diff --git a/examples/nblibrary/rof/global_discharge_gauge_compare_obs.ipynb b/examples/nblibrary/rof/global_discharge_gauge_compare_obs.ipynb
index 1bdcc54..6c8d951 100644
--- a/examples/nblibrary/rof/global_discharge_gauge_compare_obs.ipynb
+++ b/examples/nblibrary/rof/global_discharge_gauge_compare_obs.ipynb
@@ -10,38 +10,38 @@
     "tags": []
    },
    "source": [
-    "## NB1. ROF global monthly, annual, seasonal flows analysis <a id='top'></a>\n",
+    "## ROF global monthly, annual, seasonal flows analysis <a id='top'></a>\n",
     "\n",
-    "Use \n",
+    "Use the following datasets\n",
     "\n",
     "1. reach-D19 gauge link ascii\n",
     "2. D19 flow site geopackage\n",
     "3. D19 discharge netCDF\n",
-    "4. monthly and yearly flow netCD (history file)\n",
+    "4. history netCD including river discharge\n",
     "\n",
     "[1. Setupt](#setup)\n",
     "\n",
     "[2. Loading data](#load_data)\n",
     "\n",
-    "- monthly history files (directory from CESM or postprocessed) from archive. \n",
-    "\n",
-    "- Reference data is monthly discharge estimates at 922 big river mouths from Dai et al. 2019 data (D19)\n",
+    "- Read monthly history files from archive. \n",
+    "- Reference data: monthly discharge estimates at 922 big river mouths from Dai et al. 2019 data (D19)\n",
     "\n",
     "[3. Large 24 river analysis](#24_large_rivers)\n",
     "\n",
-    "- Plotting time seriese (annual, seasonal cycle) and scatter plots at Large 24 selected rivers with D19 referece data\n",
+    "- Plotting time seriese (annual, seasonal cycle).\n",
+    "- if D19 referece data is available, scatter plots at Large 24 selected rivers against D19 referece data\n",
     "\n",
     "[4. Large 50 rivers analysis](#50_large_rivers)\n",
     "\n",
-    "- Annual flow summary table at large 50 selected rivers. \n",
+    "- Annual flow summary table at large 50 selected rivers.\n",
+    "- if D19 referece data is available, Scatter plot of annual flow against D19 reference data.\n",
     "\n",
     "[5. 922 rivers analysis](#922_rivers)\n",
     "\n",
-    "- summary statistics (%bias, rmse, correlation) at all 922 river sites - map and boxplots.\n",
-    "\n",
-    "[6. basin area vs annual flow](#flow_diagnoisis)\n",
-    "\n",
-    "Moved this to another notebook"
+    "- run only if reference flow is available\n",
+    "- error statistics (%bias, rmse, correlation) at all 922 river sites.\n",
+    "- plot error statistic on the global map\n",
+    "- plot boxplots including case(s) for each error metric\n"
    ]
   },
   {
@@ -73,6 +73,7 @@
     "from dask.distributed import Client, LocalCluster\n",
     "\n",
     "import scripts.colors as colors\n",
+    "import scripts.metrics as metric\n",
     "from scripts.utility import load_yaml\n",
     "from scripts.utility import no_time_variable\n",
     "\n",
@@ -194,7 +195,7 @@
    "source": [
     "# Spin up cluster (if running in parallel)\n",
     "client = None\n",
-    "if serial:\n",
+    "if not serial:\n",
     "    cluster = LocalCluster(**lc_kwargs)\n",
     "    client = Client(cluster)\n",
     "\n",
@@ -213,9 +214,9 @@
    "metadata": {},
    "source": [
     "### 2.1. Monthly/annual flow netCDFs\n",
-    "- month_data (xr dataset)\n",
-    "- year_data (xr dataset)\n",
-    "- seas_data (xr dataset)"
+    "- month_data (xr dataset) - read from archive\n",
+    "- year_data (xr dataset) - computed from monthly\n",
+    "- seas_data (xr dataset) - computed from monthly"
    ]
   },
   {
@@ -386,13 +387,6 @@
     "dr_q_obs_seasonal = ds_q_obs_mon.groupby(\"time.month\").mean(\"time\")"
    ]
   },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "metadata": {},
-   "outputs": [],
-   "source": []
-  },
   {
    "cell_type": "markdown",
    "metadata": {},
@@ -971,14 +965,14 @@
    "metadata": {},
    "source": [
     "------\n",
-    "## 3. Anaysis for all 922 sites <a id='922_rivers'></a>"
+    "## 5. Anaysis for all 922 sites <a id='922_rivers'></a>"
    ]
   },
   {
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "### 3.1 Compute metris at all the sites (no plots nor tables)"
+    "### 5.1 Compute metris at all the sites (no plots nor tables)"
    ]
   },
   {
@@ -1042,9 +1036,9 @@
     "                    )\n",
     "\n",
     "                # compute %bias, correlation, RMSE\n",
-    "                bias[ix] = np.nanmean(q_sim - q_obs) / np.nanmean(q_obs) * 100.0\n",
+    "                bias[ix] = metric.pbias(q_sim, q_obs)\n",
     "                corr[ix] = np.corrcoef(q_sim, q_obs)[0, 1]\n",
