Add data processing exercise

e4efeaf5 · karin.meier-fleischer · Ralf Mueller · 6647b9d9 · e4efeaf5 · e4efeaf5
Commit e4efeaf5 authored 3 years ago by karin.meier-fleischer Committed by Ralf Mueller 3 years ago
--- a/exercises/data_processing.ipynb
+++ b/exercises/data_processing.ipynb
+{
+ "cells": [
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "# Exercise: CMIP6 - Plot Time Series of annual means\n",
+    "\n",
+    "<br>\n",
+    "\n",
+    "In this task, the global annual means of the CMIP6 variable `tas` are to be calculated and plotted.\n",
+    "\n",
+    "Since one time series alone would be quite boring we want to generate a total of 4 time series of <br>\n",
+    "the scenarios \n",
+    "\n",
+    "    SSP 1-2.6\n",
+    "    SSP 2-4.5\n",
+    "    SSP 3-7.0\n",
+    "    SSP 5-8.5\n",
+    "\n",
+    "For more information about the scenarios see\n",
+    "\n",
+    "https://www.dkrz.de/en/communication/climate-simulations/cmip6-en/the-ssp-scenarios\n",
+    "\n",
+    "The data are stored in several separate files and are to be merged with CDO and then their field mean average calculated. \n",
+    "\n",
+    "### In the exercise you will:\n",
+    "\n",
+    "1. Learn to use the netCDF input files\n",
+    "1. Process the data with CDO or xarray\n",
+    "1. Visualize the processed data\n",
+    "\n",
+    "An example plot of the scenarios of some more ensembles and their mean values can be found on the DKRZ <br>\n",
+    "page for CMIP6\n",
+    "\n",
+    "https://www.dkrz.de/de/kommunikation/klimasimulationen/cmip6-de/cmip6-aktivitaeten-am-dkrz-ueberblick\n",
+    "\n",
+    "### Input Data\n",
+    "\n",
+    "**Historical**:\n",
+    "\t/pool/data/CMIP6/data/CMIP/MPI-M/MPI-ESM1-2-LR/historical/r1i1p1f1/Amon/tas/gn/v20190710/\n",
+    "\n",
+    "**Scenario SSP 1-2.6**:\n",
+    "    /pool/data/CMIP6/data/ScenarioMIP/MPI-M/MPI-ESM1-2-LR/ssp126/r1i1p1f1/Amon/tas/gn/v20190710/\n",
+    " \n",
+    "**Scenario SSP 2-4.5**:\n",
+    "    /pool/data/CMIP6/data/ScenarioMIP/MPI-M/MPI-ESM1-2-LR/ssp245/r1i1p1f1/Amon/tas/gn/v20190710/\n",
+    "\n",
+    "**Scenario SSP 3-7.0**:\n",
+    "    /pool/data/CMIP6/data/ScenarioMIP/MPI-M/MPI-ESM1-2-LR/ssp370/r1i1p1f1/Amon/tas/gn/v20190710/\n",
+    "\n",
+    "**Scenario SSP 5-8.5**:\n",
+    "    /pool/data/CMIP6/data/ScenarioMIP/MPI-M/MPI-ESM1-2-LR/ssp585/r1i1p1f1/Amon/tas/gn/v20190710/\n",
+    "\n",
+    "### CDO hints:\n",
+    "\n",
+    "1. Merge files along time (mergetime)\n",
+    "1. Compute the annual mean (yearmean)\n",
+    "1. Compute the field means (fldmean)\n",
+    "1. Take care of the shapes\n",
+    "1. Important: clean temporary files\n",
+    "\n",
+    "### Xarray hints:\n",
+    "\n",
+    "1. Open multiple files at once\n",
+    "1. Compute the weights (http://xarray.pydata.org/en/stable/examples/area_weighted_temperature.html)\n",
+    "1. Compute the annual mean (resample)\n",
+    "1. Compute the field means (mean)\n"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## Import CDO module\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from cdo import *\n",
+    "\n",
+    "cdo = Cdo()"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## Import Matplotlib module\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from matplotlib import pyplot as plt\n",
+    "\n",
