{ "nbformat": 4, "nbformat_minor": 0, "metadata": { "colab": { "provenance": [] }, "kernelspec": { "name": "python3", "display_name": "Python 3" }, "language_info": { "name": "python" } }, "cells": [ { "cell_type": "markdown", "source": [ "# HeatSafeBern – Proof of Concept\n", "## Urban Heat Decision Support System für Outdoor-Sportler:innen in Bern\n", "\n", "Modul: BCS, Frühlingssemester 2026 \n", "Institution: Berner Fachhochschule (BFH) \n", "\n", "### Über dieses Notebook\n", "\n", "Dieses Jupyter Notebook demonstriert die technische Machbarkeit von HeatSafeBern,\n", "einem datenbasierten Entscheidungsunterstützungssystem\n", "für Outdoor-Aktivitäten bei Hitze im urbanen Raum Bern.\n", "\n", "Datenquelle: Smart Urban Heat Map Bern, Open Data REST API (CC-BY) \n", "Methodik: Humidex-Berechnung + relativer Comfort Score (0–100) \n", "Visualisierung: Interaktive Folium-Karte \n", "\n", "### Struktur\n", "1. Setup & Konfiguration\n", "2. Datenabfrage & Bereinigung\n", "3. Hitzebelastungsberechnung (Humidex + Comfort Score)\n", "4. Visualisierung (Karte, Routen, Tabelle)\n", "5. Historische Analyse zur Bestimmung der kühlsten Trainingszeit\n", "6. Zusammenfassung" ], "metadata": { "id": "HQh-kM7wEQoM" } }, { "cell_type": "markdown", "source": [ "# Annahmen-Limitationen\n", "\n", "Luftlinien statt echtes Routing\n", "\n", "Feste historische Stationsauswahl\n", "\n", "Comfort Score relativ pro Snapshot\n", "\n", "Keine personenbezogenen Empfehlungen / keine Gesundheitswarnung\n", "\n", "Kein GPS-Standortrouting, aufgrund Google-colab Einschränkung\n", "\n", "Nur ~60 Punktmessungen\n" ], "metadata": { "id": "e1yIZ0owM5Tg" } }, { "cell_type": "markdown", "source": [ "## 1. Setup & Konfiguration\n", "Installation der benötigten Libraries und Definition der globalen Konfigurationsparameter." ], "metadata": { "id": "woNBdMqOEUXc" } }, { "cell_type": "code", "source": [ "# ============================================================\n", "# HeatSafeBern PoC – Setup\n", "# Datenquelle: Smart Urban Heat Map Bern (CC-BY)\n", "# https://smart-urban-heat-map.ch\n", "# ============================================================\n", "\n", "# Libraries installieren (nur in Google Colab nötig)\n", "!pip install folium geopandas branca --quiet" ], "metadata": { "id": "k4a_l1GYEZoU" }, "execution_count": 30, "outputs": [] }, { "cell_type": "code", "source": [ "# Benötigten Libraries importieren\n", "import requests\n", "import pandas as pd\n", "import geopandas as gpd\n", "import folium\n", "import io\n", "import numpy as np\n", "from math import radians, cos, sin, asin, sqrt\n", "from datetime import datetime\n", "\n", "# Konfiguration RTD und historische-Daten\n", "API_BASE = 'https://smart-urban-heat-map.ch'\n", "URL_LATEST = f'{API_BASE}/api/v2/latest'\n", "URL_TIMESERIES = f'{API_BASE}/api/v2/timeseries'\n", "\n", "BERN_LAT = 46.9480 # Zentrum Bern\n", "BERN_LON = 7.4474\n", "GEO_RADIUS_KM = 10 # Maximaler Radius für Stationsfilter\n", "\n", "print(f'Setup abgeschlossen – {datetime.now().strftime(\"%d.%m.