "                fb=ols.ols(flux[:150],temp[:150])\n",
    "                feedbacks[mn][runType][varx][tx]=fb.b[1]\n",
    "                \n",
    "                fb=ols.ols(flux[20:150],temp[20:150])\n",
    "                feedbacks_20_150[mn][runType][varx][tx]=fb.b[1]\n",
    "                \n",
    "                # calculate forcing as y-intercept of R vs. T regression\n",
    "                if tx == 'ts':\n",
    "                    \n",
    "                    fb=ols.ols(flux[:150],temp[:150])\n",
    "                    forcingts[mn]=fb.b[0]/2\n",
    "                    \n",
    "                    fb=ols.ols(flux[20:150],temp[20:150])\n",
    "                    forcingts_20_150[mn]=fb.b[0]/2\n",
    "\n",
    "                else:\n",
    "\n",
    "                    fb=ols.ols(flux[:150],temp[:150])\n",
    "                    forcingta[mn]=fb.b[0]/2\n",
    "                    \n",
    "            else:\n",
    "                # split time series into individual time series and create feedback for each one, then average \n",
    "                length=17*12\n",
    "                temp=temp[:len(temp)//length*length].reshape(len(temp)//length,length)\n",
    "                flux=flux[:len(flux)//length*length].reshape(len(flux)//length,length)\n",
    "                temp=anomaly(temp.T).T;flux=anomaly(flux.T).T\n",
    "\n",
    "                slope=[]\n",
    "                for t1,f1 in zip(temp,flux):\n",
    "                    slope.append(stats.linregress(t1,f1).slope)\n",
    "                feedbacks[mn][runType][varx][tx]=np.average(slope)\n",
    "                    \n",
    "    print('{} '.format(mn),end=\"\")\n",
    "    sys.stdout.flush()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### calculate ECS"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ACCESS1-0 ACCESS1-3 BCC-CSM1-1 BCC-CSM1-1-M CCSM4 CNRM-CM5 CNRM-CM5-2 CanESM2 FGOALS-g2 FGOALS-s2 GFDL-CM3 GFDL-ESM2G GFDL-ESM2M GISS-E2-H GISS-E2-R INMCM4 IPSL-CM5A-LR IPSL-CM5A-MR IPSL-CM5B-LR MIROC5 MPI-ESM-LR MPI-ESM-MR MPI-ESM-P MRI-CGCM3 NorESM1-M "
     ]
    }
   ],
   "source": [
    "ecs2={};forcing2={}\n",
    "runType = 'abrupt4xCO2';varx='flux';tx='ts'\n",
    "# loop through models\n",
    "for mn in models:\n",
    "    # loop over run types\n",
    "\n",
    "    temp=np.copy(modelOutput[mn].abrupt4xCO2.ts)\n",
    "    control=modelOutput[mn].piControl.ts\n",
    "    temp -= np.average(control) # subtract average control temperature\n",
    "    temp=temp[:len(temp)//12*12];temp=temp.reshape(len(temp)//12,12).mean(axis=1)\n",
    "\n",
    "    flux=np.copy(modelOutput[mn].abrupt4xCO2.flux)\n",
    "    control=modelOutput[mn].piControl.flux\n",
    "    flux -= np.average(control)\n",
    "#     print np.average(control)\n",
    "    flux=flux[:len(flux)//12*12];flux=flux.reshape(len(flux)//12,12).mean(axis=1)\n",
    "\n",
    "    fb=ols.ols(flux[:150],temp[:150]);fb.est_auto()\n",
    "    ecs2[mn]=-fb.b[0]/fb.b[1]/2\n",
    "    forcing2[mn]=fb.b[0]/2\n",
    "\n",
    "    print('{} '.format(mn),end=\"\")\n",
    "    sys.stdout.flush()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "## load CERES data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Do two different slope calculations: 1) regress detrended R and T and 2) use estimate of F to regress R-F and Ts"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/anaconda/lib/python3.6/site-packages/ipykernel_launcher.py:5: UserWarning: WARNING: valid_min not used since it\n",
      "cannot be safely cast to variable data type\n",
      "  \"\"\"\n",
      "/usr/local/anaconda/lib/python3.6/site-packages/ipykernel_launcher.py:5: UserWarning: WARNING: valid_max not used since it\n",
      "cannot be safely cast to variable data type\n",
      "  \"\"\"\n"
     ]
    }
   ],
   "source": [
    "## calculate slope of detrended data\n",
    "ovar=[]\n",
    "with Dataset('../energyBalance/ceres_era.nc') as fni:\n",
    "    for varx in ['gtoa_net_all_mon','t_trop','t2m']:\n",
    "        ovar.append(anomaly(fni[varx][:].squeeze()))\n",
    "\n",
    "        # detrend the data\n",
    "        fitx=ols.ols(ovar[-1],np.arange(len(ovar[-1])))\n",
    "        ovar[-1] = ovar[-1] - np.arange(len(ovar[-1]))*fitx.b[1]\n",
    "        ovar[-1]=anomaly(ovar[-1])\n",
    "\n",
    "CERESdetrendedFlux = ovar[0]\n",
    "CERESdetrendedTa = ovar[1]\n",
    "CERESdetrendedTs = ovar[2]\n",
    "timeCERES=np.arange(len(CERESdetrendedFlux))/12.+2000+2/12.+1/24."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### identify models that agree with CERES and those that don't"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "First, vs. $\\Theta$"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "numSortRuns=numRuns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "there are 15 \"good\" models and 10 others\n"
     ]
    }
   ],
   "source": [
    "# find \"good\"/\"bad\" models that agree (and not) with CERES data\n",
    "# compare slope of detrended R vs. detrended Ta\n",
    "slope=monteCarlo1(CERESdetrendedTa,CERESdetrendedFlux,runs=numSortRuns,autocorr=True)[0]\n",
    "\n",
    "modellist = namedtuple('modellist', ('theta_detrended','lambda_detrended'))\n",
    "slope_list=modellist(theta_detrended=slope,lambda_detrended=0)\n",
    "\n",
    "low,high=np.percentile(slope,5),np.percentile(slope,95) # 5-95% uncertainty\n",
    "avgceres=np.average(slope)\n",
    "\n",
    "modname=[ii for ii in feedbacks.keys()]\n",
    "p=np.array([feedbacks[ii]['piControl']['flux']['t_trop'] for ii in modname])\n",
    "\n",
    "good=[];bad=[] # store names of good/bad models\n",
    "\n",
    "for mm,vv in zip(modname,p):\n",
    "    if (vv > low) & (vv <= high): \n",
    "        good.append(mm)\n",
    "    else:\n",
    "        bad.append(mm)\n",
    "\n",
    "print('there are {} \"good\" models and {} others'.format(len(good),len(bad)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Second, vs. $\\lambda$"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "there are 22 \"good\" models and 3 others\n"
     ]
    }
   ],
   "source": [
    "# find \"good\"/\"bad\" models that agree (and not) with CERES data\n",
    "# compare slope of detrended R vs. detrended Ts\n",
    "slope=monteCarlo1(CERESdetrendedTs,CERESdetrendedFlux,runs=numSortRuns,autocorr=True)[0]\n",
    "slope_list=modellist(theta_detrended=slope_list.theta_detrended,lambda_detrended=slope)\n",
    "\n",
    "low,high=np.percentile(slope,5),np.percentile(slope,95) # 5-95% uncertainty\n",
    "avgceres=np.average(slope)\n",
    "\n",
    "modname=[ii for ii in feedbacks.keys()]\n",
    "p=np.array([feedbacks[ii]['piControl']['flux']['ts'] for ii in modname])\n",
    "\n",
    "good3=[];bad3=[] # store names of good/bad models\n",
    "\n",
    "for mm,vv in zip(modname,p):\n",
    "    if (vv > low) & (vv <= high): \n",
    "        good3.append(mm)\n",
    "    else:\n",
    "        bad3.append(mm)\n",
    "\n",
    "print('there are {} \"good\" models and {} others'.format(len(good3),len(bad3)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "### save names of good and bad model in named tuple\n",
    "good_list=modellist(theta_detrended=good,lambda_detrended=good3)\n",
    "bad_list=modellist(theta_detrended=bad,lambda_detrended=bad3)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### print stats of $\\Theta$ and $\\lambda$ from CERES + ERAi data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "slope +/- 5-95% CI: -1.09 +/- -1.37 - -0.80\n",
      "upper, lower error bars: 0.29, 0.28\n"
     ]
    }
   ],
   "source": [
    "# theta_detrended: calculate slope of detrended R vs. detrended Ta\n",
    "slope=slope_list.theta_detrended\n",
    "avgslope=np.median(slope);p5=np.percentile(slope,5);p95=np.percentile(slope,95)\n",
    "print('slope +/- 5-95% CI: {:.2f} +/- {:.2f} - {:.2f}'.format(avgslope,p5,p95))\n",
    "print('upper, lower error bars: {:.2f}, {:.2f}'.format(p95-avgslope,avgslope-p5))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "slope +/- 5-95% CI: -0.81 +/- -1.45 - -0.15\n",
      "upper, lower error bars: 0.66, 0.64\n"
     ]
    }
   ],
   "source": [
    "# lambda_detrended: calculate slope of detrended R vs. detrended Ts\n",
    "slope=slope_list.lambda_detrended\n",