+    "%matplotlib inline"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## Input files"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "data_ssp126 = '/pool/data/CMIP6/data/ScenarioMIP/MPI-M/MPI-ESM1-2-LR/ssp126/r1i1p1f1/Amon/tas/gn/v20190710/tas_*.nc'\n",
+    "data_ssp245 = '/pool/data/CMIP6/data/ScenarioMIP/MPI-M/MPI-ESM1-2-LR/ssp245/r1i1p1f1/Amon/tas/gn/v20190710/tas_*.nc'\n",
+    "data_ssp370 = '/pool/data/CMIP6/data/ScenarioMIP/MPI-M/MPI-ESM1-2-LR/ssp370/r1i1p1f1/Amon/tas/gn/v20190710/tas_*.nc'\n",
+    "data_ssp585 = '/pool/data/CMIP6/data/ScenarioMIP/MPI-M/MPI-ESM1-2-LR/ssp585/r1i1p1f1/Amon/tas/gn/v20190710/tas_*.nc'"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## Compute the field mean of the yearly mean"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "tas126_ymean = cdo.fldmean(input='-yearmean -mergetime ' + data_ssp126, returnXArray='tas')\n",
+    "tas245_ymean = cdo.fldmean(input='-yearmean -mergetime ' + data_ssp245, returnXArray='tas')\n",
+    "tas370_ymean = cdo.fldmean(input='-yearmean -mergetime ' + data_ssp370, returnXArray='tas')\n",
+    "tas585_ymean = cdo.fldmean(input='-yearmean -mergetime ' + data_ssp585, returnXArray='tas')"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## Keep an eye on the data shapes\n",
+    "\n",
+    "```\n",
+    "print(tas126_ymean.shape)\n",
+    "print(tas245_ymean.shape)\n",
+    "print(tas370_ymean.shape)\n",
+    "print(tas585_ymean.shape)\n",
+    "\n",
+    "(86, 1, 1)\n",
+    "(86, 1, 1)\n",
+    "(86, 1, 1)\n",
+    "(86, 1, 1)\n",
+    "```\n",
+    "Remove the lat and lon dimensions\n",
+    "\n",
+    "```\n",
+    "data126 = tas126_ymean[:,0,0]\n",
+    "data245 = tas245_ymean[:,0,0]\n",
+    "data370 = tas370_ymean[:,0,0]\n",
+    "data585 = tas585_ymean[:,0,0]\n",
+    "```\n"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "To prevent this step use CDO's option `--reduce_dims` within the CDO command line.\n",
+    "\n",
+    "For example\n",
+    "\n",
+    "`tas126_ymean = cdo.fldmean(options='--reduce_dim', input='-yearmean -mergetime ' + data_ssp126, returnXArray='tas')`"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "data126 = cdo.fldmean(options='--reduce_dim', input='-yearmean -mergetime ' + data_ssp126, returnXArray='tas')\n",
+    "data245 = cdo.fldmean(options='--reduce_dim', input='-yearmean -mergetime ' + data_ssp245, returnXArray='tas')\n",
+    "data370 = cdo.fldmean(options='--reduce_dim', input='-yearmean -mergetime ' + data_ssp370, returnXArray='tas')\n",
+    "data585 = cdo.fldmean(options='--reduce_dim', input='-yearmean -mergetime ' + data_ssp585, returnXArray='tas')"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## Define x-axis"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "x = range(2015,2101,1)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## Create the time series plot\n",
+    "\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "fig, ax = plt.subplots(figsize=(8,4))\n",
+    "\n",
+    "ax.plot(x, data126, label='ssp126')\n",
+    "ax.plot(x, data245, label='ssp245')\n",
+    "ax.plot(x, data370, label='ssp370')\n",
+    "ax.plot(x, data585, label='ssp585')\n",
+    "\n",
+    "plt.grid()\n",
+    "\n",
+    "ax.set_title('CMIP6:  tas global yearly mean')\n",
+    "\n",
+    "ax.set_xlabel('years')\n",
+    "ax.set_ylabel(data126.long_name+' ['+data126.units+']')\n",
+    "\n",
+    "legend = ax.legend(loc='upper left', shadow=True, fontsize='small')\n",