%Y %H:%M\")}')" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "GqL_NP0DEky8", "outputId": "b2a48e9c-12d6-407a-c1f0-ce24c96d5664" }, "execution_count": 31, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "Setup abgeschlossen – 13.03.2026 17:20\n" ] } ] }, { "cell_type": "markdown", "source": [ "## 2. Datenabfrage & Bereinigung\n", "### 2.1 Hilfsfunktionen\n", "Definition der Kernfunktionen für Distanzberechnung, Humidex und Comfort Score." ], "metadata": { "id": "GwipgxTDwaGw" } }, { "cell_type": "code", "source": [ "#Haversine-Funktion zur Distanzberechnung zwischen zwei GPS-Koordinaten.\n", "# Warum wird das benötigt?\n", "# Für die Routenlogik im PoC werden keine echten Strassenwege berechnet,\n", "# sondern vereinfachte Pseudo-Routen als Luftlinie zwischen Messstationen.\n", "# Die Funktion liefert die Distanz in Kilometern und dient damit als Grundlage,\n", "# um Stationen geografisch sinnvoll zu sortieren bzw. die ungefähre\n", "# Gesamtdistanz einer Route abzuschätzen.\n", "# Dadurch bleibt die Logik des PoC technisch einfach und nachvollziehbar,\n", "# ohne einen externen Routing-Service integrieren zu müssen.\n", "\n", "def haversine(lat1, lon1, lat2, lon2):\n", " \"\"\"Berechnet die Distanz in km zwischen zwei GPS-Koordinaten\"\"\"\n", " R = 6371\n", " lat1, lon1, lat2, lon2 = map(radians, [lat1, lon1, lat2, lon2])\n", " a = sin((lat2-lat1)/2)**2 + cos(lat1)*cos(lat2)*sin((lon2-lon1)/2)**2\n", " return 2 * R * asin(sqrt(a))\n", "\n", "#Berechnung des Humidex aus Temperatur und relativer Luftfeuchtigkeit.\n", "# Warum wird das benötigt?\n", "# Die Rohdaten der API enthalten nur Temperatur und Luftfeuchtigkeit.\n", "# Für die Bewertung der wahrgenommenen Hitzebelastung reicht Temperatur allein\n", "# jedoch nicht aus, da hohe Luftfeuchtigkeit die gefühlte Belastung verstärkt.\n", "# Der Humidex ist deshalb die zentrale absolute Kennzahl des PoC.\n", "# Er bildet die kombinierte Wirkung von Temperatur und Luftfeuchtigkeit ab\n", "# und dient als Grundlage für den relativen Comfort Score sowie für die\n", "# historische Analyse der typischerweise kühlsten Tageszeit.\n", "# Ungültige Werte (z.B. fehlende Daten oder physikalisch unsinnige Luftfeuchte)\n", "# werden bewusst als NaN behandelt, damit sie die weitere Berechnung nicht verzerren.\n", "\n", "def calculate_humidex(temp, rh):\n", " \"\"\"Berechnet den Humidex aus Temperatur (°C) und relativer Luftfeuchtigkeit (%) und gibt bei ungültigen Werten NaN zurück.\"\"\"\n", " if pd.isna(temp) or pd.isna(rh) or rh <= 0 or rh > 100:\n", " return np.nan\n", " try:\n", " gamma = np.log(rh/100) + (17.27*temp)/(237.7+temp)\n", " dewpoint = (237.7 * gamma) / (17.27 - gamma)\n", " e = 6.11 * np.exp(5417.7530 * (1/273.16 - 1/(273.15 + dewpoint)))\n", " return round(temp + 0.5555 * (e - 10), 1)\n", " except Exception:\n", " return np.nan\n", "\n", "# Relativer Comfort Score auf Basis der Humidex-Werte aller aktuellen Stationen.\n", "# Warum wird das benötigt?\n", "# Der Humidex ist ein absoluter Hitzewert, eignet sich aber im städtischen Raum\n", "# nur begrenzt für den direkten Vergleich zwischen nahe beieinanderliegenden\n", "# Stationen, da die Unterschiede oft klein sind.