    "avgslope=np.median(slope);p5=np.percentile(slope,5);p95=np.percentile(slope,95)\n",
    "print('slope +/- 5-95% CI: {:.2f} +/- {:.2f} - {:.2f}'.format(avgslope,p5,p95))\n",
    "print('upper, lower error bars: {:.2f}, {:.2f}'.format(p95-avgslope,avgslope-p5))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "average ECS of the good models: 2.90\n",
      "std of 0.48\n",
      "range of 2.04 to 3.86\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAgcAAAHOCAYAAAAMrqraAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvNQv5yAAAIABJREFUeJzs3XucU9XV//HPEoVBsdxUULBVi7bqCOqgBbUo1Sq1FWi91VbEWjuKVpFqn6JVvEy1WlSsj5eKF4p4Aav1EYqPgAql/AAFRMvghQL61JEBFREFHAZh/f5IMp3MZGZyOcmcJN/365UXyTk7K2tiTNbZZ5+9zd0RERERidmptRMQERGRcFFxICIiInFUHIiIiEgcFQciIiISR8WBiIiIxFFxICIiInFUHIiIiEgcFQciIiISR8WBiIiIxNm5tRNoLXvssYfvt99+zbbZvHkzu+22W8avpTj5kYvi5CZOmHJRnNzECVMuxR5nyZIlH7v7ni0Gc/eivJWVlXlLZs+e3WKbZChOdmMoTn7FCVMuipObOGHKpdjjAIs9id/IvDitYGaDzOwdM1tpZqMT7G9nZlOi+18xs/1yn6WIiEhhCH1xYGZtgHuB7wGHAOeY2SENmv0c2ODuvYBxwG25zVJERKRwhL44AI4GVrr7anevBSYDQxq0GQJMjN5/GjjRzCyHOYqIiBSMfCgOegDv13tcFd2WsI27fwlsBLrmJDsREZECY5HxCeFlZmcCp7j7hdHHw4Cj3f2yem2WR9tURR+virZZ3yBWOVAO0K1bt7LJkyc3+9qbNm2iQ4cOGf8NipMfuShObuKEKRfFyU2cMOVS7HEGDhy4xN37thgsmVGLrXkD+gMz6j2+Gri6QZsZQP/o/Z2Bj4kWPk3ddLVCOOKEKRfFyU2cMOWiOLmJE6Zcij0OBXS1wiLgQDPb38zaAj8GpjZoMxUYHr1/BvBy9E0QERGRFIV+EiR3/9LMfkmkd6AN8Ii7Lzezm4hUQFOBh4FJZrYS+IRIASEiIiJpCH1xAODuzwPPN9g2pt79GuDMXOclIiJSiPLhtIKIiIjkkIoDERERiaPiIMsuuOACBg8ezNq1a5N+zrPPPkv37t3ZtGlTFjMTERFJTMVBFq1atYqJEyfy+eefU1FRkfTzhg4dSvfu3Rk7dmwWsxMREUlMxUEW3XzzzbF5GJgwYULSvQdmRnl5Offccw81NTXZTFFERKQRFQdZsnXrVh5//HF22inyFm/fvp3rr7+eUaNGUVpaSocOHejevTunnXYab7/9dqPnn3XWWXz66af89a9/zXXqIiJS5FQcZMn06dOpra1l+/btANTW1vLoo4/y4Ycfcu211zJ9+nTuv/9+ampq6NevX6NehT322IODDz6YF154oTXSFxGRIpYX8xzko0TjBXbs2EGnTp348Y8jczRt376dU045hW7duvHkk08yatSouPZHHHEECxcuzEm+IiIiMeo5yILq6moWLVrUaHttbS0PPvggRx55JJ06dWLnnXdmt912Y9OmTbzzzjuN2u+5556sWbMmFymLiIjUUc9BFlRUVNDU0g7btm1jy5YtPPHEE+yxxx7stNNOnHrqqQkHHrZv314DEkVEJOdUHGTBggUL2LFjR5P727dvz6mnngpEioVPPvkkYbtPPvmErl27ZiVHERGRpqR9WsHM2kVXSjzEzPYMMql8t3TpUiZOnAjA+++/z+zZs3F3hg4dyje/+U2WLl1a13bSpEl1gxYbevfdd/nGN76Rk5xFRERiUioOzGx3MxthZnOBjcBKoBJYa2bvm9mDZnZUNhLNNwMGDADg1Vdfrds2aNAg3n77bUaNGsVLL73EH/7wB8aMGUOnTp0aPd/dWbRoUV0cERGRXEm6ODCzUcB7wAXALGAIcDhwENAfuJ7IaYpZZvaCmR0YeLZ5ZL/99uPoo49m2rRpddt+8Ytf8Nvf/pYpU6Zw2mmnMX36dKZNm0bHjh0bPX/+/Pl88skndVc2iIiI5EoqYw6OAY5398om9r8KPGJmFwM/B44H/pVhfnltxIgRjBw5krPPPhuAnXbaid/97nf87ne/i2v33nvvNXruY489xnHHHUdpaWkuUhURkVZQXV3NyJEjmTFjBt27d2/tdOok3XPg7mcCpyXRbqu73+fuD2WUWQEYNmwYPXr04LnnnkvpeWvXrmXixIncfPPNWcpMRETCoKKigmXLlqW0/k4upDog8VozO6m5Bma2fwb5FJQ2bdrwyCOP0K5du5Se995773HHHXdkZbxBrEpNZZVIEREJXnV1NRMmTMDdU1p/JxdSLQ6uAJ40s68m2mlmxwKa0q+efv36MXTo0JSfM2LEiKzkE9YqVUSk2FRUVNRd9r59+/ZQfS+nVBy4+4PAc8CzZhZ3OGxm5wAvAn8LLj0JUpirVBGRYhL7Pq6trQUiM+iG6Xs5nXkOLgG2A3+KbTCzG4BHgWvd/efBpCZBC3OVKiJSTOp/H8eE6Xs55eLA3WuB04Hvm9mvzOxxYBQw1N3vCDpBCUbYq1QRkWLR8Ps4Jkzfy6lOgjTJzEYC+wHlwG1AP6C/u08PPj0JStirVBGRYpHo+zgmLN/LqfYc9CAy2dHfgaej21YC3zOzE8zsK0EmJ8HIhypVRKRYLFiwoNH3cUxtbS3z58/PcUaNpTog8Tvu3gXoBZwD3BGNcTXwMrDBzFYEnqVkJB+qVBGRYrF06VLcve4WW38ndqu//k5rSWtVRndfDawG/hLbZmb7AX2BI4NITIKTD1WqiIiER6pjDu4ws2+bWaPnuft77v60u18TVHJm1sXMZpnZv6L/dm6i3XYzez16mxrU6xeKfKhSRUQkPFIdc7Ar8CSwzsz+bGZDzax9FvKKGQ285O4HAi9FHyfyhbsfHr0NzmI+IiIiBS/VMQcj3L0n8H3gA+B3wMdmNtXMLjCzPQPObwgwMXp/IpDaVIMiIiKSsnQmQcLdX3X337p7KdCHyNUL5wNVZjbPzK4ysx4B5NfN3aujr1kN7NVEuxIzW2xmC81MBYSIiEgGzN2DC2a2BzA4epvn7rcn8ZwXgUTrVP4WmOjuneq13eDujcYdmNk+7r7GzA4gctXEie6+KkG7ciLzM9CtW7eyyZMnN5vbpk2b6NChQ0t/QosUJz9yUZzcxAlTLoqTmzhhyqXY4wwcOHCJu/dtMVj9gWlhuwHvAHtH7+8NvJPEc/4MnNFSu7KyMm/J7NmzW2yTDMXJbgzFya84YcpFcXITJ0y5FHscYLEn8fub1GkFMysxs54Jth+azPMzMBUYHr0/nMiiTw1z6BxbBCrac3Es8GaW8xIRESlYLRYHZvZDYAUw3cyWm9m36u2elLXMIm4Fvmtm/wK+G32MmfU1s4eibQ4GFpvZG8Bs4FZ3V3EgIiKSpmQmQRoDlLn7R2bWF5hoZje7+xOAZTM5d18PnJhg+2Lgwuj9+cBh2cxDRESkvurqakaOHMmMGTPo3j3RsLn8lsxphbbu/hHU/SgPAC4yszFAcKMZRUREsiz2o57pmjIVFRUsW7asYKefT6Y4+NDMesceRI/mv0ukO793k88SEREJmSB+1GOL2bl7wS5el0xxMAz4sP4Gd69193OA47OSlYiISMCC+lGvv5hdoS5e12Jx4O5V7p7wHXT3/xd8SiIiIsEL4kc9VmDEFrOrra0tyN6DtGZINLOeZnanmT1pZreY2Zlm1ivo5ERE8kFQ57Ele4L6Ua9fYMQUYu9BWsUB8AyRdQ++AI4G7gNWmNlGM5sbVHIiIvmg0AenFYIgftQbFhgxhdh7kG5xUAoMdfcL3P0kd98T+CpwLjAjsOxEREKukAenFUqPSFA/6okKjJhC6z1ItzhYDMRN4BwdmzDN3W/OPC0RkfxQyIPTghrZ39oFRlA/6gsWLGhUYMTU1tYyf/78tHMMm3SLgyuBCjPr1GJLEZECVciD04Ic2d/ap1yC+lFfunRp3PoDs2fPjnu8dOnSINNuVekWB58DuwHvmNn9ZnaemR1mZm0CzE1EJNSCHpwWhqPsmCBH9rf2KZdi+lEPSrrFwRSgU/TfvYEK4A3gczN7NaDcRERCKxuD08JwlA3ZGdlfaKdcCl26xcGBRJZFvtzdh7r714CuwGlECgYRkYIW9OC0sBxlQ3ZG9hfSKZdikG5xsBDoUn+Du29w95fc/Y7M0xIRCbegB6eF5Sg7myP71XuQP9ItDh4AbjSzPYJMRkQkXwR5HjtMR9lB9IgU03wAhSrd4mAycAKRiY8eNbNLzKy/me0aXGqtK0wDg0SksIXpKDuIHpFimg+gUKVbHOxLZHzB7UAJMBL4B7DRzN4MKLdWFZaBQSJS2MJ2lB1Ej0gxzQdQqHZO50nu/gHwATA9ti3aa9CHPFnGedu2bVRXVyfct27dOh555BHcnUceeYTy8nL22muvrLxWsccJUy6Kk5s4YcolDHFGjx7d7FH26NGj+f3vf59SzHXr1nH55ZczadKkjL67IL2/6/nnn497XFlZSWlpady2dN6r1v5vVSxxIMWeAzO7w8y+bWaNnufuW9x9gbs/EEhmrWjcuHG4OwA7duxg3LhxrZyRiBSqJUuWNHuUvXjx4pRjjhs3jsrKSn13SdpSPa2wK/AksM7M/mxmQ82sfRbyajXr1q1jypQpcQODpkyZwocfftjKmUm2rVu3jiuvvDLj/9ZBxZHiMGvWLNasWVN3mzlzZtzjWbNmpRQv9h3m7vrukrSlVBy4+wh37wl8n8hphd8BH5vZVDO7wMz2zEaSuVS/1yBGvQfFIaijLR21SWtSz6cEIa0Bie7+qrv/1t1LiYwz+DtwPlBlZvPM7Coz6xFgnjnRsNcgRr0H4RbEkXpQR1s6apPWpJ5PCUqqYw4azWvg7ivd/Q53HwD0AB4BjgPOCSbF3EnUaxCjCjy8gjhSD+poK4xHbTrNUTzU8ylBSbXn4N9mdp+Z9Uq0090/dvdHolMq3x5AfjmVjYFBkl1BHKkHdbQV1qM2neYoDur5lCClWhxcCPQF3jazZ83smCzk1GqCHhgk2RfEkXpQR1thPGrTaY7ioZ5PCVKqAxKfcPejge8ABsw1swVm9iMzs6xkKNKEII7UgzraCutRWxhPc0h2qOdTgpTugMS57j4U+CbwGvAo8K/oNMoFdWmjhFcQR+pBHW2F8agtrKc5JDvU8ylBSnf6ZKBuMOKlRKZTnghUAP8OIjEAMzvTzJab2Q4z69tMu0Fm9o6ZrTSz0UG9voRXUEfqQR1thfGoLYynOUQkP6Q0fbKZ3Qt0Ajom+LdD4NlBJfAjIqtANpVTG+Be4LtAFbDIzKa6e0Gs8SCJJXOknsyUsw2PphJN85qMoOIEpaXiadSoURlPqysihSvVnoMRwA+BVUR+sH8LDCOyQuOBQDciCzEFwt3fcvd3Wmh2NLDS3Ve7ey2RFSOHBJWDhFMYj9TDJIynOUQkf1hTXyAJG5sdCVxF5Gh+OvAHd38lS7nVf905wFXu3ugb38zOAAa5+4XRx8OAb7n7LxO0LQfKAbp161Y2adKkZl+3pqaGkpLMax3FyY9cCinOiBEjWLVqVZP7v/71r3P//ffnLJ+gYyhOfsUJUy7FHufkk09e4u5NnqaPSem0gru/BvzEzPYFRgEzzeyfRIqEaanEijGzF4HuCXb91t2fSyZEolQTNXT38cB4gD59+nhL3b5BdQ0rTn7kUkhx/vGPf4Qqn7Dmoji5iROmXBQnOeku2fw+8CszuwG4CLjXzP4A3A5MinbvJxvrpHRyqKeKyIDImJ7AmgxjioiIFK1Mr1b4jEhBUAY8DvwBeC/ztFKyCDjQzPY3s7bAj4GpOc5BRESkYKS6tsL06MJKy8zs32a2EfgSWAvcBHQm8SmCtJjZD82sCugPTDezGdHt+5jZ8wDu/iXwS2AG8BbwlLsvDyoHkdaiNRFEpLWkelrhQ2AFsAH4NHrbkODfQLj7s8CzCbavAU6t9/h54PlUYq9atYrTTz89bttpp53G+eefz5YtWxg2bBibN29mt912q9t/1llncfbZZ7N+/XrKy8sbxTzvvPMYMmQIH3zwAZdffnnd9liciy66iJNPPpmVK1fym9/8ptHzR44cyYABA6isrOT6669vtP+ss86itLSURYsWceuttzbaf+ONN1JaWsrcuXP54x//2Gj/bbfdRq9evViwYEHC+HfffTc9evTgueee49FHH220f/z48XTt2pUpU6bw2GOP8fbbb/PNb36Ttm3bAjBp0iR23XVX/vznPzNtWuMhKM888wwA999/Py+++GLce1NSUsLjjz8OREbaz5s3L+65nTt35qGHHgLglltuYcmSJXH727dvz2OPPQbAmDFjWL48vj484IADGDt2LAC//vWvWb16ddz+Qw89lJtuugmAX/7yl1RXV8ftLysr45prrgHgwgsvZMOG+I/5cccdx6hRowD46U9/yvr16+M+OyeddBIjRowAaPS5g8afPYDVq1ezbt06TjnlFEaPHp3WZy9m0KBBlJaWpv3ZGz16NEcddRTLly9PuD/Zz97MmTO58847494bSO2z99RTTwHE/f+ZzmcvZtu2bUydGulsTOezt/fee3PPPfcAwXz2Vq5cGff+pPrZq6mpiXt/0vns1XfsscdSWlqa9mfvoosuYp999sn4s7do0SKuu+66Rp+dVD57DzzwQNx7A+l99mI2b97MX//617Q/e5l+7zX87L366qtx709Ln73mpDog8WeptJfCVVVVxebNm6mqquKAAw5o7XQKTm1tbV2PwYcffshnn33WyhmJSDFJ+lJGM9vf3d9Nsq0BPaMDF0OpT58+/sILLzTbJmwjSMMSZ926dfTr14+tW7dSUlLCwoUL055QJyx/U9jijB49msmTJ1NbW0vbtm0555xzkprUKVv5BBknTLkoTm7ihCmXYo+zzz77JHUpYypjDhaY2cNm1r+pBmbW2cxGAG+iiYgKlhbzyS6tiSAirS2V4uCbwCdEBgZ+ZGYvmNkEM7vfzCZH5zv4EDgXuMLd78lGwtK69MOVfVoTQURaW9LFgbt/6u6/BnoQmUb5bSLrKuxP5IqFicAR7n6su8/IRrLS+vTDlV1hXfpZRIpLyvMcuPsX7v60u1/h7j9090Hufq673+HuldlIUsJxWZt+uLJPayKISBhkNAmS5M64ceOorKxs1R8H/XBlnxaUSk4YimXJLf03zy0VB3kgdsTu7q16hK4fruybNWsWa9asqbvNnDkz7nHDpaFzKYgv56C+4MNQLEtu6b95bqk4yANhuTogzD9ckn1BfDkHESPIYllHo/khLAdIxUTFQcjp6gAJgyC+nIP6gg+yWNbRaH4IywFSMUm7ODCzjmbW18zKzKxTkEnJfxTy1QE6assfQXw5BxEjyGJZR6P5QQdIrSPl4sDMvmpm04D1wCvAq8DHZjbVzL4WdILFrNCvDtBRW34I4ss5qC/4IItlHY3mh0I+QAqzVFdl7AEsBI4AxgCnA2cA1xNZtnm+me0TdJLFqpCvDtBRW/4I4ss5iBhBFss6Gs0PhX6AFGap9hxcD7wL9HL3W4AKYLG73wwcGN3XeFktSUshXx2go7b8EMSXc1Bf8EEWyzoazQ+FfIAUdqkWB6cC17h7TfTxfsAuAO6+BbgW+H5g2RW5Qr06QEdt+SOIL+egvuCDKpZ1NJo/CvkAKexSWrIZ2BNY1cz+ldE2Ik1q7qgtk5UHJXhBfDkH9QXfsBhOdyW7ZIoVfQ7DIaj/5pK6VIuDD4FeQFUT+w+MthFJqKWjtlGjRqW9/LMEL4gv57B9wetoVKRlqRYHLwC3mNmJ7v5F/R1mVkJkDMLzQSUnhUdHbdLawlasiIRRqsXBDcASYKWZ3UNkeeZSMzsbuITIGIazAs1QCoqO2kREwi+l4sDdPzCz/sB9wO+imw1wYAZwqbuvCTZFKSQ6ahMRCb9Uew5w93eB75lZZyJjDAD+5e4bAs0sy3bZZRf23