diff --git a/01-introduction-geospatial-data.ipynb b/01-introduction-geospatial-data.ipynb
index 5c00070..cbcaa5c 100644
--- a/01-introduction-geospatial-data.ipynb
+++ b/01-introduction-geospatial-data.ipynb
@@ -182,6 +182,40 @@
     "africa.plot()"
    ]
   },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "---\n",
+    "\n",
+    "**Exercise**: create a plot of South America\n"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "<!--\n",
+    "countries[countries['continent'] == 'South America'].plot()\n",
+    "-->"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "---"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "countries.head()"
+   ]
+  },
   {
    "cell_type": "markdown",
    "metadata": {},
@@ -486,6 +520,31 @@
     "Note the different scale of x and y."
    ]
   },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "---\n",
+    "\n",
+    "**Exercise**: project the countries to [Web Mercator](http://epsg.io/3857), the CRS used by Google Maps, OpenStreetMap and most web providers.\n"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "<!--\n",
+    "countries.to_crs(epsg=3857)\n",
+    "-->"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "---"
+   ]
+  },
   {
    "cell_type": "markdown",
    "metadata": {},
@@ -512,6 +571,34 @@
     "See the [04-more-on-visualization.ipynb](04-more-on-visualization.ipynb) notebook for more details on visualizing geospatial datasets."
    ]
   },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "---\n",
+    "\n",
+    "**Exercise**: replicate the figure above by coloring the countries in black and cities in yellow\n"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "<!--\n",
+    "ax = countries.plot(edgecolor='w', facecolor='k', figsize=(15, 10))\n",
+    "rivers.plot(ax=ax)\n",
+    "cities.plot(ax=ax, color='yellow')\n",
+    "ax.set(xlim=(-20, 60), ylim=(-40, 40))\n",
+    "-->"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "---"
+   ]
+  },
   {
    "cell_type": "markdown",
    "metadata": {},
@@ -629,6 +716,37 @@
    "source": [
     "See http://geopandas.readthedocs.io/en/latest/gallery/create_geopandas_from_pandas.html#sphx-glr-gallery-create-geopandas-from-pandas-py for full example"
    ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "---\n",
+    "\n",
+    "**Exercise**: use [geojson.io](http://geojson.io) to mark five points, and create a `GeoDataFrame` with it. Note that coordinates will be expressed in longitude and latitude, so you'll have to set the CRS accordingly.\n"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "<!--\n",
+    "df = pd.DataFrame(\n",
+    "    {'Name': ['Hotel', 'Capitol', 'Barton Springs'],\n",
+    "     'Latitude': [30.28195889019179, 30.274782936992608, 30.263728440902543],\n",
+    "     'Longitude': [-97.74006128311157, -97.74038314819336, -97.77013421058655]})\n",
+    "df['Coordinates']  = list(zip(df.Longitude, df.Latitude))\n",
+    "df['Coordinates'] = df['Coordinates'].apply(Point)\n",
+    "gdf = geopandas.GeoDataFrame(df, geometry='Coordinates', crs={'init': 'epsg:4326'})\n",
+    "-->"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "---"
+   ]
   }
  ],
  "metadata": {
@@ -647,7 +765,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.5.5"
+   "version": "3.6.5"
   }
  },
  "nbformat": 4,
diff --git a/02-spatial-relationships-operations.ipynb b/02-spatial-relationships-operations.ipynb
index c413dcc..945093a 100644
--- a/02-spatial-relationships-operations.ipynb
+++ b/02-spatial-relationships-operations.ipynb
@@ -125,8 +125,41 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "You can recognize the abstract shape of Belgium.\n",
+    "You can recognize the abstract shape of Belgium."