+    "\n",
+    "legend.get_frame().set_facecolor('lightgray')"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## Delete temporary files"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "cdo.cleanTempDir()"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## Xarray not CDO"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import xarray as xr\n",
+    "import numpy as np"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "ds126 = xr.open_mfdataset(data_ssp126)\n",
+    "ds245 = xr.open_mfdataset(data_ssp245)\n",
+    "ds370 = xr.open_mfdataset(data_ssp370)\n",
+    "ds585 = xr.open_mfdataset(data_ssp585)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "### Yearly means"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "tas126 = ds126.tas.resample(time=\"Y\").mean()\n",
+    "tas245 = ds245.tas.resample(time=\"Y\").mean()\n",
+    "tas370 = ds370.tas.resample(time=\"Y\").mean()\n",
+    "tas585 = ds585.tas.resample(time=\"Y\").mean()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "weights = np.cos(np.deg2rad(ds126.lat))\n",
+    "weights.name = \"weights\""
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "tas126_weighted = tas126.weighted(weights)\n",
+    "tas245_weighted = tas245.weighted(weights)\n",
+    "tas370_weighted = tas370.weighted(weights)\n",
+    "tas585_weighted = tas585.weighted(weights)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "tas126_weightedmean = tas126_weighted.mean(('lon', 'lat'))\n",
+    "tas245_weightedmean = tas245_weighted.mean(('lon', 'lat'))\n",
+    "tas370_weightedmean = tas370_weighted.mean(('lon', 'lat'))\n",
+    "tas585_weightedmean = tas585_weighted.mean(('lon', 'lat'))"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "fig, ax = plt.subplots(figsize=(8,4))\n",
+    "\n",
+    "ax.plot(x, tas126_weightedmean, label='ssp126 weighted mean')\n",
+    "ax.plot(x, tas245_weightedmean, label='ssp245 weighted mean')\n",
+    "ax.plot(x, tas370_weightedmean, label='ssp370 weighted mean')\n",
+    "ax.plot(x, tas585_weightedmean, label='ssp585 weighted mean')\n",
+    "\n",
+    "plt.grid()\n",
+    "\n",
+    "ax.set_title('CMIP6:  tas global yearly mean')\n",
+    "\n",
+    "ax.set_xlabel('years')\n",
+    "ax.set_ylabel(data126.long_name+' ['+data126.units+']')\n",
+    "\n",
+    "legend = ax.legend(loc='upper left', shadow=True, fontsize='small')\n",
+    "\n",
+    "legend.get_frame().set_facecolor('lightgray')"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "Python 3 unstable (using the module python3/unstable)",
+   "language": "python",
+   "name": "python3_unstable"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.8.6"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 4
+}
+%% Cell type:markdown id: tags:
+
+# Exercise: CMIP6 - Plot Time Series of annual means
+
+<br>
+
+In this task, the global annual means of the CMIP6 variable `tas` are to be calculated and plotted.
+
+Since one time series alone would be quite boring we want to generate a total of 4 time series of <br>
+the scenarios
+
+    SSP 1-2.6
+    SSP 2-4.5
+    SSP 3-7.0
+    SSP 5-8.5
+
+For more information about the scenarios see
+
+https://www.dkrz.de/en/communication/climate-simulations/cmip6-en/the-ssp-scenarios
+
+The data are stored in several separate files and are to be merged with CDO and then their field mean average calculated.