nvvZtu88847LbZJhuLkRy6Kk5s4YcpFcXITJ0y5KE5yUi4OYqLFwKuBZCEiIiKhoYWXREREJI6KAxEREYmj4kBERETihLo4MLMzzWy5me0ws77NtHvPzJaZ2etmpmvhREREMpD2gMQcqQR+BDyQRNuB7v5xlvMREREpeKEuDtz9LQAza+1UREREioY1NZVtmJjZHOAqd094ysDM3gU2EJmM6QF3H99Eu3KgHKBbt25lkydPbvZ1N23aRIcOHTLIXHHyKRfFyU2cMOWiOLmJE6Zcij3OwIEDl7h7k6fp67h7q96AF4mcPmh4G1KvzRygbzMx9on+uxfwBjCgpdctKyvzlsyePbvFNslQnOzGUJz8ihOmXBQnN3HClEuxxwEWexK/za1+WsHdTwogxprovx+a2bPA0cDcTOOKiIgUo1BfrZAMM9vNzHaP3QdOJtLzICIiImkIdXFgZj80syqgPzDdzGZEt+9jZrGlobsB88zsDSLTOU939xdaJ2MREZHcq66uZuTIkaxduzaQeKEuDtz9WXfv6e7t3L2bu58S3b7G3U+N3l/t7n2it0Pd/ebWzVpERCS3KioqWLZsGRUVFYHEC3VxICIiIs2rrq5mwoQJuDsTJkwIpPdAxYGIiEgeq6ioYMeOHQBs3749kN4DFQciIiJ5KtZrUFtbC0BtbW0gvQcqDkRERPJU/V6DmCB6D/KmODCzNma21Mz+lmDf+Wb2UXThpdfN7MLWyFFERCRXGvYaxATRe5A3xQEwEnirmf1T3P3w6O2hXCUlIiLSGhL1GsRk2nuQF8WBmfUEvg/oR19ERARYsGBBo16DmNraWubPn5927FafPjlJdwH/BezeTJvTzWwAsAIY5e7vN2xQf+Gl9u3b841vfKPZF92xYwc77ZR5/aQ4+ZGL4uQmTphyUZzcxAlTLoUW56CDDmoyzpYtWxL9zh2ZVOBkFmBozRvwA+C+6P0TgL8laNMVaBe9fzHwcktxtfBSOOKEKRfFyU2cMOWiOLmJE6Zcij0OSS68lA+nFY4FBpvZe8Bk4Dtm9lj9Bu6+3t23Rh8+CJTlNkUREZHCEfriwN2v9sgUyvsBPybSK3Bu/TZmtne9h4NpfuCiiIiINCNfxhw0YmY3EekemQpcbmaDgS+BT4DzWzM3ERGRfJZXxYG7zwHmRO+Pqbf9auDq1slKRESksORVcSAiIiLxxo8fX3d/xYoVlJeXZxwz9GMOREREJLdUHIiIiEgcFQciIiISR2MORERE8kj9MQbJ7E9nDIJ6DkRERCSOeg5ERERaQTauMgiKeg6y7LLLLuOaa64BYM6cOZgZL774YrPP+eKLL9h77735y1/+kosURURE4qjnIItWrVrFAw88wH//93+n9Lz27dvzX//1X1x99dUMHTqUXXbZJUsZiohIvmnYwxDEGIOG1HOQRXfddRd9+vRpcWnoRM4//3zef/99nn322SxkJiIi6Ro/fjzjx49nxYoVLQ4OzFfqOciSrVu38thjjzFmzJhG+zZu3Mj555/Pc889x/bt2znttNO4++676dq1a12bzp07c8opp/DQQw9x1lln5TJ1ERHJglxcZRAU9RxkycKFC/n000/59re/3WjfFVdcgZnx5JNPcvPNNzN16lTOOOOMRu0GDBjA3//+d2pqanKRsoiICKCeg6xZuHAhZkbv3r2ZP39+3L5DDz2UCRMmADBo0CC6dOnCueeey0svvcSJJ55Y1+6II46gtraW1157jWOOOSan+YuISH6I9TDMmTOHE044IZCYeVMcmFkbYDHwgbv/oMG+dsCjQBmwHjjb3d/LeZL1rFmzhq985Su0bdu20b6GpwnOPPNMzjvvPBYsWBBXHOy55551sUREpHXk0+mAoORNcQCMBN4CvpJg38+BDe7ey8x+DNwGnJ3L5BqqqamhXbt2Cfd169Yt7nHbtm3p3LkzH3zwQdz29u3bA5FLG0VEJDOtPa9ALq4yCEpejDkws57A94GHmmgyBJgYvf80cKKZWS5ya0rXrl3ZsGFDwn3r1q2Le1xbW8uGDRvo0aNH3PZPPvkEgD322CM7SYqIiCSQF8UBcBfwX8COJvb3AN4HcPcvgY1A1yba5sQ3v/lNtm3bRlVVVaN9Tz31VNzjv/zlL+zYsYP+/fvHbX/33XcB0roUUkREJF3m7q2dQ7PM7AfAqe5+iZmdAFyVYMzBcuAUd6+KPl4FHO3u6xu0KwfKAbp161Y2efLkZl9706ZNdOjQIa28165dyznnnMONN97IkUceSYcOHXj99dcZNWoUe+65J2VlZQwcOJCqqioefvhhDjzwQO666664GPfeey9z5sypmykxk3yC+ruCjhOmXBQnN3HClIvi5CZOWHJZsWJF3OODDjqoIOLEJPP+DBw4cIm7920pVj6MOTgWGGxmpwIlwFfM7DF3P7demypgX6DKzHYGOgKfNAzk7uOB8QB9+/b1lkZ1Zjryc9y4cbz77rsMGDAgLs6f/vQnpk6dyi233ML27dsZMmQId999d6PTB+Xl5QwfPrzuuUGNRA1TnDDloji5iROmXBQnN3FaK5eWBhI2/HFO9px/w+el+7eF9bsd8qA4cPergasB6vUcnNug2VRgOLAAOAN42UPQJTJixAhGjhzJ2WdHxkaecMIJxNL60Y9+1OxzX3nlFVatWhWqASoiIq2lurqakSNHMmPGDLp3797a6RS8fBlz0IiZ3WRmg6MPHwa6mtlK4FfA6NbL7D+GDRtGjx49eO6551J+7q233srw4cMz7mYSESkEFRUVLFu2jIqKitZOpSiEvuegPnefA8yJ3h9Tb3sNcGbrZNW0Nm3a8Mgjj9DS2IaGampqOOKII/jFL36RpcxERPJHdXU1EyZMwN2ZMGEC1113Xav2HmRj0qGwyaviIB/169cv5emPS0pKEq7JICJSjCoqKtixI3Kx2vbt26moqODee+9t8Xn5NK9A2OTtaQURESl8sV6D2tpaIDIvzIQJE1i7dm0rZ1bYVByIiEho1e81iIn1Hkj2qDgQEZGsGD9+POPHj2fFihUtXlaYSMNegxj1HmSfxhyIiEgoJeo1iEll7EFMMQwkDEraxYGZdQHaAx+7+9bgUhIREYEFCxY06jWIqa2tZf6hYBUCAAAgAElEQVT8+TnOqHgkXRyYWXfgfGAQcDTQrt6+94G5wJPAC2GYgEhERNLX2isYAixdujTusY74c6fF4iC6ImIF8BPgcyKzEN4OfAR8AXQB9ge+BfwN+D8zu87dH89W0iIiEj4tjSvQpYT5I5meg3eAmcBQYKa7b2+qYbSQ+CnwBzPbx93HBpOmiIiI5EoyxcGx7v56MsGiqyLeZmZ3AftlkpiIiIi0jhaLg2QLAwAzO9jd34oOUHwno8xERCRndEpA6ktmzMEj7n5BEu0OB2YA3YJITERE0tNaKxhquuLCkcwkSMPN7L7mGphZX+AlYFMgWYmISNq0gqFkKpni4GLgIjO7M9FOM+sPvAisB44PMLdY/BIze9XM3jCz5WZ2Y4I255vZR2b2evR2YdB5iIhkU6azCcY0XMFQswhKOpIZc/CgmbUD7jazGne/JrbPzI4ncvniv4ET3T0bn8KtwHfcfZOZ7QLMM7P/dfeFDdpNcfdfZuH1RUSaFIb5AOrTCoYShKQmQXL3e8yshMglilvd/UYzOxl4FlgJnOTuH2UjweiESrHTFbtEb5pkSUSkgaZWMLzuuutyOvYgRtMV56+kF15y99uB64DrzezPwHPAW8DAbBUGMWbWxsxeBz4EZrn7KwmanW5m/zSzp81s32zmIyKFITZwr1C63rWCoQTFUp3p2MxuAq4FFgKD3P2zbCTWxGt3ItJbcZm7V9bb3hXY5O5bzexi4Cx3/06C55cD5QDdunUrmzx5crOvt2nTJjp06JBx3oqTH7koTm7ihCmXcePGMW3aNAYPHswVV1yRVowVK1bEPT7ooIPSel5LWoq7fv16fvKTnyRci6Bdu3Y88cQTdOnSJenXS/fvSiRM/82LPc7AgQOXuHvflmK1WBxE101o2KgnkemTGy645O7+tZZeNBNmdj2wOdqTkWh/G+ATd+/YXJy+ffv64sWLm32toLrCFCc/clGc3MQJIkZ1dTWDBg3K6FK96upqDjjgAGpqamjfvj2rV69OKlaqgwWbOlcfVJyYSy65hIcffjhhcdC2bVsuvPDClFYwjAnLf3PFCSaOmSVVHCQz5uAlWvEcv5ntCWxz90/NrD1wEnBbgzZ7u3t19OFgIqc7RKRA1b9UL50fvFiMdAbuhZVWMJQgJXO1wvk5yKM5ewMToz0COwFPufvfoqc3Frv7VOByMxsMfAl8QmT1SBEpQA0v1UtnsF3YBu4FQSsYSpCSXrK5tbj7P4EjEmwfU+/+1cDVucxLRFpHEEf8zQ3cy1XvgS4dlDBLZvrkI939tVSCRi973M/d3047MxGRBoI44m8YIybZWPpRl2KQzKWMc81sqpkNMrNm25vZV83sGuBd4AeBZCgiEhXEpXqJYqQbS6RQJVMcfAP4mMi8BuuihUKFmV1uZheZ2WgzGx+dh2A18DPgyqauJhCR4pXJvAItHfEnG1MD90Ra1mJx4O4fRFdl3Be4HegA/Aq4C7gfuAU4EVhKpLfgIHd/ImsZi0jeymRBoKCO+JcuXYq7191mz54d97jhwL5cKS8vp7y8nIMOOkinJqTVJT0g0d0/JHIJ4W1QNyFRCbDe3bdlJz0RKRSZXmUQ1iN+TREshSjp6ZMbcvdP3X2tCgMRSUaiqwxSEdYjfpFClHZxICKSrKauMiiUNQ1ECo2KAxHJOi0IJJJfVByISFYFdZWBiOSOigMRySrNKyCSf1QciEhWhfUqAxFpWkZrK5jZqcAhQDXwrLtvCSQrESkYWhBIJP8ks7bCZcBQdz+x3radgOnAyYBFN//LzI5z94+ykqmIiIjkRDKnFX4IvNNg2wXAKcDjQB/gx8BewLWBZiciIiI5l0xx8E1gXoNtZwKfABe6+zJ3fwoYC5wacH6YWYmZvWpmb5jZcjO7MUGbdmY2xcxWmtkrZrZf0HmIiIgUi2SKgy5AVeyBmbUBjgNmu3v9UUavAj2DTQ+ArcB33L0PcDgwyMz6NWjzc2CDu/cCxhGd4llERERSl0xxsA6oPwH6kUB7YEGDdtuBxEOSM+ARm6IPd4nevEGzIcDE6P2ngRPNzBCRwGSyoqKI5Bdzb/g726CB2dNAV+C77v6lmd0PlAOHuvvb9dpdBlzi7gcHnmSkt2IJ0Au4191/02B/JTDI3auij1cB33L3jxu0K4/mTrdu3comT57c7Otu2rSJDh06ZJy/4uRHLorTvHHjxjFt2jQGDx7MFVdc0aq5KE5+xQlTLsUeZ+DAgUvcvW+LweovXJLoRqQr/wtgDbAc2AE8k6Ddq8CEluJlcgM6AbOB0gbblwM96z1eBXRtLlZZWZm3ZPbs2S22SYbiZDeG4mQ/zpo1a7ykpMQBb9++vVdXV7daLoqTf3HClEuxxwEWexK/ty2eVnD314EBwIvAauB64Kf125hZDyKnH55osRrJgLt/CswBBjXYVQXsG81lZ6AjkQGTIhKATFdUFJH8ktQMie6+yN3Pc/fT3L3C3Wsa7P8gum9W0Ama2Z5m1il6vz1wEvB2g2ZTgeHR+2cAL0crJBHJkFZUFCk+LRYH0csELzOz/s20OSbaZpdg0wNgb2C2mf0TWATMcve/mdlNZjY42uZhoKuZrQR+BYzOQh4ieSnTgYRaUVGk+CTTc1AO3ACsaKbNO0RON1wQQE5x3P2f7n6Eu/d291J3vym6fYy7T43er3H3M929l7sf7e6rg85DJFlBjeoPKk5FRQXLli1L68dcKyqKFKdkioMzgfHuvr6pBtF944GfBJWYSL7K5Mc46DixH3d3T+vHXCsqihSnZIqD3sDfk2g3N9pWpGhl+mMcdJxMBxJqRUWR4pRMcVACbE6i3WYikyOJFK2gRvUHESeIgYRLly6Nu7xp9uzZcY8brrgoIoUhmeJgLXBgEu0OjLYVKUpBjeoPKo4GEopIupIpDmYDF0eXaU4ouu9i4OWgEhPJtbCM6g8ijgYSikgmkikO/gAcBjxlZns13Bnd9lS0zdhg0xPJnTCM6g8qjgYSikgmkpkh8S3gPCLLMb9vZvPM7PHobR7w7+i+YdG2InknLKP6g4qjgYQikolkZ0j8C5ErER4G9gR+GL3tGd3W292fzlaSItkWllH9QcXRQEIRyURSxQGAu69090vc/Rvuvmv09g13v9TdV2YzSZHmZDpWIEyj+vWjLiJhkHRx0BIz28nMugQVTwpfGGYAjD1fo/pFRP5j55YamNknwEnu/lr0sQHPAVc0mKb4KGA+0CYbiUrhqf+jfu+996YVo+FYgeuuu47u3bun/PymBgCmGk9Ewmnbtm1UVVVRU1NDx44deeutzIfIhTlOSUkJPXv2ZJdd0lvyqMXiAOjUoN1OwA+IrLcgkpZMf9RjEo0VSKXQSGYAYLqFi4iER1VVFbvvvjv77bcfmzZtYvfdd8845ueffx7KOO7O+vXrqaqqYv/9908rVmCnFURSEZYZADWqX6Q41NTU0LVrVyKd34XNzOjatSs1NTVpxwh9cWBm+5rZbDN7y8yWm9nIBG1OMLONZvZ69DamNXKV5IRpBkANABQpHsVQGMRk+reGvjgAvgSudPeDgX7ApWZ2SIJ2/3D3w6O3m3KboqRCMwCKiIRbssVBDzM7wMwOAA5ouC26vWc2EnT36thgSHf/HHgL6JGN15Ls0wyAIpIvqqurOf7440N7sPHee+9RWlqaldjJFgdPA/+K3t6Obvufetv+Bfwl8OwaMLP9gCOAVxLs7m9mb5jZ/5rZodnOpdilexmiZgAUkXxRUVHBvHnzivJgw9y9+QZmw1MJ6O4TM8qo6Tw6AH8Hbnb3vzbY9xVgh7tvMrNTgT+6e6OVJM2sHCgH6NatW9nkyZObfc1NmzbRoUOHjHMPS5wVK1bEPT7ooIPSjjVu3DimTZvG4MGDueKKK5J+3i9+8QtWrmx6zqxevXrx4IMPppxPWN5jxcmvXBQnN3HCkEvHjh3p1asXEDkQadOm+avu165dS+/evampqaGkpIRly5bRrVu3uDbJxKnvtttu46mnnqJnz5506dKFI444ghNOOIErrriCL774gv333597772Xzp07889//jPh9qVLl3LppZfSvn17+vfvz6xZs3jllVcS5rNy5Uo2btwYl8PAgQOXuHvfFpOtP/gqrDdgF2AG8Ksk278H7NFcm7KyMm/J7NmzW2yTjLDEeeCBB+Ju6VqzZo2XlJQ44O3bt/fq6uq0Y4XlvVGc3MUJUy6Kk5s4YcjlzTffrLv/2Weftdh+xIgR3rZtWwe8bdu2fskllzRqk0ycmEWLFnmfPn18y5Yt/tlnn3mvXr187Nixfthhh/nzzz/v7u7XXXedjxw50t3dDzvsMJ8zZ06z26+66io/9NBDm8yn/t8cAyz2JH5HQz8gMTrp0sPAW+5+ZxNtukfbYWZHEzldsj53WRaXIC5DFBEJq6CuqKpv3rx5DBkyhPbt27P77rtz2mmnsXnzZj799FOOO+44AIYPH87cuXPZuHEjn376Kccff3yz24cNG5bhX9q00BcHwLHAMOA79S5VPNXMLjazi6NtzgAqzewN4G7gx9EKqWAENdVwEHkE/T+NiEiYZGNK9SB+ktw9Z5djhr44cPd57m7u3tv/c6ni8+7+J3f/U7TNPe5+qLv3cfd+7l5wo9HSWT9g/PjxcbdU9zeVh9YhEJFCla3LpI877jimTZtGTU0NmzZtYvr06ey222507ty5bgD1pEmTOP744+nYsSOdO3fmH//4R9z2Tp060bFjR+bNmwfA448/nsFf2rzQFwfSeKrh1jpK19wCIlLosnWZ9FFHHcXgwYPp06cPP/rRj+jbty8dO3Zk4sSJXHvttfTu3ZvXX3+dMWMic/hNnDiRX//61422T5gwgUsvvZT+/fvTvn379P7IJCSztoK0skzXD8hGHg1pHQIRKQTZvEz6qquu4oYbbmDLli0MGDCAK6+8ksMPP5yXX3650doKhx9+OAsXLmwUo6ysjDfeeKPu8Q033JB2Ps1Rz0HIhekcv+YWEJFC13BK9Ya3TKZULy8v5/DDD+fII4/k9NNP58gjjwww82Cp5yDkmjvH39JRenl5edzjhuMKGu5vScP/KebMmcMJJ5yQUgwRkWL1xBNPtHYKSVPPQYjpHL+IiLQGFQdZFLsKYMWKFUlfDVCf1g8QEZHWoOIgxHSOX0REWoPGHIRY0Of4Y2MMNFZARKR5mY7RynfqORAREcmBTJZYzubyzImo5yBALY0rKPZKVERE8oOKAxEREZo/wKu/L5MDuy+//JLhw4ezdOlSDjroIB599FFuvfVWZs6cyRdffMExxxzDAw88gJmxZMkSLrjgAnbddde6xZlyRacVmhCWhY5ERKRwvPPOO5SXl/PPf/6Tr3zlK9x3332Ul5ezaNEiKisr+eKLL/jb3/4GwM9+9jPuvvtuFixYkPM8VRw0IZ2FjkRERJqz7777cuyxxwJw7rnnMm/ePP7xj3/wrW99i8MOO4yXX36Z5cuX53R55kR0WiGBhgsdXXfddXTv3r3F5wU9I6GIiBSWhksumxm/+tWvWLJkCfvuuy833HADNTU1OV2eOZHQ9xyY2b5mNtvM3jKz5WY2MkEbM7O7zWylmf3TzDKasDrRQkciIlLYysvL626p7EvFv//977rTBE8++WTdWII99tiDTZs28fTTTwPkdHnmREJfHABfAle6+8FAP+BSMzukQZvvAQdGb+XA/em+WJgWOhIRkcJy8MEHM3HiRHr37s0nn3zCiBEjGD58OIcddhhDhw7lqKOOqmubq+WZEwn9aQV3rwaqo/c/N7O3gB7Am/WaDQEedXcHFppZJzPbO/rclGSy0JGIiEhT9ttvP958881G28eMGcPYsWMbbc/V8syJWOT3ND+Y2X7AXKDU3T+rt/1vwK3uPi/6+CXgN+6+uMHzy4n0LNCtW7eyyZMnx8Vfv349P/nJTxJOWdyuXTueeOIJunTpknLemzZtokOHDik/rxjihCkXxclNnDDloji5iROGXDp27EivXr2AyAFfmzZtMs4n7HFWrlzJxo0b49oMHDhwibv3bTFYc+tWh+kGdACWAD9KsG86cFy9xy8BZc3FKysr84ZGjBjhbdu2daDRrW3btn7JJZc0ek4yZs+endbziiFOmHJRnNzECVMuipObOGHI5c0336y7/9lnnwWQTfjj1P+bY4DFnsRvbj6MOcDMdgGeAR53978maFIF7FvvcU9gTaqvo4WOREQKl+dRT3mmMv1bQ18cWORajoeBt9z9ziaaTQXOi1610A/Y6GmMN1i6dGlc5TR79uy4xw0XQhIRkfxQUlLC+vXri6JAcHfWr19PSUlJ2jFCPyAROBYYBiwzs9ej264Bvgrg7n8CngdOBVYCW4CftUKeIiISUj179qSqqoqPPvqImpqajH44Y8Icp6SkhJ49e6YdK/TFgUcGGTY7E0T0PMqluclIRETyzS677ML+++8PRJatP+KIIzKOWahxIA9OK4iIiEhuqTgQERGROCoOREREJE7oxxxky3vvvUffvs3PA7F582Z22223jF9LcfIjF8XJTZww5aI4uYkTplwUh+TWHkpmMoRCvCWaBKmhMEzcUehxwpSL4uQmTphyUZzcxAlTLsUeh0KaBElERERyJy+KAzMbZGbvRJdkHp1gfzszmxLd/0p0DQYRERFJQ+iLAzNrA9xLZFnmQ4BzEizZ/HNgg7v3AsYBt+U2SxERkcIR+uIAOBpY6e6r3b0WmExkieb6hgATo/efBk6MTrssIiIiKcqH4qAH8H69x1XRbQnbuPuXwEaga06yExERKTDmIV+EwszOBE5x9wujj4cBR7v7ZfXaLI+2qYo+XhVts75BrHKgHKBbt25lkydPbva1w7AGeaHHCVMuipObOGHKRXFyEydMuRR7nIEDBy5x9+av44fwX8oI9Adm1Ht8NXB1gzYzgP7R+zsDHxMtfJq66VLGcMQJUy6Kk5s4YcpFcXITJ0y5FHscCuhSxkXAgWa2v5m1BX5MZInm+qYCw6P3zwBejr4JIiIikqLQz5Do7l+a2S+J9A60AR5x9+VmdhORCmgq8DAwycxWAp8QKSBEREQkDaEvDgDc/Xng+QbbxtS7XwOcmeu8REQkv1RXVzNy5EhmzJhB9+7dWzud0MqH0woiIiKBqKioYNmyZVRUVLR2KqGm4iDLLrvsMq655pqk248bN47evXuzY8eOLGYlIlJ8qqurmTBhAu7OhAkTWLt2bWunFFoqDrJo1apVPPDAAwwfPrzlxlEXX3wxH374IRMnTmy5sYiIJK2ioqLuwGv79u3qPWiGioMsuuuuu+jTpw/f+MY3kn5O+/btOe+887j99tuzmJmISHGJ9RrU1tYCUFtbq96DZuTFgMRs2LZtG9XV1Rm3acrWrVuZNGkSo0aNiouzfv16br/9dmbOnMn69evp2rUrxxxzDLfffjvt2rUD4MQTT2Ts2LFMnTqVo446KpB8gvq7go4TplwUJzdxwpSL4uQmThhyGT16dKPTtdu3b2f06NH8/ve/z3k+YY4D6jnImtdee42NGzfyrW99q27bp59+yuDBg5k6dSoXXXQRkyZN4tprr2Xbtm1s27atrl1paSm77747s2fPbo3URUQKyrp165gyZUpdr0FMbW0tU6ZM4cMPP2ylzMKraHsOsm3JkiWYGQcffDArVqwAYPz48fzf//0f//u//8thhx1W1/aHP/xh3HN32mknDj74YF577bWc5iwiUojGjRtHU/Pi7dixg3HjxqXde1Co1HOQJevWrWP33Xenbdu2ddvmzp3L4YcfHlcYNKVr166sW7cumymKiBSFJUuWNOo1iKmtrWXx4sU5zij81HOQJVu3bo0rDAA2bNjAIYccktTzS0pKqKmpyUZqIiJFZdasWXGPKysrKS0tbaVs8oN6DrKkc+fObNy4MW5bly5dkh4Z++mnn9KlS5dspCYiItIsFQdZ0qtXL7Zt28aaNWvqtg0YMIClS5eyfPnyFp//73//m69//evZTFFERCShtIsDM2sXXSnxEDPbM8ikCkG/fv0AeP311+u2lZeX87WvfY2zzz6bhx56iHnz5jF16lQuvfRSNm3aVNdu48aNrF69Ou5KBxERkVxJqTgws93NbISZzQU2AiuBSmCtmb1vZg+a2VHNRykO++67L0cccQQzZ86s29axY0eee+45vve973HPPffw05/+lBtvvJGdd96ZXXbZpa7dSy+9RNu2bfne977XGqmLiEiRS3pAopmNAq4FVgNTgZuBNcAXQBegFPg2MMvMFgKXufu/As84j5x33nmMGTOGc889t27bHnvswdixY5t93jPPPMMPfvADjTkQEZFWkcrVCscAx7t7ZRP7XwUeMbOLgZ8DxwNFXRycccYZ3HfffUybNo2+ffsm9ZzKykrmz5/Pyy+/nOXsREREEkv6tIK7nwmclkS7re5+n7s/lFFmBaBNmzbceeedddMiJ+Ojjz7izjvvZP/9989iZiIiIk1LdZ6Da81skbu/2FQDM9vf3d/NMK+CUVZWllJxMHDgwCxmIyIi0rJUr1a4AnjSzL6aaKeZHQsszDgrERGRIrBu3TquvPLK0K3vkFJx4O4PAs8Bz5pZ3OGwmZ0DvAj8Lbj0RERECte4ceOorKxk3LhxrZ1KnHTmObgE2A78KbbBzG4AHgWudfefB5OaiIhI4YqtFunuoVsdMuXiwN1rgdOB75vZr8zscWAUMNTd7wg6QRERkUJUf7XI2OqQYZHqJEiTzGwksB9QDtwG9AP6u/v04NMTEREpPLFeg9hqkbW1taHqPUi156AHcD3wd+Dp6LaVwPfM7AQz+0qQyYmIiBSi+r0GMWHqPUh1QOJ33L0L0As4B7gjGuNq4GVgg5mtCCo5M+tiZrPM7F/Rfzs30W67mb0evU0N6vVFRESC1rDXICZMvQdpLbzk7qvd/S/uPtrdv+vuewAHAGfznx6FIIwGXnL3A4GXoo8T+cLdD4/eBgf4+iIiIoFK1GsQE5beg1THHNxhZt82s0bPc/f33P1pd78muPQYAkyM3p8IDA0wtoiISM4tWbKkUa9BTG1tLYsXL85xRo1ZU9VLwsZm9xOZQrkdMB34H2CGu3+RleTMPnX3TvUeb3D3RqcWzOxL4HXgS+BWd/+fJuKVExlISbdu3comTZrU7OvX1NRQUlKSwV+gOPmUi+LkJk6YclGc3MQJUy7FHufkk09e4u4tLvaTUnFQ9ySzo4kc1Q8B9ifS5f8/wDR3/yjFWC8C3RPs+i0wMcniYB93X2NmBxAZ+3Ciu69q7nX79OnjL7zwQrO5VVZWUlpamsyfoThpWLduHeeddx6TJk1ir732atVcFCd3ccKUi+LkJk6YcgkqTr5+f+2zzz5JFQfpjjl41d1/6+6lQB8iVy+cD1SZ2Twzu8rMeiQZ6yR3L01wew5YZ2Z7A0T/TThKw93XRP9dDcwBjkjn75LcCuvMYCIiLSn076+0ioP63H2lu9/h7gOIXOr4CHAckasZMjUVGB69P5zI1M1xzKxzbCpnM9sDOBZ4M4DXliwK88xgIiLNKYbvr4yLg/rc/WN3f8Tdh7r77QGEvBX4rpn9C/hu9DFm1tfMYktCHwwsNrM3gNlExhyoOAi5MM8MJiLSnGL4/kqqODCzEjPrmWD7ocGn9B/uvt7dT3T3A6P/fhLdvtjdL4zen+/uh7l7n+i/Dwfx2mFdKasQhH1mMJF8p++v7CmW768WiwMz+yGwAphuZsvN7Fv1djc/3D+PFfr5pNYU9pnBRPKdvr+yp1i+v5LpORgDlLl7HyLn/R8xs59E91nWMmtFxXA+qbXkw8xgIvmsUL+/wtAbUkzfX8kUB21jlye6+2JgAHCRmY0BUr8OMg8Uw/mk1pIPM4OJ5LNC/f4KQ29IMX1/JVMcfGhmvWMP3H09kcGBBwO9m3xWniqW80mtJR9mBhNpLZkeHRfq91dYekOK6fsrmeJgGA3mF3D3Wnc/Bzg+K1m1omI5n9RaZs2axZo1a+puM2fOjHs8a9as1k4xY2Ho/pT8lOnRcaF+f4WlN6QYvr9iWiwO3L3K3dc2se//BZ9S6wnz+ST94OSPMHR/Sm4F8f9npkfHYf7+ykSh9oaEXVrzHJhZTzO708yeNLNbzOxMM+sVdHK5FubzSfrByQ9h6f6U3Ari/89Mj47D/P2ViULtDQm7dCdBeobIugpfAEcD9wErzGyjmc0NKrlcC+v5JP3g5I+wdH9K7gTx/2cQR8dh/v5Kt1elUHtD8kG6xUEpMNTdL4iujbAn8FXgXGBGYNnlWFjPJ+kHJz+o+7M4BfH/ZxBHx2H+/kq3V6VQe0PyQbrFwWKgQ/0N0bEJ09z95szTkhj94OQPdX8WnyD+/wzz0XFQV0+k26sS1t6QYpBucXAlUGFmnVpsKRnRD05+CPMXvGRPEP9/hvnoOMirJ9L5W8LaG1IM0i0OPgd2A94xs/vN7DwzO8zM2gSYW9HTD07+CPMXvDQtDOfDw3p0HPTVE/reyi/pFgdTgE7Rf/cGKoA3gM/N7NWAcit6+sHJH2H9gpfmheF8eFiPjrNx9YS+t/KHNfXhbvZJZpuBo919eb1tnYEjgcPd/Y7gUsyODh06eO/e8RM8nnbaaZx//vls2bKFYcOGsXnzZnbbbbe6/WeddRZnn30269evp7y8vFHM8847jyFDhvDBBx9w+eWX122Pxbnooos4+eSTWblyJb/5zW8aPX/kyJEMGDCAyspKrr/+et544w22bNnS5N+w66670qdPn7rHN954I6WlpcydO5c//vGPjdrfdttt9OrViwcffJAXXnih0f67776bHj168Nxzz/Hoo4822j9+/Hi6du3KlClTeOqppxq9P5MmTWLXXXflz3/+M9OmTWv0/GeeeQaA+++/nxdffDHuvSkpKeHxxx8HIl8q8+bNi3tu586deeihyCrdt9xyCxByCEAAACAASURBVEuWLInb3759ex577DEAxowZw/Lly+P2H3DAAYwdOxaAX//616xevTpu/6GHHspNN91EZWUlf/rTn6iuro7bX1ZWxjXXXAPAhRdeyIYNG+L2H3fccYwaNQqAn/70p6xfvz7uvTnppJMYMWIEAKeffnqj96bhZy8m9v6k+9mLGTRoEL/4xS+S/uw1NHr0aI466qi6//YNJfvZmzlzJnfeeWfcewOpf/bqvzeQ3mevtraW1157DXenpKSEhQsX8vjjjyf92Uv2/89kP3sAv/zlL1m5cmXc+5PqZ6+mpibu/Unns1f/vYlp164dpaWltG3bNu75iT57iZ4fY2YceeSRdXGS/ewtWrSI6667rtFnJ5XP3gMPPBD33kB6n72YzZs389e//jXlz15M7HuvsrKSWbNmpfy9t/fee3PPPfcAke+9V199Ne79SfTZW7BgwRJ379so0QZ2bqlBExYCXepvcPcNwEvRmwSg/g8//Kc4WbRoEbfeemsrZSVSGKqqqurux45o99prr6SfH/v/M/YFXVlZyVNPPdWoMM1H9d+bmB07dlBVVcUBBxyQ1vMb7k8mjrSedHsOzgIuBs5y948DzyoH+vTp44mOnuurrKyktLQ049dSnPzIRXFyEycMuaxbt45+/fqxdevWum2x3oNUCoSg8glTnETvTUyy79F3v/vdZoukQw89NK3TJa393hRCnH322SepnoN0xxxMBk4gMvHRo2Z2iZn1N7Nd04wnIpIzOh/etCDGUoR1HIUkL93iYF/gNOB2oAQYCfwD2GhmbwaUm4hI4HQVUPM0uFYgzTEH7v4B8AEwPbYt2mvQhwJcxlnCLXY52qRJk9LuEi5ken/iJXNk/Pvf/z7HWYVHw6P6oLq8Jb+k1HNgZneY2bfNrNHz3H2Luy9w9weCS0+kZVqUqnl6f+LpyFikZameVtgVeBJYZ2Z/NrOhZtY+C3mJJEWLUjVP709jOh8u0rKUigN3H+HuPYHvEzmt8DvgYzObamYXmNme2UhSpClalKp5en9EJB1pDUh091fd/bfuXkpknMHfgfOBKjObZ2ZXmVmPAPMUaUTTszZP74+IpCvVMQd7NNzm7ivd/Q53HwD0AB4BjgPOyTQ5MzvTzJab2Q4za/K6TDMbZGbvmNlKMxud6etKftDlaM3T+yMi6Uq15+DfZnafmfVKtNPdP3b3R9x9qLvfHkB+lcCPgLlNNYgu9nQv8D3gEOAcMzskgNcORKZLnkpiuhyteXp/RCQTqRYHFwJ9gbfN7FkzOyYLOdVx97fc/Z0Wmh0NrHT31e5eS2SCpiHZzCsVGimeHVqUqnl6f0QkE+lOnzwA+BXwA2ARMBZ41tMJltzrzQGucvdG1xiZ2RnAIHe/MPp4GPAtd/9lgrblQDlAt27dyiZNmtTs69bU1FBSUpJ23uvXr2f48OHU1tbStm1bHn30Ubp06dLyE7OUTxjjpBtjxIgRrFq1qsn9X//617n//vtzlk/Y4oT5/Wnt90Zxch8nTLkUe5yTTz45ewsvuftcYG709MIo4FFgrZndCUxw9y+SjWVmLwLd/397Zx6/21Tv8ffXdIxlyFQUEkrpGG+kLhoUEbllSKWIkKKreUBH5RYXIV23DA1IhCKzQ4bMB8ecUIRIV6JBnM/947ue89vPPs/z/Nbae/1+v+d3rPfr9bx++9nD97f23s9e+7u+6zv02PRFSWfHiOjVxF47SjoOOA68tsJoiT3aJv/43Oe63R/OO++8VslVJmMe77GSccUVV2Rvy9wkZ5ivzzC1pcgZHznD1JYiJ46m6ZOB2c6Ie+PplE8CpgG/T5TxVkmv7fGJUQwAHgr/v8MKwMMpbRgLiqd4oVAoFCYrSZYDMzsGWBx4cY+/i2ZvXRzXA68ys5Xx3As7ADtNUFtmM8hT/IWcmrVQKBQKw0+q5WBPYFvgt8D/AF8EPoBXaHwVsCxeiCkLZratmT0EbAica2YXhPUvNbNfAkh6Dvg4cAFwJ3CapAktqF48xQuF3pTonUJhcpCqHKwHnAXsAWwD/FHSZZJukvRbSY+HiIEsSDpT0gqSpkhaVtLmYf3Dkrao7PdLSatJeqWkr+X6/00pnuKFQm9K9E6hMDlITZ98k6SdcCvB74ALzewKM9tqTFo3SSmFXQqFOSl1HgqFyUPT9MkPSvoU7gj4c+AYM7vTzHY1swWytnASUgq7FApzUuo8FAqTh7bRCk8BhwLrAj8Gvgk80L5ZhcL4M2zz4cPWnjaU6J1CYXKRWlvh3FBYaaaZ/d7M/gI8BzwKfBVYgt45CwqFoWfY5sOHrT1tKHUeCoXJRarl4DE8dPCnuMVgH9wxcRNgLeDlwGIZ21cojAvDNh8+bO1pQ4neKRQmH6kOiR+WtJ+kr0r6tqQfSPqFpCsk3SbpIUnPjFVjC4WxYtjmw4etPW0o0TuFwuQjWjkISYZi9zUzW3H0PQuFiWfY5sOHrT1tKdE7hcLkI8Vy8Gsz+76ZbdhvBzNbwsz2BO5giCojFgqDGLb58GFrT4emDpIleqdQmHykKAdrAH/GMxU+bmbnm9kJZnasmZ1qZrfiPgk7A/tKOnosGlwo5GTY5sOHrT1V5iYHyUKhMJho5UDSk5I+DbwMT6N8F15XYWU8YuEkYG1Jb5R0wVg0tlDIzbDNhw9bezrMTQ6ShUJhdJJLNodyzKeHT6EwqRm2+fBha0+HXg6SpYBYoTD3kqwcFApzE/X57pz10JswbO2B/g6S++23H8sss8yEtq1QKIwNrTIkFgqFuZ9hdZAsFApjR1EOCoVCX4bZQbJQKIwdRTkoFAp9GVYHyUKh0E3uWixFOSgUCn0ZVgfJQqHQTe5Q4+KQWCgU+jKMDpKFQqGbeqhxDmfh1pYDM1vMzDY0s93N7NtmNr2tzMLwMzeVEy4UCoXJzFjUYkkt2byWmb3fzL5hZueY2QPAk8BVwDeBdYA7W7eqMPSUbHmFQqEw8YxVLZaUwks/A2YA3wN2A7YAbsZLNq8saXFJG0vaq1WLCmNKjhF/yZZXKBQKw8FYhRqnWA62BHYHFgVWBP4L2Bx4C/CnVq0ojBs5RvxzUznhQqFQmKyMZahxinLwdeAnkp6X9A9JnwfWD587zGzLxq0ojAs5RvxzWznhQqFQmKyMZahxSuGlgyQ9XVt3m6Q3AocAPwrVGbPlUzWz95rZ7WY2y8zWG7DfA2Y208xuNrMSW9WHHCP+ki2vUCgUhoOxDDXOkudA0rHAa4B58WqNubgNeA/wq4h9N5U0VVJfJeKFTI4Rf8mWVygUCsPDRRddxMMPPzz7c+GFF3Z9r4cip5Atz4GkR4D3mtkWGWXeCWBmuUTOZv7552f55ZcfuM/dd9896j4xDIOcadOm9RzxH3fccRxzzDGNZTSV1WEYrk2RM75ycsh45JFH2H///bngggtYbrnlJrw9Rc7kaUuRE0f2DImSfplbZsy/BS40sxvNbPcJ+P9DzSOPPMIJJ5zQc8R/wgkn8Oijj0bJ+fWvfz3QhHX11Ve3bmuhEMO0adOYOXMm06ZNm+imFApzJdZvJDhuDTC7GOil+n9R0tlhn8uA/SX1nEAxs5dKejj4O1wE7CNpjqmIoDjsDrDsssuue+qppw5s29NPP82iiy6acjpDKefwww/nl7/8Jc8999wc2+abbz623HJL9t1333FrT24ZRc7kktNWxhNPPMFOO+3Es88+y5QpUzj55JNZcsklJ6w9Rc7kassLXc6mm256Y9T0u6Sh/wCXAetF7nsgrkgM3G/dddfVaEyfPn3UfWKYaDlTp04Vbl3p+Zk6deq4tie3jCJncslpK2PPPffUAgssIEALLLCA9tprrwltT5EzPjKKnDxygBsU8S6d9IWXzGwRM1usswy8HXdkLARmzJjRddOnT5/e9X3GjBkT3cRCIYr6FFnq1FihUIhjqJUDM9vWzB4CNgTONbMLwvqXmlnHt2FZ4EozuwW4DjhX0vkT0+JCoTCWTJs2jVmzZnWte/7554vvQaGQmaFWDiSdKWkFSVMkLStp87D+YUlbhOX7JL0+fNaU9LWJbXWhUBgLcjnWFgqF0Rlq5aBQKBQ69LIadCjWg0IhL0U5KBQKk4ISSlsojB9FOSgUCpOC4lhbKIwfRTkoFAqFQqHQRVEOCoVCoVAodFGUg0KhUCgUCl0U5aBQKBQKhUIXRTkoFAqFQqHQRVEOCoVCoVAodFGUg0KhUCgUCl0U5aBQKBQKhUIXRTkoFAqFQqHQRVEOCoVCoVAodFGUg0KhUCgUCl2YpIluw4RgZo8Dvxtlt5cAf8rw74qcydGWImd85AxTW4qc8ZEzTG15oct5haSlRxP0glUOYjCzGyStV+SMnZxhakuRMz5yhqktRc74yBmmthQ5cZRphUKhUCgUCl0U5aBQKBQKhUIXRTkYzHFFzpjLGaa2FDnjI2eY2lLkjI+cYWpLkRNB8TkoFAqFQqHQRbEcFAqFQqFQ6KIoB4VCoVAoFLooykGhUCgUCoUuinIwADNbJ5Ocec3s/TlkFQoThZltPdFtGCbMbBsz29/MNp/othQKuZlvohswLPRQBAw428y2wh03b4qQ8SJgb+BlwM+Bi4CPA/sDNwM/jmzLxsAqkn4Qvp8OLBk2Hyzp0kg5RwF9PU4lfSJSzrcHbU+QY8B7Q5tOBzYD3g3cBXxX0qwYOQPkLw7sLelrDY/fGNgAuE3ShW3aUpH5YUknRO77qUHbJf13pJyfjyJn1Je8mb2nvgo4xszmCzJ+FtOWHnIbX+PwfC0t6be19WtJujVBzhr4M3qtpKcr698h6fxIGd8B1gSuBqaZ2QaSpsW2oSZrXmAJSX8K3xcAdgH2k/TqiOOXAw4AZgFfAfYBtgPuBD4p6ZEm7ar9j7cBn5H0toh9VwS+hV/j84BvSfpX2HaWpG0iZLwY+DywDdDJ5vcYcDZwiKQnE9u/HICkR81saeBNwN2Sbk+R00d2q36nImd1YH9JH008bv7O9a2se0nn99SUYjkY4QbgaOCw8DkUWAr477Acww+B1YGZwG7AhcB/AO+W9O6EthwU2tNhdeDTwIHAZxLk3ADcGD5bV5Y7n1g+BmwMPFyTmSrnGOB9wAfwa/WxIO/NwOGxQsxsRTM7zszOMbPdzGxhMzsMuAdYJkHOdZXlj+L3fzHgADP7XKycUTgoYd9DgZ3x392ioS3VTywbAisAVwSZh9U+MZwGfAR4F7BV+LtIZTmKXNfYzN6HK5FnmNntZrZ+ZfOJCXI+gb9g9gFuM7Pqc/n1WDn4b3YzSZ8HNsFfYsmY2Q7An4FbzexyM9sUuA94JxBrbTwRuAN4EJgO/B3YEr//301sz2Zmdo+ZPW1mPzKz15jZDcAhwLGRYo4HLsOv8fLA5Wa2VNj2ikgZpwH/B2wiaSlJSwGbhnU/jZQBgJntAfwauMbM9gTOwX/DPzOzXRPk5Op31jKzC83sNjM72MyWNbMzgEvw+xgrZ1Mzewh4OMhbqbK5/eBGUvl4OOd/AJcDW1TW3Z8oY2ZleV78h7xYg7ZcX/v+s8ryVQ3Pb0aLa7MU/iKfjltDdsNHOqlyZoa/8wNPAAuE7/NVr12EnOm4orQ5rlTcCpwCLNf0mgDX46NS8JdgSntu7fOZCfwzQc5UvBO+Gfg+8FZCuHHiec0LvAM4CZgBHAysmShjfbyz2rPThtTnIfM1vhlYPixvgCsK70n9bYd7smhYXglXTj/ZQM5Ng74nyLkNWDUsrwP8E9i2xTX+ff26pcrClZ0puMLzVOf6pNyr2vedgduBV8ZeJ3xUn7xtwD1fOPRjT3f6CWCJlOuTsd+5FrcMrQ58EvgDbmlZMFHO9Z3nGn9//QZ4Q/030fTT6uC57YOP1g7HNdOXA/clHp+rw/jNgG33NpTZqC095LwMnyZ5GPhA4rHVTuz82raUh/SW2vc/AlManMstoYNYCrihX1sj5PwRf7G/ovZZCXi44XXeCDgKNw1v3eJ+TQkd0ePAPonHzhM6r+n4Cznpech8jWfWvi+PW60+kfLbBu6ofV8UOB+3EKb8Bv9GtxL4t8ryrQly6n3GXU2ucWX54Nq26Lb0ac9vG7Tn9vqLDld07wUeiZRxIW4lXbayblngs8DFTc+pR9+R8hvM1e/UlacHgXnb3PfwfU3gbmDblGei36f4HFSQzz/uZ2Zr46OuFFMuwOvN7KmwbMBC4bu5eL0oUs5dZralpHOrK83sXfjNnxCCX8aOwNvwucSUKQWAR81sUUlPS3pHRe5ywLOJbVkCv64AjwILm9kiAJL+HCnmxfg5GCAzW04+J7loRXYM5+Cj0Zt7tPOyBDmdY5YG1gZeBzyEz7WmypiCm5Z3xJWUbwNJfgJyH5Ajg89L9LRPjVzX+K9m9koFfwNJj5jZJsBZeKcYy6NmNrVzryQ9HZ6r4/HrHcuovgCRLFPzNVm0+l1xfiZnV56rL3VWmtmquMk7hcVr/iZW/a44X5PvAf+GW2I7x11sZu8FvhnZju2Bz+FTEh2T/R9xX673RcroMKsyL79lZ6WZLUji1HqmfmfB8I7pyHkaWCv4ZKEI/7bAvzrPUzjudjN7C94fvTJSRl9KhsQ+hBu1mKSnRt05//9eFTgXd3bq/FDWxUeT75IU9cCb2V8ZcUhcGB/dQKKyYmYH4XN0dwKn4qP+52KOjZS/CLCIpKiXoJk9gDtf9Xq5SNIqLduzEG4qvL+NnAb/98N4p7gg7rB5Wuw1qck5CXgtrsCdKum2rA3NgJktjI8Ko66xmb0eeEbSvbX18wPvkxTr7LsC8FynQ61te6Okq2Lk9JH9LknnJB5zwKDtklJ8VlpjZoOcZyXpI+PWmEyY2ctxi0Xdae9lwKslXRwp5wEy9DthwNDvxStJm0XKeSvwuKRbautfDHxcbR0ki3LgWCaP/B5yd5eUnO86jPzez8io6HbgZEn/aNKONpjZLNxJ6u9hVedH01Ey1oqUMzA0NEFjHvQ/XibpD5H7rg+8RNJ5tfVbA3+QlGoZqcpIvu/hOs8Efh9WdT2ciogyqMh5poeMaKXQMkQ8BDlLDtqeMNoaM5o+oz3k3CQpS/hzW8zsHEnRjqORMreTdEbEfmPVl2Y/p1yk9Dtj3I4lgBWVEL3TV1ZRDhwzexZ3DjoNn0/v0g4lndRQ7tB0GND4pTXQw1jS7yLlzMKVnMc7q7rFxGnMo/yP30t6eeS+lwG7SHqgtn5V4Lg27Wly383s3wdtl3T5oO05MbPH8bnQU3AHqvrzENWWcM8fAjqWpvo9T7Ly1KxhHf6COxb+p6T7UuQFmVmeUTObIWntxGPG6kWa3JYImVHP1hj2pY3OqcdvxsL31OneQf8jpd+phwl3ETl1U5V3GR6NNh/uuPs4cLmkgaHRo1F8DkZYHo/B3x7vyH4CnCHp/1rKTZlX9QO644R/CRyqxDjhAXyMxMpdsS//CP4Tj7/+Oz49caYqceaZSLneS9UVAwBJ99pI6NV4tKPzf+d44ZrZOpksKqlK4XK4b8mOwE74NNcpSo8LPwr3fr8KVzSuVLsRyX/jL5yT8Wu8Q2jr3bjfwCYNZCbfqz7s0eCYjzHgRdqCGZnkVIlt21j1pU3P6RL8N/IzfJrt96Ps34SU+3Y6/hLv+Ch1Kcwk+gYBL5b0lJntBpwg6QAza205aOXNOLd+aOGR30PWCg2OuQjvNKbinevV+IsMWoaoNDkeWBF/mV8BfAGYv7LtrAbyVg5yrsU7xakZ793vE/btG/kxaNtY3fc+cnJFmTSWQ4uIh3C84THqx+Ed4jeBlRu25doe664Jf29pKLPJM/qeQZ8EOVnChMfjk/JsVY7J1pe2bPuLgQ8DF+COknsBS07EtcGjCU7FrV1fJoSytvjfM3GF7EJg/bAuKUql16dYDmq09ci3WvY1SQ+F9dHZ1/BY8E7ykn3MbGfgV2EuvO080FYNjjkeOAO4BtgV9yDeStITxCc1mY2k+83sbGAhPCHSaoxo0aNi/TM/GrB4QlMuNrOvAV9SeKKC/IOAqCyU/ajc9+gMiX3INZJsYsFqHfEAbrcFppvZDHykPw2Pyf7fVFm45/n78NEXeHz37H+VKsxCxkYzS83YmGX0F56h7wLfDQ5yOwK3m9lnJf0woT09MbPzJL0zYf+Z9H+2lk38322jm/rJTTonAEl/AU4Izrrb44OuBXFLVOz/zdLvSDoTODM4Yr8bOCxYKr+oZlOHX8WVnislXW9mq+DPVyuKz0Egh0e+efa1vYOMqXjykLPDtuh5TTO7HVhXFefD4Jn6Xdyrf/mENuVIFXuzpKmV7zvjqU23Bn6acF6r4C+Hd+Pz2acC5yjRydLMPjRouyLnNMPD+T08hr/TyU/Fk4vspgxTHilzkX2O30bSWRnasUJHYYncP0vEQ6UD3B5Pg/sz4CeSHmwobxXgSDwLJHjmu/3wRDLrSrpylOOvk7RBWP4o/ryeCbwd+IWkQyLbsS1+TqviGRdPUS2SIoXai/RG4DBJUdnyBjj6Gv58pfQXrf2LMvWl2c4pyNsIv75vAq7Ef4NXJMrI0u9U5HWSle2AP2ufk3RBioyxpCgHgRwe+UHr3lAeO70SPrr4oaQjU5xpzGw/3Ax8eW392sA3FZHfPOw/VMpKuMa34p3pU8zpjR+txfeQPTvet8Gxq1CJClGiU9uA+T0DVpM0pUm7av9jDUl3ZZATZcnIEfEQ5DyDj2JOwZPg1O95oxoNTak+h2Z2PZ4R9fGgxFwjKSXXQV35SR79ZXqRPo+byntZh94gaaEUeT3kJ4VoZupLs52TeQjik/j1vZQR51hvXAufnib9jnmK7B3xQcnFuPJ9w+CjomVnc4AvykEgk8Z8h6TXVL4viisId+B52Kf2PXgMGEJl5UAGF4JqHNPd9KEwL3LTCRkVfq9OlvTPBBl/xFOq1h2uDLha0ktT29Xjf7SyQOSWk/D/TmRwTHfjuPmGESG34E6L8wAXSFqvsq1JtEGr0V+mF+lteMrlOUzJZvagpBVj29NHftJ1ztSXZjsn684r0IlSqDRn3KOSOoOkK0N76gpzowiVIDtblErxORhhfjwpS1cSFDN7E+5ME0OW7GuWL7xp3o5pXNID5hnlTg8Pb/QctKSe2fEkzcDNoLFyDozdtwFN5tRfg2dcu4qRLH6bAF80s61jzbpkypA44L4nzWmOYsmImjc2T1L0L41EyawObAE8EOZMo5C0S+y+DWjij5ElY2OP0d+RDUd/Kzc4ps6B9M/0t08G+UnXuf7yD/Ppb8ad9mL9Dg4k0zlJ2iRl/0Sa/AY/nL0VI5w7+i5xFMtBwMzOAb6gWvIIM1sPOEDSqI58lin7mmWKEzazS4FPVV9a5iV3jwfeL2neSDm5SjZ/E8/P/93a+v3wjISfjZHTR/Zekr6TeMwlePnXi2rr34qbhzdt2p4mmMdj/ydefKfOYZJeEimntSXDzH4F7CrpN+Z5H67DS46/Bi8MFlVR0TwV8F8kfb+2fh9ceT0iRk4f2Qerki64DZaesXEsR3/JmRbHEvNy1NeNvufs/c/BLSi3mdnyeJbXG/CUvse1uedNCD5SppqDZ/A5eUbSyS1kJ/c7feQ0nhatyGiihPWXV5QDx9xb+bV9ts2MmYs0s80kXRqWV652NGb2ntj51XCTW8cJD6GycgfwWnnO/ur6efDQm57XP0LuwvhL63eSHh9t/8pxd0lao8+2OyVF5c+3TFkAgzL3JUlX99h2v6SoUaaZfR+Pd57DOc/MTpa0U4SM2b95M5uGh33tHaZhboydmw/m4XUkPVtbPwVXMqKya/aQ26gjzHivdmHwFFmjRD9BdqoZfyv8+fld+P4VPJ/I73A/o+g04JYhQY+Z3S5pzbD8BWANSR80s8XwqrIxUyU5z2kG8GZJf62tfxEwXdK6sbJ6yF4ED03cUdKWo+0/QE6T6YmxVcI0QXGnw/YhQ8w73dW/clVobFMFcbPK8sq1beMei407+yVv67Hv1sAD4WHYArgfD7N8FPhQgpx76FFVDQ9x6lsZs8f+s/CUx/eFz/2VT3QlQ2BJYOEmv5PcHypx0vi0yzaV79H5BBhQlnnQth77noMrluAx3Y8Av8B9RPYd73vVR/YSkF5iu4ecpFwkuAVj4bD8rvC7Xjc8pxckypoVnqvjw+eEyuf4SBk3V5YvAXbotW0cz6lvzP+gbQOOWQAvZ30a7lh9ArDVeN7zcMztleUvAD8Iy4s1Oa855LcVMLd8cG/qj/ZYvyse9pJ0g+s3u+HNXwfPlHgz8H3gNYnHD5uycj3wqh7rX0WtnO8ocm7BcyOsj1c0WyWsXybxhfOl8NJZqbJuJdwP4SsJco4MbfoOHirV+gWR+wPsnrj/j4BD8TDBP1Y66sVJVA6olN2trF828V5l6Qhz3SvgK/iIGDxJ1KXAn/EKmm9tea82SNy/WrL5eOCzle9JzzkZEvTgSts+Qdb/AYuH9QsROQjIfE534lFV9fWLkVAiG/evOh4Pm/0RnjPmgTb3uiJ7rwbHtFbCBsrPcWJzwyd0VlcDlwGHhc/leBz1cpEysryMgYNwp6kf4VrzfA3PadiUlXfi4Wy74A6ar8Odc+7BQ8qanNfMftsiZX0cH0n+KXx+xwRlAcRHS70+M2k5EmjQoS6El8w9Enh9Zf1GJCiFwAfDi+bfQ2e8GO70eR1pVp5sHWGme3U7I9Oyu+NWtXnxUs7XJchpnWkx/EYWxR34fgesV9l2R8PfyyJ42uyzcb+Kf084dhk8zPls4O2V9ZsC+4/3OeGDmfOYcxBwLvDpBDmz8HfCypV1raxNteu9M3BuwjGtlbBBnxKtEJD0R2Cj4IXcmfs+V8GHIJJVzKvZWWWZ8D3FK/nLuMnz9eHzdfNS30lVEOmeE9WAbQOxOWOxP68GJZslnWdm2wCfZsTj+HZgO0kzE0TNY159bB48Y161xnpSfXZJRwNHh/lQVJuXTJAj2