\n", "# Deshalb werden die Humidex-Werte innerhalb des aktuellen RTD-Snapshots\n", "# auf eine Skala von 0 bis 100 normalisiert:\n", "# 100 = kühlste Station, 0 = wärmste Station.\n", "# Dadurch wird sichtbar, welche Stationen im Vergleich zum restlichen Stadtgebiet\n", "# aktuell am angenehmsten bzw. am belastendsten sind.\n", "# Falls alle Stationen denselben Humidex-Wert haben, wird neutral 50 vergeben,\n", "# da in diesem Fall keine sinnvolle Differenzierung möglich ist.\n", "\n", "def calculate_comfort_score(humidex_series):\n", " \"\"\"Normalisiert die Humidex-Werte eines aktuellen RTD-Snapshots auf einen relativen Comfort Score von 0 bis 100.\"\"\"\n", " min_h = humidex_series.min()\n", " max_h = humidex_series.max()\n", " if max_h == min_h:\n", " return pd.Series([50.0] * len(humidex_series), index=humidex_series.index)\n", " return ((max_h - humidex_series) / (max_h - min_h) * 100).round(1)\n", "\n", "# Übersetzt den numerischen Comfort Score in eine leicht verständliche Ampelfarbe.\n", "# Warum wird das benötigt?\n", "# Der Comfort Score ist für die interne Berechnung geeignet, soll für Nutzer:innen\n", "# im Frontend aber möglichst intuitiv dargestellt werden.\n", "# Deshalb wird der Score in vier visuelle Klassen übersetzt:\n", "# grün = kühl, orange = mässig, rot = heiss, grau = keine Daten / Ausreisser.\n", "# Die Ampelfarbe wird später sowohl in der Karte als auch in der Tabelle verwendet\n", "# und bildet die Grundlage für die schnelle visuelle Orientierung im PoC.\n", "\n", "def get_ampel_color(score):\n", " \"\"\"Ordnet einem Comfort Score eine Ampelfarbe zu und markiert fehlende oder ungültige Werte als grau.\"\"\"\n", " if pd.isna(score): return 'gray'\n", " elif score >= 66: return 'green'\n", " elif score >= 33: return 'orange'\n", " else: return 'red'\n", "\n", "# Diese Funktion sortiert eine bereits ausgewählte Menge von Stationen nach\n", "# geografischer Nähe.\n", "# Die Stationen selbst wurden zuvor fachlich ausgewählt\n", "# (z.B. Top-5 nach Comfort Score). Hier geht es nur noch darum,\n", "# diese Stationen in eine möglichst plausible Reihenfolge zu bringen,\n", "# damit sie als Pseudo-Route visualisiert werden können.\n", "#\n", "# Vorgehen:\n", "# Ausgehend vom aktuellen Punkt wird jeweils die nächstgelegene verbleibende\n", "# Station gewählt (Nearest-Neighbour-Heuristik).\n", "#\n", "# Die Distanzlogik beeinflusst somit nur die Reihenfolge der Stationen,\n", "# nicht aber die fachliche Auswahl der \"kühlsten\" oder \"wärmsten\" Stationen.\n", "\n", "def sort_route_by_proximity(df_stations):\n", " \"\"\"Ordnet eine fachlich vorselektierte Menge von Stationen mittels Nearest-Neighbour-Heuristik in eine geografisch sinnvolle Reihenfolge.\"\"\"\n", " stations = df_stations.copy().reset_index(drop=True)\n", " route = [stations.iloc[0]]\n", " remaining = stations.iloc[1:].copy()\n", " while len(remaining) > 0:\n", " last = route[-1]\n", " distances = remaining.apply(\n", " lambda row: haversine(last.geometry.y, last.geometry.x,\n", " row.geometry.y, row.geometry.x), axis=1)\n", " nearest_idx = distances.idxmin()\n", " route.append(remaining.loc[nearest_idx])\n", " remaining = remaining.drop(nearest_idx)\n", " return pd.DataFrame(route)\n", "\n", "# Diese Funktion bestimmt die geografisch sinnvollste Reihenfolge für eine bereits\n", "# ausgewählte Menge von Stationen.