+
+### In the exercise you will:
+
+1. Learn to use the netCDF input files
+1. Process the data with CDO or xarray
+1. Visualize the processed data
+
+An example plot of the scenarios of some more ensembles and their mean values can be found on the DKRZ <br>
+page for CMIP6
+
+https://www.dkrz.de/de/kommunikation/klimasimulationen/cmip6-de/cmip6-aktivitaeten-am-dkrz-ueberblick
+
+### Input Data
+
+**Historical**:
+	/pool/data/CMIP6/data/CMIP/MPI-M/MPI-ESM1-2-LR/historical/r1i1p1f1/Amon/tas/gn/v20190710/
+
+**Scenario SSP 1-2.6**:
+    /pool/data/CMIP6/data/ScenarioMIP/MPI-M/MPI-ESM1-2-LR/ssp126/r1i1p1f1/Amon/tas/gn/v20190710/
+
+**Scenario SSP 2-4.5**:
+    /pool/data/CMIP6/data/ScenarioMIP/MPI-M/MPI-ESM1-2-LR/ssp245/r1i1p1f1/Amon/tas/gn/v20190710/
+
+**Scenario SSP 3-7.0**:
+    /pool/data/CMIP6/data/ScenarioMIP/MPI-M/MPI-ESM1-2-LR/ssp370/r1i1p1f1/Amon/tas/gn/v20190710/
+
+**Scenario SSP 5-8.5**:
+    /pool/data/CMIP6/data/ScenarioMIP/MPI-M/MPI-ESM1-2-LR/ssp585/r1i1p1f1/Amon/tas/gn/v20190710/
+
+### CDO hints:
+
+1. Merge files along time (mergetime)
+1. Compute the annual mean (yearmean)
+1. Compute the field means (fldmean)
+1. Take care of the shapes
+1. Important: clean temporary files
+
+### Xarray hints:
+
+1. Open multiple files at once
+1. Compute the weights (http://xarray.pydata.org/en/stable/examples/area_weighted_temperature.html)
+1. Compute the annual mean (resample)
+1. Compute the field means (mean)
+
+%% Cell type:markdown id: tags:
+
+## Import CDO module
+
+%% Cell type:code id: tags:
+
+``` python
+from cdo import *
+
+cdo = Cdo()
+```
+
+%% Cell type:markdown id: tags:
+
+## Import Matplotlib module
+
+%% Cell type:code id: tags:
+
+``` python
+from matplotlib import pyplot as plt
+
+%matplotlib inline
+```
+
+%% Cell type:markdown id: tags:
+
+## Input files
+
+%% Cell type:code id: tags:
+
+``` python
+data_ssp126 = '/pool/data/CMIP6/data/ScenarioMIP/MPI-M/MPI-ESM1-2-LR/ssp126/r1i1p1f1/Amon/tas/gn/v20190710/tas_*.nc'
+data_ssp245 = '/pool/data/CMIP6/data/ScenarioMIP/MPI-M/MPI-ESM1-2-LR/ssp245/r1i1p1f1/Amon/tas/gn/v20190710/tas_*.nc'
+data_ssp370 = '/pool/data/CMIP6/data/ScenarioMIP/MPI-M/MPI-ESM1-2-LR/ssp370/r1i1p1f1/Amon/tas/gn/v20190710/tas_*.nc'
+data_ssp585 = '/pool/data/CMIP6/data/ScenarioMIP/MPI-M/MPI-ESM1-2-LR/ssp585/r1i1p1f1/Amon/tas/gn/v20190710/tas_*.nc'
+```
+
+%% Cell type:markdown id: tags:
+
+## Compute the field mean of the yearly mean
+
+%% Cell type:code id: tags:
+
+``` python
+tas126_ymean = cdo.fldmean(input='-yearmean -mergetime ' + data_ssp126, returnXArray='tas')
+tas245_ymean = cdo.fldmean(input='-yearmean -mergetime ' + data_ssp245, returnXArray='tas')
+tas370_ymean = cdo.fldmean(input='-yearmean -mergetime ' + data_ssp370, returnXArray='tas')
+tas585_ymean = cdo.fldmean(input='-yearmean -mergetime ' + data_ssp585, returnXArray='tas')
+```
+
+%% Cell type:markdown id: tags:
+
+## Keep an eye on the data shapes
+
+```
+print(tas126_ymean.shape)
+print(tas245_ymean.shape)
+print(tas370_ymean.shape)
+print(tas585_ymean.shape)
+
+(86, 1, 1)
+(86, 1, 1)
+(86, 1, 1)
+(86, 1, 1)
+```
+Remove the lat and lon dimensions
+
+```
+data126 = tas126_ymean[:,0,0]
+data245 = tas245_ymean[:,0,0]
+data370 = tas370_ymean[:,0,0]
+data585 = tas585_ymean[:,0,0]
+```
+
+%% Cell type:markdown id: tags:
+
+To prevent this step use CDO's option `--reduce_dims` within the CDO command line.