mcBPnsAbwAAHRxJREFUnIXfl5PxB//vg3dPItWrelE8KVC9lsJTQGymTyT9wLyI01fx50r4PT9AtWqYo/BgcGJ8CFdSzwcwL689f4KcXPfq2SAH3PnzVEnPA3cGp99YcmRaPCIc/xRwp0LURAg1fiShLVX+gRe0egp4OT7VFoW8zPjHeqyfbmaxiYeynZOkQ83saTyz66L4dX0Gd0Y+NkHUuvjv5WIzuw/vC6OcunsR/He2wJWwd+BZaL878KBudsWfq7cC20t6Mqx/Az7V0YrikBjI4ahkmarq5YgTDnKeBH6FdzhvCsuE7xtLWiJSTpaSzbmwTHXVg6x5cf+JP4XvC+CWjf0U75CYLQugeUbLHXGT5R24onBhE2WsJjc1Q+KpwLH136yZbY6P+Ed1asyJmS2Dd4TLA8copDoOyvy6kg6NlJPlXpnZNfj89x/xok/rKjjJDXJ07SEnS6ZF89TLy+Dm+Flh3XLAAkooNNQjRDM5QY+ZXSlp47D8Q0kfqGxLSb6W5ZxqMhfF33uNBgEVOW/Er9N2uBJzpiILm5lZp6DZ5rjF6SfAUZJWatOm2v84VNL+rWQU5cCxzKVlK3JbV9Vr4Zk9bMrKxrh/wA/C99NxJzyAgxOtNK0xsx2A/8FHEb/BY6t/iPtGTIu9bzZGWQDNbHvgGOC/JH2riYweMmMzJM72OO+xrW9kT2QbspYxT+kIc90rM3sDcCKuYBwhaVpYvwU+7bJj9Akwh9KSlGkxPJ9PyusHdF7w2+Dm+KNVixQZRVbrEE3rzkLZda8tMklPznPqI/8cSe9qIyPImQf3RdhBUlT+gnCNr8DLxXcUyvuavmP6/I/Wyc7KtMIIuUvLdvgebgaNZlCIiplFh6j06lyaKCv1l39TZQX3pagmMFkdH6UvgjuZNVIOwmhgNXz+78nR9q/wJXzEd695Lvdf4w95dJKfwE/xTnSN8KkSaxoGZo+WdmBkHnE/PPd/Lg4izuQ4yFSfZMbvQa5iUh3eh88rx5DlXkm6psfxSPolXmY9lcZmfNxrflvgL2Y2FT/Hb+BTkt/BLRyxfIT2xd0GHR8rO+c59eJlLY+v9jvXKq0mQtbpiX7Nay2hrdPC3PQhg6NSD5lDFaJCgygF8oWRXV/7/rPK8lUJcr5TWd4YdyicjhdzSnFsrDuNRnsuj8UHd3a6CXcEXBW3qsz+JMgZ5Nj4z0gZ5/a6lrhT6Xktz/PgzNftwQm4V0dUlj9Z23Zigpxqf3MoFce7lPtdWT4UT2kO7n/TPqQtMUQTn4LcFje538eIg+V2wG+H5JyiwjJrx2Tpd2oy3wgcHfrU80iIKqr3D5XPUsBDba9RsRxUkF/xXKVlOzSpF/CvyvJbOv9f0l+DSaoNTTTKlTVSle/DwEWqJDXBnYdi6EoBLKmacCWlHOwbKsvT8Bj8m8wLKJ1G/MhtGfMMfh0WrX5XZCGojAlbXoGPrPbAPeBh5H4JiDU7LsuADImRMvYFzjUvj9yxDq2HV0NsZY5Vg6yGA3yCjITftOXL2PjmyvKH8KiODik+OJcwYsafAnzQzD7Y2ai4TIvV898Mr5iKpFnBkTma8Ns9TdJd5omqzsOLtj1nZjtJujhCzOV4LpLOcjW77K/m3L13UyrLrc6pFwo1PczsJ5K2jzwsV79TbcdVwFXmRfLehr93onwX8OdS9P79/6vHuiSKchDo46i0jhIclfrMk20afjwp82TZPLN7MJHKyl1mtqWkrvzf5vUn7m7QLoAXKUyTSLovOBjG8r+4Nabf91i+Rug4wrnsjDscrY17H28eI0T5HJJy1HoQbmJ+FSPRO5fjIazPxzbEPCV01ZRsjHRoUmR1R/J1hB+h9zTfcbivSaxyYH2WU8lhxr/UzE7DR59LEKbnwpRk6tz89viLD1zpMbw/XA04CXdSHIgi595HIec5DWLD0XfpSZt+p4um0xOKzJjalKIcjPAYczoqrW9m60O0o1KuebIsISpDqKx8CjjHzP4DN5+Dz79tRNpodA3z4kIGrGRmS0j6v+AcFN0etagCOaco/S0svwf4vtwX40Yz2ytWyAAnrAdwD/2o+yVp1wHbYqMMjsBrjXT93sxrjRxB92hwEJcAy+HK9qlq6GWesSNUr+so6Z+WNiSthtN2ljvHR78oJJ1YXxdkPRksmTHsi7/Ul8ejkDrK0nLAF2PbEsgVotlJkb0dnlNg9rGSvhpxeM5zykWWfsfMviNpr7C8MR6R9FtgVTPbQ+630ggzeyVufdhRLZyGoSgHVXI4Ki0kqVPBcWd8Xuuw8OOZYxTXD+WJE4YhU1bkRXzWYqREMriZ8WOS/hErB080U+WZ8HdJPHNdFJav+qUF7f9vuGWlWoglh2PZVBLu1wATPBBdP2Al1YqQhWNvMC//HYWkbczsxbjS9L9mtiAeunVqZDv60rQjNLNl5XlNutYl/vtqdUcYUXYhLYdIDjP+6pJODfKmzG6ENCNEVaTwTzN7LR6iuSndjp4LJ8o6G3eyvJHexcQGke2cgrNxz02kDW6y9Dtknp4I1pTt8XwJa+F9RlK0TE/aOi2UT5eDyMzK8k3A5pXvKSler6ws/7C2LSXT4pg69QRZ0dkbcSe7N/ZY/ybglRNwvz406JMg5yO4tekm4PzK+rWBS8b7fpGhfgAZao30OG4evNP6E14ttImM5fFR5XW4h/8BwOsSjs+SsTHjb7B1pkUypknHX1x3AU8AX66s3wLPwZAi67YW1yXnOU0f9JmAe149txtr26Id2IGP4tMt9wAH44rB/bnaWSwHgaoTUA+kWrnPPuSaJ1ukslyPNU8xfWZx6rEBSU3wTjU2VPMIPPqizt9JMFUH017PTSQkZVKLynk1Oceb2QWEhC2VTY+QVrs9lxNWjrDc683so5K6nHHNbFdGHBSjMLONcKXgTbjj3baSUixgmJfX3RFYAR9d7QacrcSpIeXL2NirjQdKOjDxsBxm/EH+D0k/HOUN0bzazF6ntOynHXKeU5bS67n6HTJNT+A5UH4N7KSRDJJt/VdmU5SDEdbvsc7wF9bL8OQ4o5FrnmzQDU65+cOmrKykDKZqMqUZtpH01j2RtPWg7RU5mwOLSTq9tuktuC9L7Dx7lvsl6ZNh/nwT4APAUWZ2IZ7xMDZyYl/gTDN7P93RCgvgUx9RmGezfBKP5d6dkGSsY+pVfM6NbB1hUAJaKQJ92BpPpJVCDjO++iz3+j4QMztC0r5h+ZOSjqxsO1HSLgniNgZ2MbP78WmFlJdotnOC2blZdmJE8bkTOFlpU1u50pvnmp54KfBe4L/DtNhptHdYn01RDgKSZifnCR3r+4HP4qWAvxYpQ3gn2JFTTRaUkiRjcfO0qvOE5U7In+FznbEMm7IyaP59oVghkqbaSJrhk2meZnhDPEb5FOBamnudH0Rvq8cleAKjiyLlZHPCCr/FxmG5ylNrBNyZUvio+O10X2PhFpIYxqwjtHwZG5v8fvbF6yssDRyukYx5WwAzImWsEPxnrLLcaU9qsp9cIZrgOTGaku2czOzVuKJ9AX5NDR8MfsHMNpN0V4ycXP2O+mSUladwj06YFvY/FjjWzFbE+47HzOxOPJ1zLyttNCV9coVgxtsF+E/8ZfENSdEhdjYgsyEQndnQzAY6+alhqJA1T8N8H35N5sGrMnZGN4bPi78yUs4pwKV9TNVvV3y8cV1uozTD5uFHnTzna+GJf07RnMWGRpNza7/R0KBtPfZdo9NRmdkUSf+sbHtDMPnGyMlW62EYqXSEO+Kj61YdoUWm9I2QM49CDYDxxMw+NGh7yvSZdac+7rouTZUo87oYswcGiohYyXxOp+NOn6fV1m+HW6K2i5VVO75pv5NreqKf/NXxTK+torGKchAws72BT+KjvUP6aXejyJidj97MvoDXe5+dLKjtTW/QnqFSVsKI70zcRD6HqVrSozFygqx6muHT8JfE07EyavKm4C+bbwFflXRUwrH34OWrn6utnx8vL/uqSDmzO996R5zSMdsY1XrIiXnxmc9IeltLOa07QjM7WImJmczsKAZYzRQZ6ZLDjG8eAbKYpMdr65cBnlJCJJCZ3YJPR82Dj7Y3YcQiMl3S6xNkbQ0chlt9HsOTfN2pPjU7asfmPKe7Ja2euq3P/q37HTO7mQHTEynvnnA99sane4VbM76jWjROE8q0wghH4T/gjYFfVBzAUrS5MctsaM0KhWTJbNjUUtFDThZTtZldjnuZn4ZbejrzhguY2ZIp84hBKdgSVwxWAr5Ngmkv8DM8TO/jkp4JchdpICuXE1a2Wg9tMbPN8ERQLwXOAr4O/AA/n6jpuoqsfh1hqxFSqmIQSKpUOIAcZvxv47lH6vf1bXh/tmdCe7KEaAam4dEPF0ta20YqPsaQ85yeabiti1z9Tq7pCfPKkCfjBcA6z9Q6wLVm9n559sXGFMtBwDJUHjSzXwAX4smCjsdfzk+aJwu6IUZjHiA72fRpZjdLmhqWLwH+VyOxw7O3NWxPrqpmuyuy1GnlmAcY6ajmyL6nyOpmZnYSrqSch3uJ3zbKIf3kzIeHEu2Gp0wGL57zfTwcLCqDXy7LwTARfB72w50J34l3Yl+ujpAj5VQ7ws7Lax38hRrdEVq+jI2Y2dL4aPhepRX8qspobcY3szskvabPtr7VNccaM7tB0nrBGrG2POrmOkkbRByb7ZzM7CGgVyp0w2vDrBgp5wEy9Ds95DadnrgG2FPSjNr6qcD/SPq3Ju3pUCwHgSbTCD3IkiyoD7HOSVXGMg1z66pmgY8Rn0scyJpm+AP4yGE14BM9rEVRL4qg7X/OzA7CczmAvzBSvZmzOGFZvloPOZCky8LyWWb2eKpiEDgMTxZTfQ7ONrMz8bLbsR1hloyNZrYbbgX5LbByUHIHRr/0IUemxUFWpXkatKlbeLMQTYAnzZOD/Qr4sZk9RohWifm3A7alntOgtOjfixWSsd/pNT3RpPrqi+qKAYCkm4N1uBVFOQj0GFHM3kTki0IDMhviCTcao2aFQoZNWelFsoe39QkdNLOdgMclRUUHSGrdcdbk/R2vfIiZHcdI8aRYPl1ZrputU8zYWWo9ZKIabROaNPI9wf8hS0eofBkb9wXWlPS4eWa7HwNNlIMcZvzHzGwDSddVV5qnfn+8zzEpJIVomtmqePGvd+Pz6fvh0V+voLtk+yCynVPbaafK/87S72ScFjULORJqK5ckh1JYphXyEaYVBjkpRcXNj/I/fi/p5W3l5CJRWel1/AqSHko85hpgqx7OSsvhzkFNi6lko+k0gHkExSGSPj3qzv1l3NJxHDOz44G7Jf1Xm3a1aMsgJVQdpTdCzp3ARn06wqslzZG4J0LmPHjUw1HA1xVZhTMcOzTTPma2Af6iOZFuR98P4s6a17aUnzSlae4I/QXVcpqY1+U4QNKoyc5ynpOZXSjp7WH585K+EXtsTU6WfifjtOjueJbE/emuVfNfeOr+/4mR049iOcjLoT3WdW5++zqjiYyHskJCVTPrLo9cXd9pT2znvHD9AQ3HPxocAYeBx5ocJOl5M1u35f82y1ProTXK5MwKHA5caGa9OsLDUwRZhoyNdE/7zPFd8XU5erUvyYwv6brwMt0bH4mCZ338t2DNbEvq77F1srPM57R0Zfm9eO2BJmTpd3JNT0g6zswexh0/Oz4YtwMHS/pFW/lFOcjL4sAKko4BMLPr8B+m8IRKUVi+QiFDpazQrBxyLxY0s/nUO3QwOplSLqxH2JWkd1iDsKvADPPsjT+l4k2dYII/Ai/09RQeOtbJKLg2nn1x3LA8acmzdYSWL2Nj3bKTlFJ6FFLN+EsDL5F0QG39mmamXi+0AbJ6hmhWFPgYpad1srOc50SDjIp9yNLv5JqeAJB0Dl6iPTtFOcjLZ3Ankw4L4KawRfA5/p9GyjlswLaobF6BoVJWBs39JY74c4UO5iJn2BV4GtUn6M4eGB2CqHy1HnKQIy05kK0jfIAMGRuVqS5HH1IV96PwTHl1VsAza8aW6YY8IZo56nLkPKdVgrJtlWUYMePHWlBz9Tv9MqpeSkJGVTNbEy9Y9/Pw/XBGMugenaDo9pZffA7yYWbXS1q/8v1oSR8Py9dISi2f2rY9V+Hzcw+G7zfjZuZFgBMkvSVSzkBnSiUUNgleusvj3vTPhtH1vsAukl4aKSNL6GAubMhCyQaMTN4PPJYyMsncrmpa8juAr/UyP/c5dkw7wlQsU12OPrKTMi0O+o2Z2W1KKGcdjmkVomkZkp3lPCcz+/ceq2dbUCVdHiknV8hyroyqv8Cz+F4dvt8BfBnPGrqdpG1i5PSjWA7yskT1S0cxCCxNApanUMgC6k6Ze6WkJ4AnEufIclU12xfX+u8FppjZkXj88Q9ImNdUvtDBXGQNJTOzFfCR0xvxTuxKPAQx1nEzV62HLNicacn/QwlpyQOH0D1XvDkjHeFXgFYdoaVnbMxSlyOTGX+Q9S4pZNkyhGgqT7KzbOdEJgtqxn4n17To8h3FIPCUpDOCrD0S2zQHRTnIy7V9zGl74KWNo7BMhUIYPmVld2B1SX82s5fjSsKbFVkzoI7ahw7mInco2Ql4wp/3hu87h3WxL66hcdi07rTk71DzfCJZOkLLl7FxOUbqcuxEw7oc5DHj/8bMtpCXVZ6Nmb0TuC9RVq4QzbYh3DnPKdd0L5Cl38k1PdHlw1WzTC+T2KY5kVQ+mT7hhlyNPxCHhc9leHa4ZRPknA68r8f67YAzEuT8GPhoj/V74B1ZrJxX4/PVJ+Id/b7AScDDeP2IWDk31b7flvHa35RLVoP/vQE+l30gPmLfCh+93497V6fKuzlm3YDj7wHm67F+fuA343xtZgGP4p3prfVPgpy7B51vgpwZeL2AKbi14SncKtPmHKfglpHHgX0aHL80/rJavOH/Xy3c8xPxPAL7hOfzHmC1RFn1Z3RCnqvM53R97fvRleVrWrYz+frgg/JDgD/h0y43ht/OIcD8CXKm9+pf8Bwnl7W9B8XnYAwIo5PZHtVKrx2QpVBImM8/C6+lXg3/moJnm4sqzmGZqpqZZ0g7tbJqh+p3tQv/Ol/SO5oe3xYbyfvfMaPejndCyaFkZnYx3imeElbtCHxY8T4ih+BJaHqNTP4kKdqU2hYze1VoS70i5CuAhyXdGylnOl5E7Nra+jfgeSE2iZRTz0/wW0VWFe0hq16X4+d4fPkfEmR0mfGBRpkWQ1t2wn9/wn9/vwF2lLR3gpwxe0ZTyXhO90patc+2xvc/HN+43zHPVNt4esI81PMneF9R7d8/hCe9i7ZW95RflIPho96BxW4bIG9YlJVsZVgH/I9DJe0/+p75CA5cS0u6o7Z+TdwBMGlqIUy5HI3Pawu3Rn1SkSb5YXLYtAwJccL+WTpC8/Lj1d/HodXvigwXtXx1OW4DNlXFjK8WSbzMw1V3BN6HW67OkHR0wvFj/oymkuGcfoyPpHtN924iKbYY1KD/0arfMbPjJCVPi4ZBycfpDu89JnbgN1B2UQ6GD8tUKCRje7IqK2OJTUAGSTM7FThWNa/nEDXwIUkpYVc529VqZJKpDX09y81spqTXJchq3RFavoyNsxjJQdEry11UXY4elowmyv9q+Ah/RzwE9ifA/pIGFpMbZnKeUy4L6ij/o1W/0/C+v1wNa4NEyS/KwfBhZgcM2q5MucJjyaWs2BiGf1X+x4MToDxlCbsyL5DUD0ma1qiBNB+ZtGUUk27fbT32HdOOcKLIYcYPisoVwK6daRozu08NqgSOxzMa2Y5s51SR2cqCOorsVv1Ok+kJ667iekbs9G4sJVphCBnvl38EWaqakS/8a8l+m5rKbEmusKteteUXwQtoLYVnB2zKei2ObUOOhDjgI7/WHaFlytiYkRyZFrfDlYrpZnY+rlw0fQ6yPKMZyHlOAARloLFCMJb9TkcxSJyeqP7PxkpTX+HFcjB8WKZCIcOGeVGhTvjXWjQM/zKz+3FTbq8HUm1GF00ws3Nx83avsKtPSHpnA5mL4ZEhu+IFaA5r4txYkTchDpuWISFOkDO7+I8lFgKqyTmq12pCxkZJk3bAFJxOt8Gfr81w7/4zJV2YICPLM5qLHOeUsS1j3u+kTE/ULAfZp3eLcjCE1DrCCZ/THwtlJXgi7wh8C/iqpF6d9qQgzI+egzsOVl+AGwLvknRPgqwlgU/hmQRPAo5UrRJhhIw5aj2E9U1rPbTGuhPiNHGKzd4RmjXP2JiLsTLjh9/Re3FnzaiU0D1kDNUzmuOchp2U6Qkzex63NhqePOlvnU0k+L30lV+Ug+FjrDXCBu3JpqxYnvCvVwBPSvpL+L4pPrp4AB/BP9u0fU2x7rArcGe5k1NexGb2LeA9