-    "                rmse[ix] = np.sqrt(np.mean((q_sim - q_obs) ** 2))\n",
+    "                rmse[ix] = metric.rmse(qsim, qobs)\n",
     "\n",
     "        mon_pbias[case] = bias\n",
     "        mon_corr[case] = corr\n",
@@ -1059,7 +1053,7 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "### 4.2. Spatial metric map "
+    "### 5.2. Spatial metric map "
    ]
   },
   {
@@ -1093,8 +1087,13 @@
     "    }\n",
     "\n",
     "    meta = {\n",
-    "        \"bias\": {\"name\": \"%bias\", \"vmin\": -100, \"vmax\": 100, \"cm\": colors.cmap11},\n",
-    "        \"corr\": {\"name\": \"correlation\", \"vmin\": 0.2, \"vmax\": 1, \"cm\": colors.cmap12},\n",
+    "        \"bias\": {\n",
+    "            \"name\": \"%bias\",\n",
+    "            \"vmin\": -100,\n",
+    "            \"vmax\": 100,\n",
+    "            \"cm\": colors.cmap_RedGrayBlue,\n",
+    "        },\n",
+    "        \"corr\": {\"name\": \"correlation\", \"vmin\": 0.2, \"vmax\": 1, \"cm\": colors.cmap_corr},\n",
     "        \"rmse\": {\"name\": \"RMSE\", \"vmin\": 0, \"vmax\": 1000, \"cm\": mpl.cm.turbo},\n",
     "    }\n",
     "\n",
@@ -1133,7 +1132,7 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "### 4.3 Boxplots of Error metrics (RMSE, %bias, and correlation coefficient)\n",
+    "### 5.4 Boxplots of Error metrics (RMSE, %bias, and correlation coefficient)\n",
     "Boxplot distribution is based on metrics sampled at 922 sites.\n",
     "\n",
     "The box extends from the Q1 to Q3 quartile values of the data, with a line at the median (Q2). \n",
diff --git a/examples/nblibrary/rof/global_discharge_ocean_compare_obs.ipynb b/examples/nblibrary/rof/global_discharge_ocean_compare_obs.ipynb
index d08065e..8b24be6 100644
--- a/examples/nblibrary/rof/global_discharge_ocean_compare_obs.ipynb
+++ b/examples/nblibrary/rof/global_discharge_ocean_compare_obs.ipynb
@@ -10,9 +10,9 @@
     "tags": []
    },
    "source": [
-    "## NB2. ROF monthly, annual, seasonal discharge at ocean outlets <a id='top'></a>\n",
+    "## ROF monthly, annual, seasonal discharge at ocean outlets <a id='top'></a>\n",
     "\n",
-    "Use \n",
+    "Use the following datasets\n",
     "\n",
     "1. reach-D19 gauge link ascii\n",
     "2. D19 flow site geopackage\n",
@@ -21,20 +21,20 @@
     "\n",
     "[1. Setupt](#setup)\n",
     "\n",
-    "[2. Loading data](#load_data)\n",
     "\n",
-    "- monthly history files (directory from CESM or postprocessed) from archive. \n",
+    "[2. Loading discharge data](#load_discharge_data)\n",
     "\n",
-    "- Reference data is monthly discharge estimates at 922 big river mouths from Dai et al. 2019 data (D19)\n",
+    "- Read monthly history files from archive. \n",
+    "- Reference data: monthly discharge estimates at 922 big river mouths from Dai et al. 2019 data (D19)\n",
     "\n",
-    "[3. Read river, catchment, gauge information](#read_meta)\n",
+    "[3. Read river, catchment, gauge information](#read_ancillary)\n",
     "\n",
     "- catchment polygon (geopackage)\n",
     "- gauge point (geopackage)\n",
     "- gauge-catchment link (csv)\n",
     "- outlet reach information (netCDF) including discharging ocean names\n",
     "\n",
-    "[4. Ocean discharge line plots](#922_rivers)\n",
+    "[4. Ocean discharge line plots](#24_large_rivers)\n",
     "\n",
     "- total seasonal flow for oceans. \n"
    ]
@@ -197,7 +197,7 @@
    "source": [
     "# Spin up cluster (if running in parallel)\n",
     "client = None\n",
-    "if serial:\n",
+    "if not serial:\n",
     "    cluster = LocalCluster(**lc_kwargs)\n",
     "    client = Client(cluster)\n",
     "\n",
@@ -208,7 +208,7 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "## 2. Loading data <a id='load_data'></a>"
+    "## 2. Loading discharge data <a id='load_discharge_data'></a>"
    ]
   },
   {
@@ -331,7 +331,7 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "## 3. Reading river, catchment, gauge infomation  <a id='read_meta'></a>\n",
+    "## 3. Reading river, catchment, gauge infomation  <a id='read_ancillary'></a>\n",
     "\n",
     "- catchment polygon (geopackage)\n",
     "- gauge point (geopackage)\n",
@@ -484,7 +484,7 @@
    },
    "source": [
     "------\n",
-    "## 4. plot annual cycle for global oceans <a id='24_large_rivers'></a>\n",
+    "## 3. plot annual cycle for global oceans <a id='24_large_rivers'></a>\n",
     "\n",
     "TODO: Referece flow plot should be independent from cases (network). Currently the last case plotted looks better matched with reference flow. "
    ]