diff --git a/Jupyter_Notebooks/get_metrics_joint_dom.ipynb b/Jupyter_Notebooks/get_metrics_joint_dom.ipynb
index f09b16593481688838a04bf6c98c076b9c9e65c8..60be31ea28d06ecabe78f4f909bddd29f98edfd1 100755
--- a/Jupyter_Notebooks/get_metrics_joint_dom.ipynb
+++ b/Jupyter_Notebooks/get_metrics_joint_dom.ipynb
@@ -24,9 +24,12 @@
    "source": [
     "# Evaluation over a smaller (joint) domain\n",
     "\n",
-    "In this Jupyter Notebook, evaluation will be performed on a subregion which is nested into the target region of the evaluated trained video prediction models. The resulting netCDF-files can be used in the meta-postpro\n",
-    "The following cells will first merge all forecast files under `indir` into a single netCDF-file.<br>\n",
-    "Then the data is sliced to the domain defined by `lonlatbox` and all subsequent evaluation is performed on this smaller domain.<br>\n",
+    "In this Jupyter Notebook, evaluation will be performed on a subregion which is nested into the target region of the evaluated trained video prediction models. The resulting netCDF-files can be used in the meta-postprocessing to perform an evaluation on a joint region such as in Figure 10a of the [manuscript](https://doi.org/10.5194/gmd-2021-430). <br>\n",
+    "\n",
+    "### Approach\n",
+    "\n",
+    "In the following cells, all forecast files under `indir` will be merged into a single netCDF-file first.<br>\n",
+    "Then the data gets sliced to the domain defined by `lonlatbox` and all subsequent evaluation is performed on this smaller domain.<br>\n",
     "The evaluation metrics are then saved to a file under `indir` named `evaluation_metrics_<nlon>x<nlat>.nc` where `nlat` and `nlon` denote the number of grid points/pixels in latitude and longitude direction of the smaller domain, respectively. <br>\n",
     "\n",
     "Thus, first let's define the basic parameters:"
@@ -39,9 +42,9 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "indir = \"/p/project/deepacf/deeprain/video_prediction_shared_folder/results/era5-Y2007-2019M01to12-92x56-3840N0000E-2t_tcc_t_850/savp/20210901T090059_gong1_savp_cv12/\"\n",
+    "indir = \"<path_to_trained_model/with_larger_target_domain>\"\n",
     "model = \"savp\"\n",
-    "# define domain. [3., 24.3, 40.2, 53.1] corresponds to the smallest domain tested in the GMD paper\n",
+    "# define domain. [3., 24.3, 40.2, 53.1] corresponds to the smallest domain tested in the GMD paper (with 72x44grid points)\n",
     "lonlatbox = [3., 24.3, 40.2, 53.1]"
    ]
   },
@@ -297,9 +300,9 @@
  ],
  "metadata": {
   "kernelspec": {
-   "display_name": "PyDeepLearning-1.0",
+   "display_name": "Python 3",
    "language": "python",
-   "name": "pydeeplearning"
+   "name": "python3"
   },
   "language_info": {
    "codemirror_mode": {
@@ -311,7 +314,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.8.5"
+   "version": "3.9.6"
   }
  },
  "nbformat": 4,