wHH4eTzdsC3HAefXPe/N0ydvLCm11sOEk7MjtDkzNn5D6Rkbs2BmjzPAjF93ch2nNrV+RudGcvU7o0xP3CJphQzNbU1RDoYQM3sS+BX+Y3lTWAaSC4Xkak8WZSVj+Ne1uEn6YTObClyMp9ZdC/iXpN2ayM2Jmb0EeEIJD1hwwvonXimwjQf8UNV6GCasO2PjIWqesTFXe4bNjJ/lGZ0bydXvDNu0aD+KcjCEWKZCIRnbk0VZyRj+Nbs4iZkdCsyS9BkzmwfPJBhVuCQXFhLxAH/GnQZ/CLwEr6vwQUnnj3N77pT06tRtLwTCb/AxPCPdHJ3feP92qgyDGT/XMzo3Mmz9zlgzaZ1v5nKyFArJyLsry4eGv7OVlVghkpKLEPWh+j83Az4f5M8ymxDn6qOBL+BVAi8F3inpGjNbAzcXj6tyQP5aD3MTqzMgY+P4N6enGX+iSo/nfEbnRrL0O8M4LdqLohwMJ1kLhWRg2JSVS83sNLzewxKE8CQzWx73ih9v5ut4T5vZVxUKSUm6a4KUlU8Dp5nZiXQ7SH6Q7t/VC5HD8YyNXdMJ5lkuD6d3Ncsxo2bGP6iY8YeaXP3OacC2wF/C9MRP8emJqcB38MymE05RDoaTLKWWMzJsysq+wPbA8riDXScd8HJ4SejxZlZluZ6FcNzn7SRdZ55ueG/c8Q7cQfLf2oRDziWs1CsiQdINZrbS+DeHD+Bm/NWAT1SUyRe8GX8IydXvLCSpY6XaGXf2PKwzPZGttS0pysFwkq3UciaGSlmRJDP7BzAvPur6Q1g/Y7zbEni9mT1F8KQPy4TvC453Y8Io+CWSDqitX9PMVA9xfIEx6H4sNG6tCBQz/uQhY78zbNOiPSk/zOHkWjP7aH2leaGQVpW2GjJUyoqZHQvsR8gaaGZfHu82VJE0r6QXSVpM0nxhufM9JUNiLo6i931ZAThynNsybFzf59lKzdhYeIGRsd+51MxOM7MjGY5p0Z6UaIUhxMahUEhie8a8qllie24DXi/peTNbGLhC0rrj2YZhZlC4oiXUepgbsUwZGwsvPHL1OyH5Vmd64rRO/gjz6pPLSLogZ7ubUqYVhpAwL7yRdRcKOVcZC4Uksh9wlpntRA9lZQLa86yk5wEk/c2GyRY3HOSq9TDXERTrjWoZGyfy2SpMHrL0O0M4LdqTYjkoRGNjWNUssR1/A+7tfAVeGb53nLjmqnjjVGwMaj0UCi90cvU7YXriNXi69bcAv1CLiqtjRVEOCpOOECfcl4nOejfRWMZaD4VCwcnV70yWadEyrVCYdLzQX/6jIekeM3sd3bUeLgf20AQUXSoU5gYy9juTYlq0WA4Kkw4z+yu98weU2PA+NKn1UCgURsjV70yWadGiHBQKcxnDVuuhUCiMMFmmRYtyUCjMZZjZDYzUejiOWq0HhfLbhUKh0I+iHBQKcxlmdrOkqWG5qwqjmc0oykGhMHFMlmnR4pBYKMx9DFWth0KhMIKkxSa6DTEUy0GhMJdhZs/jxXwMrxfwt84mYMEJSulcKBQmEUU5KBQKhUKh0EUpvFQoFAqFQqGLohwUCoVCoVDooigHhUKhUCgUuijKQaFQKBQKhS6KclAoFAqFQqGL/wd6w0aCdANMQAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f1313d3c2b0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# sort models by ECS\n",
    "mod2 = [k for k in sorted(ecs2, key=ecs2.get)]\n",
    "    \n",
    "fig, axes = plt.subplots(nrows=3, ncols=1,figsize=(8,8))\n",
    "plt.subplots_adjust(bottom=0.25,hspace=0)\n",
    "\n",
    "good=good_list.theta_detrended\n",
    "\n",
    "# theta comparison\n",
    "sca(axes[0])\n",
    "\n",
    "low,high=np.percentile(slope_list.theta_detrended,5),np.percentile(slope_list.theta_detrended,95) # 5-95% uncertainty\n",
    "avgceres=np.average(slope_list.theta_detrended)\n",
    "\n",
    "for ii,vv in enumerate(mod2):\n",
    "    plot([ii],feedbacks[vv]['piControl']['flux']['t_trop'],'k^',markersize=8)\n",
    "\n",
    "fill_between([-100,100],low,high,color='0.9')\n",
    "axhline(avgceres,color='k',ls='dashed')\n",
    "\n",
    "plot(100,0,'k^',label='good',markersize=8)\n",
    "plot(100,0,'k+',label='bad',markersize=10,mew=4,color='0.6')\n",
    "\n",
    "gca().set_xticks(range(len(mod2)));\n",
    "xlim(-0.5,len(mod2)-0.5)\n",
    "# legend(loc=4)\n",
    "grid()\n",
    "ylabel(r'$\\Theta$ ($W/m^2/K$)',fontsize=14)\n",
    "# ylabel(r'$\\Theta_{iv}$/$\\Theta_{forced}$',fontsize=22)\n",
    "gca().set_yticks([-1.5,-1,-0.5,0]);\n",
    "\n",
    "# ECS\n",
    "sca(axes[1])\n",
    "goodECS=[]\n",
    "for ii,vv in enumerate(mod2):\n",
    "    if vv in good:\n",
    "        plot([ii],ecs2[vv],'k^',markersize=8)\n",
    "        goodECS.append(ecs2[vv])\n",
    "    else:\n",
    "        plot([ii],ecs2[vv],'+',color='0.6',markersize=12,mew=4)\n",
    "\n",
    "ylim(1.8,4.5)\n",
    "gca().set_yticks([2,2.5,3,3.5,4,4.5]);\n",
    "        \n",
    "xlim(-0.5,len(mod2)-0.5)\n",
    "ylabel('ECS (K)',fontsize=16,labelpad=16)\n",
    "\n",
    "gca().set_xticks(range(len(mod2)));\n",
    "grid()\n",
    "\n",
    "plot(100,0,'k^',label='good',markersize=8)\n",
    "plot(100,0,'k+',label='bad',markersize=10,mew=4,color='0.6')\n",
    "legend(loc=4)\n",
    "\n",
    "# lambda comparison\n",
    "sca(axes[2])\n",
    "\n",
    "low,high=np.percentile(slope_list.lambda_detrended,5),np.percentile(slope_list.lambda_detrended,95) # 5-95% uncertainty\n",
    "avgceres=np.average(slope_list.lambda_detrended)\n",
    "\n",
    "for ii,vv in enumerate(mod2):\n",
    "    plot([ii],feedbacks[vv]['piControl']['flux']['ts'],'k^',markersize=8)\n",
    "\n",
    "fill_between([-100,100],low,high,color='0.9')\n",
    "axhline(avgceres,color='k',ls='dashed')\n",
    "xlim(-0.5,len(mod2)-0.5)\n",
    "ylim(-1.8,0.1)\n",
    "grid()\n",
    "\n",
    "plot(100,0,'k^',label='good',markersize=8)\n",
    "plot(100,0,'k+',label='bad',markersize=10,mew=4,color='0.6')\n",
    "ylabel(r'$\\lambda$ ($W/m^2/K$)',fontsize=14)\n",
    "gca().set_yticks([-1.5,-1,-0.5,0]);\n",
    "gca().set_xticks(range(len(mod2)));\n",
    "\n",
    "sca(axes[2])\n",
    "\n",
    "gca().set_xticks(range(len(mod2)));\n",
    "gca().set_xticklabels(mod2,rotation='90');\n",
    "\n",
    "for ii,ax in enumerate(axes):\n",
    "#     ax.axvline(14.5,color='k')\n",
    "    ax.text(0.05,[0.8,0.8,0.8][ii],'(a),(b),(c)'.split(',')[ii],fontsize=16,transform=ax.transAxes)\n",
    "\n",
    "# savefig('fig9.eps')\n",
    "# savefig('fig9')\n",
    "\n",
    "print('average ECS of the good models: {:.2f}'.format(np.average(goodECS)))\n",
    "print('std of {:.2f}'.format(np.std(goodECS)))\n",
    "print('range of {:.2f} to {:.2f}'.format(np.min(goodECS),np.max(goodECS)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## other stuff"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### look at CERES regression of annual average values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x7f547b8217b8>]"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f547b4538d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f547adbb6d8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import xarray as xr, pandas as pd\n",
    "import xradd\n",
    "\n",
    "ceres1=xr.DataArray(CERESdetrendedFlux,dims='time',coords={'time': pd.date_range('2000-03-01',freq='M',\n",
    "    periods=len(CERESdetrendedFlux))},name='flux')\n",
    "ceres2=xr.DataArray(CERESdetrendedTa,dims='time',coords={'time': pd.date_range('2000-03-01',freq='M',\n",
    "    periods=len(CERESdetrendedTa))},name='ta')\n",
    "ceres3=xr.DataArray(CERESdetrendedTs,dims='time',coords={'time': pd.date_range('2000-03-01',freq='M',\n",
    "    periods=len(CERESdetrendedTs))},name='ts')\n",
    "ceres=xradd.annual(xr.merge([ceres1,ceres2,ceres3]))\n",
    "\n",
    "figure()\n",
    "plot(ceres.ts,ceres.flux,'o')\n",
    "figure()\n",
    "plot(ceres.ta,ceres.flux,'o')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "median: -0.70\n",
      "5-95 percent confidence interval: -1.46 to 0.06\n"
     ]
    }
   ],
   "source": [
    "## regression of flux vs. Ts\n",
    "slope=monteCarlo1(np.array(ceres.ts),np.array(ceres.flux),runs=numSortRuns,autocorr=True)[0]\n",
    "print('median: {:.2f}'.format(np.median(slope)))\n",
    "print('5-95 percent confidence interval: {:.2f} to {:.2f}'.format(np.percentile(slope,5),np.percentile(slope,95)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "median: -0.84\n",
      "5-95 percent confidence interval: -1.36 to -0.33\n"
     ]
    }
   ],
   "source": [
    "## regression of flux vs. Ta\n",
    "slope=monteCarlo1(np.array(ceres.ta),np.array(ceres.flux),runs=numSortRuns,autocorr=True)[0]\n",
    "print('median: {:.2f}'.format(np.median(slope)))\n",
    "print('5-95 percent confidence interval: {:.2f} to {:.2f}'.format(np.percentile(slope,5),np.percentile(slope,95)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Conclusion: using annual average in the $\\lambda$ calculation gives an uncertainty about equal to using monthly average.  But using it in the $\\Theta$ calculation gives a much larger uncertainty"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "LinregressResult(slope=-0.018490778438178007, intercept=-0.9837354919838726, rvalue=-0.028186573712292468, pvalue=0.8936051575964492, stderr=0.13673396294099832)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f4cf6b2d518>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "p1=np.array([feedbacks[ii]['piControl']['flux']['t_trop'] for ii in modname])\n",
    "p2=np.array([feedbacks[ii]['abrupt4xCO2']['flux']['t_trop'] for ii in modname])\n",
    "figure(figsize=(4.5,4))\n",
    "plt.subplots_adjust(bottom=0.15,right=0.85)\n",
    "plot(p1,p2,'o')\n",
    "xlabel('$\\Theta_{4xCO2}$',fontsize=16)\n",
    "ylabel('$\\Theta_{iv}$',fontsize=16)\n",
    "one2one=np.linspace(-1.6,-0.2)\n",
    "plot(one2one,one2one,'k',ls='dotted')\n",