+
+For example
+
+`tas126_ymean = cdo.fldmean(options='--reduce_dim', input='-yearmean -mergetime ' + data_ssp126, returnXArray='tas')`
+
+%% Cell type:code id: tags:
+
+``` python
+data126 = cdo.fldmean(options='--reduce_dim', input='-yearmean -mergetime ' + data_ssp126, returnXArray='tas')
+data245 = cdo.fldmean(options='--reduce_dim', input='-yearmean -mergetime ' + data_ssp245, returnXArray='tas')
+data370 = cdo.fldmean(options='--reduce_dim', input='-yearmean -mergetime ' + data_ssp370, returnXArray='tas')
+data585 = cdo.fldmean(options='--reduce_dim', input='-yearmean -mergetime ' + data_ssp585, returnXArray='tas')
+```
+
+%% Cell type:markdown id: tags:
+
+## Define x-axis
+
+%% Cell type:code id: tags:
+
+``` python
+x = range(2015,2101,1)
+```
+
+%% Cell type:markdown id: tags:
+
+## Create the time series plot
+
+
+%% Cell type:code id: tags:
+
+``` python
+fig, ax = plt.subplots(figsize=(8,4))
+
+ax.plot(x, data126, label='ssp126')
+ax.plot(x, data245, label='ssp245')
+ax.plot(x, data370, label='ssp370')
+ax.plot(x, data585, label='ssp585')
+
+plt.grid()
+
+ax.set_title('CMIP6:  tas global yearly mean')
+
+ax.set_xlabel('years')
+ax.set_ylabel(data126.long_name+' ['+data126.units+']')
+
+legend = ax.legend(loc='upper left', shadow=True, fontsize='small')
+
+legend.get_frame().set_facecolor('lightgray')
+```
+
+%% Cell type:markdown id: tags:
+
+## Delete temporary files
+
+%% Cell type:code id: tags:
+
+``` python
+cdo.cleanTempDir()
+```
+
+%% Cell type:markdown id: tags:
+
+## Xarray not CDO
+
+%% Cell type:code id: tags:
+
+``` python
+import xarray as xr
+import numpy as np
+```
+
+%% Cell type:code id: tags:
+
+``` python
+ds126 = xr.open_mfdataset(data_ssp126)
+ds245 = xr.open_mfdataset(data_ssp245)
+ds370 = xr.open_mfdataset(data_ssp370)
+ds585 = xr.open_mfdataset(data_ssp585)
+```
+
+%% Cell type:markdown id: tags:
+
+### Yearly means
+
+%% Cell type:code id: tags:
+
+``` python
+tas126 = ds126.tas.resample(time="Y").mean()
+tas245 = ds245.tas.resample(time="Y").mean()
+tas370 = ds370.tas.resample(time="Y").mean()
+tas585 = ds585.tas.resample(time="Y").mean()
+```
+
+%% Cell type:code id: tags:
+
+``` python
+weights = np.cos(np.deg2rad(ds126.lat))
+weights.name = "weights"
+```
+
+%% Cell type:code id: tags:
+
+``` python
+tas126_weighted = tas126.weighted(weights)
+tas245_weighted = tas245.weighted(weights)
+tas370_weighted = tas370.weighted(weights)
+tas585_weighted = tas585.weighted(weights)
+```
+
+%% Cell type:code id: tags:
+
+``` python
+tas126_weightedmean = tas126_weighted.mean(('lon', 'lat'))
+tas245_weightedmean = tas245_weighted.mean(('lon', 'lat'))
+tas370_weightedmean = tas370_weighted.mean(('lon', 'lat'))
+tas585_weightedmean = tas585_weighted.mean(('lon', 'lat'))
+```
+