\n", "# Wichtig: Die Auswahl der Stationen erfolgt bereits vorher über den Comfort Score\n", "# (z.B. Top-5 kühlste oder Top-5 wärmste Stationen im aktuellen RTD-Snapshot).\n", "# Diese Funktion entscheidet also nicht, welche Stationen zur Route gehören,\n", "# sondern nur, in welcher Reihenfolge genau diese Stationen verbunden werden.\n", "#\n", "# Dafür wird für mehrere mögliche Startpunkte jeweils eine Route mit der\n", "# Nearest-Neighbour-Heuristik erzeugt. Anschliessend wird jene Variante gewählt,\n", "# deren Gesamtdistanz am kleinsten ist.\n", "#\n", "# Ergebnis:\n", "# - die Route bleibt fachlich eine \"kühle\" bzw. \"warme\" Route, weil die Stationen\n", "# bereits nach Comfort Score ausgewählt wurden\n", "# - die Distanzlogik verbessert nur die Darstellbarkeit als zusammenhängende\n", "# Pseudo-Route auf der Karte\n", "#\n", "# Hinweis:\n", "# Es handelt sich nicht um echtes Strassenrouting und auch nicht um einen\n", "# vollständigen TSP-Optimierungsalgorithmus, sondern um eine einfache,\n", "# nachvollziehbare Heuristik für den PoC.\n", "\n", "def find_best_route(df_stations):\n", " \"\"\"Ermittelt für eine vorselektierte Stationsmenge den günstigsten Startpunkt und die distanzärmste Pseudo-Route innerhalb der Nearest-Neighbour-Heuristik.\"\"\"\n", " best_route, best_dist = None, float('inf')\n", " for i in range(len(df_stations)):\n", " stations = pd.concat([df_stations.iloc[i:], df_stations.iloc[:i]])\n", " route = sort_route_by_proximity(stations.reset_index(drop=True))\n", " dist = sum(haversine(route.iloc[j].geometry.y, route.iloc[j].geometry.x,\n", " route.iloc[j+1].geometry.y, route.iloc[j+1].geometry.x)\n", " for j in range(len(route)-1))\n", " if dist < best_dist:\n", " best_dist, best_route = dist, route\n", " return best_route, best_dist\n", "\n", "print('Hilfsfunktionen definiert')" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "my7ODEXxweaS", "outputId": "3d862dc6-050b-4e27-c1a8-cfb6e44bfd4c" }, "execution_count": 32, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "Hilfsfunktionen definiert\n" ] } ] }, { "cell_type": "markdown", "source": [ "### 2.2 Datenabfrage & Bereinigung\n", "Laden der Realtime-Daten via REST API, Filterung veralteter Messungen und geografische Einschränkung auf den Raum Bern (10km Radius).\n" ], "metadata": { "id": "8KbGATsr1RNM" } }, { "cell_type": "code", "source": [ "#RTD: (/api/v2/latest)\n", "#Ziel: Nur valide, aktuelle Messwerte im Raum Bern, damit Humidex/Score/Routen korrekt sind.