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "---\n",
     "\n",
+    "**Exercise**: produce a similar figure as above that includes France, Spain, Paris, Madrid, and a line that connects both capitals"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "<!--\n",
+    "sp = countries.loc[countries['name'] == 'Spain', 'geometry'].squeeze()\n",
+    "mad = cities.loc[cities['name'] == 'Madrid', 'geometry'].squeeze()\n",
+    "fr = countries.loc[countries['name'] == 'France', 'geometry'].squeeze()\n",
+    "geopandas.GeoSeries([sp, fr, mad, paris, LineString([paris, mad])]).plot(cmap='tab10')\n",
+    "-->"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "---"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
     "Brussels, the capital of Belgium, is thus located within Belgium. This is a spatial relationship, and we can test this using the individual shapely geometry objects as follow:"
    ]
   },
@@ -302,6 +335,32 @@
     "</div>"
    ]
   },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "---\n",
+    "\n",
+    "**Exercise**: find all of the countries contiguous to Brazil"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "<!--\n",
+    "br = countries.query('name == \"Brazil\"')['geometry'].squeeze()\n",
+    "countries[countries.touches(br)]  # or .intersects\n",
+    "-->"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "---"
+   ]
+  },
   {
    "cell_type": "markdown",
    "metadata": {},
@@ -326,7 +385,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "geopandas.GeoSeries([belgium, brussels.buffer(1)]).plot(alpha=0.5, cmap='tab10')"
+    "geopandas.GeoSeries([belgium, brussels.buffer(1), brussels]).plot(alpha=0.5, cmap='tab10')"
    ]
   },
   {
@@ -399,6 +458,52 @@
     "africa"
    ]
   },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "---\n",
+    "\n",
+    "**Exercise**: Create a single polygon for Europe"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "<!--\n",
+    "eu = countries.query('continent == \"Africa\"')\n",
+    "eu.unary_union\n",
+    "-->"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "\n",
+    "**Exercise**: Create a single polygon with all of the countries that you would have to cross to travel directly (as the crow flies) from Beijing to Sydney"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "<!--\n",
+    "bei = cities.query('name == \"Beijing\"')['geometry'].squeeze()\n",
+    "syd = cities.query('name == \"Sydney\"')['geometry'].squeeze()\n",
+    "\n",
+    "countries[countries.crosses(LineString([bei, syd]))].unary_union\n",
+    "-->"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "---"
+   ]
+  },
   {
    "cell_type": "code",
    "execution_count": null,
@@ -445,7 +550,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.5.5"
+   "version": "3.6.5"
   }
  },
  "nbformat": 4,
diff --git a/03-spatial-joins.ipynb b/03-spatial-joins.ipynb
index 12fb751..66f6f37 100644
--- a/03-spatial-joins.ipynb
+++ b/03-spatial-joins.ipynb
@@ -239,6 +239,92 @@
     "joined['continent'].value_counts()"
    ]
   },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "---\n",
+    "\n",
+    "**Exercise**: Which rivers pass within 1 degree of Paris? And for Hong Kong?\n",
+    "\n",
+    "Tip: `rivers` contains empty geometries, make sure to get rid of them before anything else\n",
+    "\n"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "<!--\n",
+    "paris = cities.query('name == \"Paris\"')\\\n",
+    "              ['geometry']\\\n",
+    "              .squeeze()\\\n",
+    "              .buffer(1)\n",
+    "\n",
+    "river_lines = rivers[rivers.geometry.notna()]\n",
+    "\n",
+    "river_lines[river_lines.crosses(paris)]\n",
+    "\n",
+    "hk = cities.query('name == \"Hong Kong\"')\\\n",
+    "              ['geometry']\\\n",
+    "              .squeeze()\\\n",
+    "              .buffer(1)\n",
+    "\n",
+    "river_lines = rivers[rivers.geometry.notna()]\n",
+    "\n",
+    "river_lines[river_lines.crosses(hk)]\n",
+    "-->"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "---\n",
+    "\n",
+    "**Exercise** [PRO]: Find all European cities within 1 degre of at least one river\n",
+    "\n",
+    "\n",
+    "This requires:"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "1. Obtaining a list of European cities\n",
+    "1. Creating a `GeoDataFrame` with all river lines\n",
+    "1. For every European city, checking whether at least one river \"crosses\" the area within 1 degree\n",
+    "1. If so, saving the record of a city"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "<!--\n",