+%% Cell type:code id: tags:
+
+``` python
+fig, ax = plt.subplots(figsize=(8,4))
+
+ax.plot(x, tas126_weightedmean, label='ssp126 weighted mean')
+ax.plot(x, tas245_weightedmean, label='ssp245 weighted mean')
+ax.plot(x, tas370_weightedmean, label='ssp370 weighted mean')
+ax.plot(x, tas585_weightedmean, label='ssp585 weighted mean')
+
+plt.grid()
+
+ax.set_title('CMIP6:  tas global yearly mean')
+
+ax.set_xlabel('years')
+ax.set_ylabel(data126.long_name+' ['+data126.units+']')
+
+legend = ax.legend(loc='upper left', shadow=True, fontsize='small')
+
+legend.get_frame().set_facecolor('lightgray')
+```
+
+%% Cell type:code id: tags:
+
+``` python
+```
--- a/exercises/data_processing.md
+++ b/exercises/data_processing.md
+# Exercise: CMIP6 - Plot Time Series of annual means
+
+<br>
+
+In this task, the global annual means of the CMIP6 variable `tas` are to be calculated and plotted.
+
+Since one time series alone would be quite boring we want to generate a total of 4 time series of <br>
+the scenarios 
+
+    SSP 1-2.6
+    SSP 2-4.5
+    SSP 3-7.0
+    SSP 5-8.5
+
+For more information about the scenarios see
+
+https://www.dkrz.de/en/communication/climate-simulations/cmip6-en/the-ssp-scenarios
+
+The data are stored in several separate files and are to be merged with CDO and then their field mean average calculated. 
+
+### In the exercise you will:
+
+1. Learn to use the netCDF input files
+1. Process the data with CDO or xarray
+1. Visualize the processed data
+
+An example plot of the scenarios of some more ensembles and their mean values can be found on the DKRZ <br>
+page for CMIP6
+
+https://www.dkrz.de/de/kommunikation/klimasimulationen/cmip6-de/cmip6-aktivitaeten-am-dkrz-ueberblick
+
+### Input Data
+
+**Historical**:
+	/pool/data/CMIP6/data/CMIP/MPI-M/MPI-ESM1-2-LR/historical/r1i1p1f1/Amon/tas/gn/v20190710/
+
+**Scenario SSP 1-2.6**:
+    /pool/data/CMIP6/data/ScenarioMIP/MPI-M/MPI-ESM1-2-LR/ssp126/r1i1p1f1/Amon/tas/gn/v20190710/
+ 
+**Scenario SSP 2-4.5**:
+    /pool/data/CMIP6/data/ScenarioMIP/MPI-M/MPI-ESM1-2-LR/ssp245/r1i1p1f1/Amon/tas/gn/v20190710/
+
+**Scenario SSP 3-7.0**:
+    /pool/data/CMIP6/data/ScenarioMIP/MPI-M/MPI-ESM1-2-LR/ssp370/r1i1p1f1/Amon/tas/gn/v20190710/
+
+**Scenario SSP 5-8.5**:
+    /pool/data/CMIP6/data/ScenarioMIP/MPI-M/MPI-ESM1-2-LR/ssp585/r1i1p1f1/Amon/tas/gn/v20190710/
+
+### CDO hints:
+
+1. Merge files along time (mergetime)
+1. Compute the annual mean (yearmean)
+1. Compute the field means (fldmean)
+1. Take care of the shapes
+1. Important: clean temporary files
+
+### Xarray hints:
+
+1. Open multiple files at once
+1. Compute the weights (http://xarray.pydata.org/en/stable/examples/area_weighted_temperature.html)
+1. Compute the annual mean (resample)
+1. Compute the field means (mean)