\n", "\n", "# API-Call mit Fehlerbehandlung\n", "# timeout verhindert \"Hängenbleiben\"\n", "response = requests.get(URL_LATEST, timeout=30)\n", "#raise_for_status stoppt sofort bei HTTP-Fehlern (crash verhindern)\n", "response.raise_for_status()\n", "\n", "# GeoJSON sauber parsen\n", "#response.json() liefert strukturierte Daten (kein fragiles Parsen über Text)\n", "geojson = response.json()\n", "# features enthält die Geometrien + Messwerte; wenn es fehlt, ist die Antwort unbrauchbar\n", "#In GeoDataFrame umwandeln, später nötig für Distanzberechnung und Kartenvisualisierung\n", "features = geojson.get(\"features\")\n", "if not features:\n", " raise ValueError(\"GeoJSON enthaelt keine 'features'.\")\n", "stations_raw = gpd.GeoDataFrame.from_features(features)\n", "#EPSG:4326 = WGS84 (Lon/Lat), Standard für GPS-Koordinaten\n", "#wichtig, damit geometry.x/geometry.y korrekt interpretiert werden kann\n", "if stations_raw.crs is None:\n", " stations_raw.set_crs(epsg=4326, inplace=True)\n", "\n", "#Transparenz/Debug:\n", "print(f\"Stationen total: {len(stations_raw)}\")\n", "print(f\"outdated == True: {(stations_raw['outdated'] == True).sum()}\")\n", "if \"measurementsPlausible\" in stations_raw.columns:\n", " print(f\"measurementsPlausible==False: {(stations_raw['measurementsPlausible'] == False).sum()}\")\n", "\n", "#Bereinigung: outdated + measurementsPlausible\n", "#copy() verhindert SettingWithCopy-Probleme bei späteren Spaltenzuweisungen\n", "stations_clean = stations_raw[stations_raw[\"outdated\"] == False].copy()\n", "\n", "#Bereinigung: measurementsPlausible filtern (falls vorhanden):\n", "# - schützt vor Messwerten, die vom Datenprovider als unplausibel markiert wurden\n", "# - falls die Spalte nicht existiert, bleibt die Pipeline dennoch lauffähig\n", "if \"measurementsPlausible\" in stations_clean.columns:\n", " stations_clean = stations_clean[stations_clean[\"measurementsPlausible\"] == True].copy()\n", "print(f\"Nach Bereinigung: {len(stations_clean)} Stationen\")\n", "\n", "# Geo-Filter\n", "#reduziert Stationen auf den relevanten Nutzungsraum (Bern & Umgebung)\n", "#verhindert, dass entfernte Stationen die Normalisierung/Route beeinflussen\n", "#vektorisierte Distanzberechnung ist schneller als DataFrame.apply schneller als DataFrame.apply)\n", "stations_clean[\"dist_km\"] = np.vectorize(haversine)(\n", " BERN_LAT, BERN_LON,\n", " stations_clean.geometry.y.to_numpy(),\n", " stations_clean.geometry.x.to_numpy()\n", ")\n", "df_rtd_bern = stations_clean[stations_clean[\"dist_km\"] <= GEO_RADIUS_KM].copy()\n", "print(f\"Nach Geo-Filter ({GEO_RADIUS_KM}km): {len(df_rtd_bern)} Stationen\")\n", "\n", "# Pflichtfelder-Check, sicherstellen das alle Variablen vorhanden sind für Humidex/Score/Visualisierung\n", "needed = [\"temperature\", \"relativeHumidity\", \"name\"]\n", "missing = [c for c in needed if c not in df_rtd_bern.columns]\n", "if missing:\n", " raise ValueError(f\"Fehlende Spalten: {missing}\")\n", "\n", "#Drop NA + Transparenz\n", "#Protokollierung, wie viele Stationen entfernt werden\n", "n_before_dropna = len(df_rtd_bern)\n", "df_rtd_bern = df_rtd_bern.dropna(subset=[\"temperature\", \"relativeHumidity\"]).copy()\n", "n_dropped = n_before_dropna - len(df_rtd_bern)\n", "if n_dropped > 0:\n", " print(f\"Stationen wegen fehlender Messwerte entfernt: {n_dropped}\")\n", "print(f\"Nach Bereinigung + Geo-Filter: {len(df_rtd_bern)} Stationen\")\n", "\n", "#Qualitätskontrolle (sollte nach dropna 0 sein):\n", "#schnelle Verifikation, dass die Daten jetzt berechnungsbereit sind\n", "print(\"\\nFehlende Werte (nach dropna):\")\n", "print(df_rtd_bern[[\"temperature\", \"relativeHumidity\"]].isna().sum())" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "hQchpDsX2Nol", "outputId": "79f9bee6-641d-46ec-e1b5-4ed2df860a5e" }, "execution_count": 33, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "Stationen total: 64\n", "outdated == True: 26\n", "measurementsPlausible==False: 0\n", "Nach Bereinigung: 38 Stationen\n", "Nach Geo-Filter (10km): 31 Stationen\n", "Nach Bereinigung + Geo-Filter: 31 Stationen\n", "\n", "Fehlende Werte (nach dropna):\n", "temperature 0\n", "relativeHumidity 0\n", "dtype: int64\n" ] } ] }, { "cell_type": "markdown", "source": [ "## 3. Hitzebelastungsberechnung\n", "### 3.1 Humidex & Comfort Score\n", "Berechnung des Humidex sowie des relativen Comfort Scores (0–100) für alle aktiven Stationen im Raum Bern. Der Comfort Score normiert die Humidex-Werte aller Stationen zueinander: 100% = kühlste Station, 0% = wärmste Station." ], "metadata": { "id": "Yy-TM5TH7b-Z" } }, { "cell_type": "code", "source": [ "# Humidex vektorisiert berechnen: np.vectorize wendet calculate_humidex\n", "# elementweise auf alle Stationen an (schneller als axis=1 apply)\n", "df_rtd_bern[\"humidex\"] = np.vectorize(calculate_humidex)(\n", " df_rtd_bern[\"temperature\"].to_numpy(),\n", " df_rtd_bern[\"relativeHumidity\"].to_numpy()\n", ")\n", "\n", "# Ausreisser-Erkennung via MAD (Median Absolute Deviation)\n", "# robuster als Z-Score bei kleinen n (ca. 31 Stationen) und nicht-normaler Verteilung\n", "# muss VOR Comfort Score berechnet werden, sonst verzerrt Ausreisser Min/Max\n", "h = df_rtd_bern[\"humidex\"].dropna()\n", "med = h.median()\n", "mad = (h - med).abs().median()\n", "\n", "if pd.isna(mad) or mad == 0:\n", " print(\"MAD Outlier Detection übersprungen (MAD=0 oder NaN).\")\n", "else:\n", " modified_z = 0.6745 * (df_rtd_bern[\"humidex\"] - med) / mad\n", " mask_out = modified_z.abs() > 3.5\n", "\n", " outliers = df_rtd_bern[mask_out]\n", " if len(outliers) > 0:\n", " print(f\"Ausreisser markiert ({len(outliers)} Station(en)):\")\n", " print(outliers[[\"name\", \"temperature\", \"relativeHumidity\", \"humidex\"]]\n", " .to_string(index=False))\n", "\n", " # Ausreisser als NaN markieren – bleiben in Karte als gray sichtbar\n", " # werden in df_rtd_valid (Route-Selektion) automatisch ausgeschlossen\n", " df_rtd_bern.loc[mask_out, \"humidex\"] = np.nan\n", "\n", "# Comfort Score normalisiert innerhalb des Geo-Filters (nur Berner Stationen)\n", "# wichtig: Score NACH Ausreisser-Erkennung berechnen\n", "df_rtd_bern[\"comfort_score\"] = calculate_comfort_score(df_rtd_bern[\"humidex\"])\n", "\n", "# Ampelfarbe zuweisen: green >= 66%, orange 33-65%, red < 33%, gray = NaN\n", "# Score wird aus aktuellem Snapshot berechnet, nicht persistent gespeichert\n", "df_rtd_bern[\"color\"] = df_rtd_bern[\"comfort_score\"].apply(get_ampel_color)\n", "\n", "# df_rtd_valid: Stationen