+    "eu_cities = joined.query('continent == \"Europe\"')\n",
+    "river_lines = rivers[rivers.geometry.notna()]\n",
+    "\n",
+    "answer = []\n",
+    "for id, city in eu_cities.iterrows():\n",
+    "    buf = city['geometry'].buffer(1)\n",
+    "    rivers_crossing = river_lines.crosses(buf).sum()\n",
+    "    if rivers_crossing > 0:\n",
+    "        answer.append(city)\n",
+    "answer = geopandas.GeoDataFrame(answer, \n",
+    "                                crs=eu_cities.crs)\n",
+    "answer\n",
+    "-->"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "---"
+   ]
+  },
   {
    "cell_type": "markdown",
    "metadata": {},
@@ -300,6 +386,45 @@
     "\n",
     "</div>"
    ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "---\n",
+    "\n",
+    "**Exercise**: if we define rural as those areas that are not within 50Km of the centre of a city, what is the extention of rural North America? What is the percentage that it covers?\n",
+    "\n",
+    "Tip: make sure your projection allows you to calculate distances in metres ([wink wink](http://epsg.io/3083))"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "<!--\n",
+    "namerica = countries.query('continent == \"North America\"')\\\n",
+    "                    .to_crs(epsg=3083)\n",
+    "urban = geopandas.GeoDataFrame(\n",
+    "            {'geometry':cities.to_crs(epsg=3083)\\\n",
+    "                              .buffer(50000)}, \\\n",
+    "            crs = namerica.crs)\n",
+    "        \n",
+    "rural = geopandas.overlay(namerica, urban, how='difference')\n",
+    "\n",
+    "print('Rural: ', rural.unary_union.area * 1e-6, ' Km^2')\n",
+    "print('Rural is %.2f%% of the total extent'%\n",
+    "      (rural.unary_union.area * 100 / \n",
+    "       namerica.unary_union.area))\n",
+    "-->"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "---"
+   ]
   }
  ],
  "metadata": {
@@ -318,7 +443,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.5.5"
+   "version": "3.6.5"
   }
  },
  "nbformat": 4,
diff --git a/04-more-on-visualization.ipynb b/04-more-on-visualization.ipynb
index 563c2df..bc34a29 100644
--- a/04-more-on-visualization.ipynb
+++ b/04-more-on-visualization.ipynb
@@ -90,6 +90,32 @@
     "ax.set_axis_off()"
    ]
   },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "---\n",
+    "\n",
+    "**Exercise**: create a 12 by 12 plot of the world rivers without axes"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "<!--\n",
+    "ax = rivers.plot(figsize=(12, 12))\n",
+    "ax.set_axis_off()\n",
+    "-->"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "---"
+   ]
+  },
   {
    "cell_type": "markdown",
    "metadata": {},
@@ -144,6 +170,36 @@
     "ax.set_axis_off()"
    ]
   },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "---\n",
+    "\n",
+    "**Exercise**: create an equal interval map of the length of world rivers"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "<!--\n",
+    "ax = rivers.assign(len=rivers.length).plot(figsize=(12, 12), \n",
+    "                                             column='len', \n",
+    "                                             scheme='equal_interval', \n",
+    "                                             legend=True, \n",
+    "                                             cmap='viridis_r')\n",
+    "ax.set_axis_off()\n",
+    "-->"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "---"
+   ]
+  },
   {
    "cell_type": "markdown",
    "metadata": {},
@@ -176,6 +232,39 @@
     "ax.set(xlim=(-20, 60), ylim=(-40, 40))"
    ]
   },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "#### `matplotlib`tic approach\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "f, ax = plt.subplots(1, figsize=(15, 10))\n",
+    "countries.plot(edgecolor='k', facecolor='none', ax=ax)\n",
+    "rivers.plot(ax=ax)\n",
+    "cities.plot(ax=ax, color='C1')\n",
+    "ax.axis('equal')\n",
+    "ax.set(xlim=(-20, 60), ylim=(-40, 40))\n",
+    "plt.show()"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "---\n",
+    "\n",
+    "**Exercise**: make a map of Europe with its rivers and cities\n",
+    "\n",
+    "---"
+   ]
+  },
   {
    "cell_type": "markdown",
    "metadata": {},
@@ -346,7 +435,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.5.5"
+   "version": "3.6.5"
   }
  },
  "nbformat": 4,
diff --git a/img/TopologicSpatialRelarions2.png b/img/TopologicSpatialRelarions2.png
new file mode 100644
index 0000000..bf0e631
Binary files /dev/null and b/img/TopologicSpatialRelarions2.png differ