    "colorplotRight(gca())\n",
    "print(stats.linregress(p1,p2))\n",
    "savefig('corr')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "average ECS of the good models: 2.90\n",
      "std of 0.48\n",
      "range of 2.04 to 3.86\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAgUAAAHYCAYAAADd1JthAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nGgT8PyWtK+8rFzv7AAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fa0cd458350>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# sort models by p/a\n",
    "mod2=[]\n",
    "for key, value in sorted(ecs2.iteritems(), key=lambda (k,v): (v,k)):\n",
    "    mod2.append(key)\n",
    "\n",
    "fig, axes = plt.subplots(nrows=3, ncols=1,figsize=(8,8))\n",
    "plt.subplots_adjust(bottom=0.25,hspace=0)\n",
    "\n",
    "good=good_list.theta_detrended\n",
    "\n",
    "# theta comparison\n",
    "sca(axes[0])\n",
    "\n",
    "low,high=np.percentile(slope_list.theta_detrended,5),np.percentile(slope_list.theta_detrended,95) # 5-95% uncertainty\n",
    "avgceres=np.average(slope_list.theta_detrended)\n",
    "\n",
    "for ii,vv in enumerate(mod2):\n",
    "    if vv in good:\n",
    "        plot([ii],feedbacks[vv]['piControl']['flux']['t_trop'],'k^',markersize=8)\n",
    "    else:\n",
    "        plot([ii],feedbacks[vv]['piControl']['flux']['t_trop'],'+',color='0.6',markersize=12,mew=4)\n",
    "\n",
    "fill_between([-100,100],low,high,color='0.9')\n",
    "axhline(avgceres,color='k',ls='dashed')\n",
    "\n",
    "plot(100,0,'k^',label='good',markersize=8)\n",
    "plot(100,0,'k+',label='bad',markersize=10,mew=4,color='0.6')\n",
    "\n",
    "gca().set_xticks(range(len(mod2)));\n",
    "xlim(-0.5,len(mod2)-0.5)\n",
    "# legend(loc=4)\n",
    "grid()\n",
    "ylabel(r'$\\Theta_{iv}$',fontsize=22)\n",
    "# ylabel(r'$\\Theta_{iv}$/$\\Theta_{forced}$',fontsize=22)\n",
    "gca().set_yticks([-1.5,-1,-0.5,0]);\n",
    "\n",
    "# ECS\n",
    "sca(axes[1])\n",
    "goodECS=[]\n",
    "for ii,vv in enumerate(mod2):\n",
    "    if vv in good:\n",
    "        plot([ii],ecs2[vv],'k^',markersize=8)\n",
    "        goodECS.append(ecs2[vv])\n",
    "    else:\n",
    "        plot([ii],ecs2[vv],'+',color='0.6',markersize=12,mew=4)\n",
    "\n",
    "ylim(1.8,4.5)\n",
    "gca().set_yticks([2,2.5,3,3.5,4,4.5]);\n",
    "        \n",
    "xlim(-0.5,len(mod2)-0.5)\n",
    "ylabel('ECS',fontsize=22)\n",
    "\n",
    "gca().set_xticks(range(len(mod2)));\n",
    "grid()\n",
    "\n",
    "plot(100,0,'k^',label='good',markersize=8)\n",
    "plot(100,0,'k+',label='bad',markersize=10,mew=4,color='0.6')\n",
    "legend(loc=4)\n",
    "\n",
    "# lambda comparison\n",
    "sca(axes[2])\n",
    "\n",
    "low,high=np.percentile(slope_list.lambda_detrended,5),np.percentile(slope_list.lambda_detrended,95) # 5-95% uncertainty\n",
    "avgceres=np.average(slope_list.lambda_detrended)\n",
    "\n",
    "for ii,vv in enumerate(mod2):\n",
    "    if vv in good:\n",
    "        plot([ii],feedbacks[vv]['piControl']['flux']['ts'],'k^',markersize=8)\n",
    "    else:\n",
    "        plot([ii],feedbacks[vv]['piControl']['flux']['ts'],'+',color='0.6',markersize=12,mew=4)\n",
    "\n",
    "fill_between([-100,100],low,high,color='0.9')\n",
    "axhline(avgceres,color='k',ls='dashed')\n",
    "xlim(-0.5,len(mod2)-0.5)\n",
    "ylim(-1.8,0.1)\n",
    "grid()\n",
    "\n",
    "plot(100,0,'k^',label='good',markersize=8)\n",
    "plot(100,0,'k+',label='bad',markersize=10,mew=4,color='0.6')\n",
    "ylabel(r'$\\lambda_{iv}$',fontsize=22)\n",
    "gca().set_yticks([-1.5,-1,-0.5,0]);\n",
    "gca().set_xticks(range(len(mod2)));\n",
    "\n",
    "sca(axes[2])\n",
    "\n",
    "gca().set_xticks(range(len(mod2)));\n",
    "gca().set_xticklabels(mod2,rotation='90');\n",
    "\n",
    "for ii,ax in enumerate(axes):\n",
    "#     ax.axvline(14.5,color='k')\n",
    "    ax.text(0.05,[0.8,0.8,0.8][ii],'(a),(b),(c)'.split(',')[ii],fontsize=16,transform=ax.transAxes)\n",
    "# if printf: \n",
    "#     savefig('Fig7.eps')\n",
    "#     savefig('Fig7')\n",
    "\n",
    "print 'average ECS of the good models: {:.2f}'.format(np.average(goodECS))\n",
    "print 'std of {:.2f}'.format(np.std(goodECS))\n",
    "print 'range of {:.2f} to {:.2f}'.format(np.min(goodECS),np.max(goodECS))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## plot $\\Theta_{iv}$ vs. $\\Theta_{4xCO2}$"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "## plot code\n",
    "missing=object()\n",
    "almost_black = '#262626' # tweek from Thorsten\n",
    "\n",
    "def plotFeedbacks(p,a,modn,ax=missing,title=missing,era='t_trop',pspines=['left','bottom']):\n",
    "    \"\"\"plot piControl feedback vs. abrupt 4xCO2 feedback calculated from\n",
    "    fluxvar vs. tempvar; era is a flat — if true, it plots ERAi feedback\n",
    "    spines tells you where to put the axes [right/left,bottom/top]\"\"\"\n",
    "           \n",
    "    if ax is missing:\n",
    "        figure(figsize=(9,7))\n",
    "        ax=gca()\n",
    "\n",
    "    ax.plot(p,a,'ko')\n",
    "    \n",
    "    invisible=set('left,top,right,bottom'.split(',')).difference(set(pspines))\n",
    "    for spx in invisible:\n",
    "        ax.spines[spx].set_visible(False)\n",
    "\n",
    "    # Only show ticks on the left and bottom spines\n",
    "    ax.yaxis.set_ticks_position(pspines[0])\n",
    "    ax.xaxis.set_ticks_position(pspines[1])\n",
    "    \n",
    "    # set axis positions\n",
    "    ax.yaxis.set_label_position(pspines[0]) \n",
    "    ax.xaxis.set_label_position(pspines[1]) \n",
    "    \n",
    "    for spx in pspines:\n",
    "        ax.spines[spx].set_linewidth(0.5)\n",
    "        ax.spines[spx].set_color(almost_black)\n",
    "    \n",
    "    for ticks in ax.xaxis.get_ticklines() + ax.yaxis.get_ticklines():\n",
    "        ticks.set_color(almost_black)\n",
    "        \n",
    "    ax.plot(np.linspace(-5,5),np.linspace(-5,5),almost_black,ls='dotted')\n",
    "#     ax.plot(np.linspace(-5,5),np.linspace(-5,5)*0.8,almost_black,ls='dotted')\n",
    "#     ax.plot(np.linspace(-5,5),np.linspace(-5,5)*1.2,almost_black,ls='dotted')\n",
    "    \n",
    "    ax.tick_params(axis='both', which='major', labelsize=14)\n",
    "\n",
    "    for ii,mn in enumerate(modn):\n",
    "        ax.annotate('{}'.format(mn), (p[ii],a[ii]),fontsize=9)\n",
    "    \n",
    "    if title is missing:\n",
    "        ax.set_title('{} {}'.format(fluxvar,tempvar),fontsize=16)\n",
    "    else:\n",
    "        ax.set_title(title,fontsize=16)\n",
    "    \n",
    "    if era != '':\n",
    "        obs=erai['flux'][era]\n",
    "        ax.axvline(obs.b[1],color='k',ls='dotted',lw=3)\n",
    "        ax.fill_betweenx([-10,10],[obs.b[1]-obs.conf90[1],obs.b[1]-obs.conf90[1]],\\\n",
    "                      [obs.b[1]+obs.conf90[1],obs.b[1]+obs.conf90[1]],alpha=0.1,color='k')\n",
    "\n",
    "#     ax.set_xlim(np.average(p)-1.1,np.average(p)+1.1)\n",
    "#     ax.set_ylim(np.average(a)-1.1,np.average(a)+1.1)\n",
    "#     xlim(-2,0.5)\n",
    "#     ylim(-2,0.5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x7fde7bb7bd90>]"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAdEAAAFMCAYAAACUO4YGAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\niv4AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fde7bbd1c50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "figure(figsize=(7,5))\n",
    "\n",
    "good=good_list.theta_detrended\n",
    "p=[feedbacks[vv]['piControl']['flux']['t_trop'] for vv in models]\n",
    "a=[feedbacks[vv]['abrupt4xCO2']['flux']['t_trop'] for vv in models]\n",
    "\n",
    "plotFeedbacks(p,a,models,ax=gca(),title='',era='',pspines=['right','top']) \n",
    "\n",
    "gca().set_xticks(np.arange(-2,0.5,0.5))\n",
    "gca().set_xticklabels('-2.0,-1.5,-1.0,-0.5,'.split(','))\n",
    "gca().set_yticks(np.arange(-2,0.5,0.5))\n",
    "gca().set_yticklabels('-2.0,-1.5,-1.0,-0.5,0.0'.split(','))\n",
    "\n",
    "xlim(-1.8,0)\n",
    "ylim(-1.8,0)\n",
    "\n",
    "xlabel('control $\\Theta$',fontsize=22,color=almost_black)\n",
    "ylabel('abrupt $\\Theta$',fontsize=22,color=almost_black,rotation=90)\n",
    "\n",
    "plot([np.average(p)],[np.average(a)],marker='*',color='r',markersize=14)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}