mit validen Humidex/Score-Werten fuer Route-Selektion\n", "# verhindert dass NaN-Stationen in Routenberechnung einfliessen\n", "df_rtd_valid = df_rtd_bern.dropna(subset=[\"humidex\", \"comfort_score\"]).copy()\n", "print(f\"Stationen fuer Routenberechnung: {len(df_rtd_valid)}\")\n", "\n", "# Qualitaetscheck: verbleibende NaNs protokollieren\n", "n_humidex_nan = df_rtd_bern[\"humidex\"].isna().sum()\n", "if n_humidex_nan > 0:\n", " print(f\"Stationen mit Humidex=NaN (Ausreisser + fehlende Werte): {n_humidex_nan}\")\n", "\n", "# Kontrollausgabe: Top/Bottom 5 zur Verifikation\n", "print(\"\\nTop 5 kühl (höchster Comfort Score):\")\n", "print(df_rtd_bern[[\"name\", \"humidex\", \"comfort_score\", \"color\"]]\n", " .sort_values(\"comfort_score\", ascending=False)\n", " .head(5).to_string(index=False))\n", "\n", "print(\"\\nTop 5 warm (tiefster Comfort Score):\")\n", "print(df_rtd_bern[[\"name\", \"humidex\", \"comfort_score\", \"color\"]]\n", " .sort_values(\"comfort_score\", ascending=True)\n", " .head(5).to_string(index=False))" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "0r8LCf7K_Nwy", "outputId": "688e04b0-7dbb-412c-d585-6a3e1931e4f4" }, "execution_count": 34, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "Ausreisser markiert (4 Station(en)):\n", " name temperature relativeHumidity humidex\n", " Eigerplatz 14.153124 48.768444 13.0\n", " Kehrsatz Breitägertenstrasse Laterne 12.235828 47.374150 10.4\n", "Köniz Blinzernplateau Strom-/Telefonmast 12.107653 49.891890 10.5\n", " Bern Hintere Länggasse 3.573282 91.127490 2.0\n", "Stationen fuer Routenberechnung: 27\n", "Stationen mit Humidex=NaN (Ausreisser + fehlende Werte): 4\n", "\n", "Top 5 kühl (höchster Comfort Score):\n", " name humidex comfort_score color\n", "Liebefeld Turnierstrasse Laterne 10.7 100.0 green\n", " Köniz Gartenstadt Laterne 10.7 100.0 green\n", " Zollikofen 3m 11.3 66.7 green\n", " Gümligen Dentenberg Laterne 11.3 66.7 green\n", " Ittigen Gantrischweg 1 Laterne 11.4 61.1 orange\n", "\n", "Top 5 warm (tiefster Comfort Score):\n", " name humidex comfort_score color\n", " Helvetiaplatz 12.5 0.0 red\n", " Hirschengraben 12.3 11.1 red\n", "Emch+Berger (Zieglerstrasse) 12.1 22.2 red\n", " Ittigen Worblentalstrasse 12.1 22.2 red\n", " BernExpo Asphalt 0.5m 12.1 22.2 red\n" ] } ] }, { "cell_type": "markdown", "source": [ "## 4. Visualisierung\n", "### 4.1 Interaktive Karte mit Ampelfarben und Routen\n", "Darstellung aller aktiven Stationen mit Ampelfarben basierend auf dem Comfort Score. Die kühlste Route (grün) verbindet die 5 Stationen mit dem höchsten Comfort Score, die wärmste Route (rot) die 5 Stationen mit dem tiefsten Score." ], "metadata": { "id": "xPqg72PMBE_c" } }, { "cell_type": "code", "source": [ "# Hilfsfunktion: sichere Formatierung (verhindert Crash bei NaN-Werten im Tooltip)\n", "def fmt(x, suffix=\"\", decimals=1):\n", " return \"n/a\" if pd.isna(x) else f\"{x:.{decimals}f}{suffix}\"\n", "\n", "# Guard: mindestens 2 valide Stationen nötig für Routenberechnung\n", "if len(df_rtd_valid) < 2:\n", " raise ValueError(\"Zu wenige valide Stationen fuer Routenberechnung.\")\n", "\n", "# Top 5 kühlste und wärmste Stationen aus validen Stationen selektieren\n", "# df_rtd_valid schliesst Ausreisser und NaN-Stationen bereits aus\n", "top5_cool = df_rtd_valid.nlargest(5, \"comfort_score\").copy()\n", "top5_warm = df_rtd_valid.nsmallest(5, \"comfort_score\").copy()\n", "\n", "# Routen optimieren: Heuristik zur Minimierung von Sprüngen (kein Strassenrouting)\n", "# testet alle Startpunkte, nimmt die kürzeste Nearest-Neighbour Route\n", "cool_route, cool_dist = find_best_route(top5_cool)\n", "warm_route, warm_dist = find_best_route(top5_warm)\n", "\n", "# Reihenfolge = geografisch (Nearest-Neighbour), nicht nach Score-Ranking\n", "print(f\"Top 5 Stationen (kühlste) – Route nach Nähe sortiert (~{cool_dist:.1f} km Luftlinie):\")\n", "for _, s in cool_route.iterrows():\n", " print(f\" {s['name']} (Score: {s['comfort_score']}%)\")\n", "\n", "print(f\"\\nTop 5 Stationen (wärmste) – Route nach Nähe sortiert (~{warm_dist:.1f} km Luftlinie):\")\n", "for _, s in warm_route.iterrows():\n", " print(f\" {s['name']} (Score: {s['comfort_score']}%)\")\n", "\n", "# Karte auf valide Stationen zentrieren (Ausreisser könnten Zentrum verschieben)\n", "m = folium.Map(\n", " location=[df_rtd_valid.geometry.y.mean(), df_rtd_valid.geometry.x.mean()],\n", " zoom_start=13\n", ")\n", "\n", "# Alle Stationen als Marker (inkl. gray für Ausreisser/NaN)\n", "for _, station in df_rtd_bern.iterrows():\n", " tooltip = (\n", " f\"{station['name']}
\"\n", " f\"Temp: {fmt(station['temperature'], '°C')} | \"\n", " f\"Luftf: {fmt(station['relativeHumidity'], '%')}
\"\n", " f\"Humidex: {fmt(station['humidex'])}
\"\n", " f\"Comfort Score: {fmt(station['comfort_score'], '%')}\"\n", " )\n", " folium.CircleMarker(\n", " location=(station.geometry.y, station.geometry.x),\n", " radius=8,\n", " color=\"black\", weight=0.5,\n", " fill=True,\n", " fill_color=station[\"color\"],\n", " fill_opacity=0.8,\n", " tooltip=folium.Tooltip(tooltip, sticky=True)\n", " ).add_to(m)\n", "\n", "# Kühlste Route (grün, durchgezogen)\n", "# Hinweis: Top-5 bedeutet Rangliste; einzelne Stationen können trotz Top-5 im Orange-Bereich liegen.\n", "cool_coords = [[s.geometry.y, s.geometry.x] for _, s in cool_route.iterrows()]\n", "folium.PolyLine(\n", " locations=cool_coords,\n", " color=\"green\", weight=5, opacity=0.9,\n", " tooltip=\"Route kühlste Stationen (Top-5) – Luftlinie\"\n", ").add_to(m)\n", "\n", "# Wärmste Route (rot, gestrichelt = visuell als \"meiden\" markiert)\n", "warm_coords = [[s.geometry.y, s.geometry.x] for _, s in warm_route.iterrows()]\n", "folium.PolyLine(\n", " locations=warm_coords,\n", " color=\"red\", weight=5, opacity=0.9,\n", " dash_array=\"10\",\n", " tooltip=\"Route wärmste Stationen (Top-5) – meiden\"\n", ").add_to(m)\n", "\n", "# Legende (unten links)\n", "legend_html = '''\n", "
\n", " HeatSafeBern
\n", "
\n", " Comfort Score:
\n", " ● Grün ≥ 66% (kühl)
\n", " ● Orange 33–65% (mässig)
\n", " ● Rot < 33% (heiss)
\n", " ● Grau = Ausreisser / keine Daten
\n", "
\n", " Routen (Luftlinie, Top-5 Stationen):
\n", " ▬ Route kühlste Stationen
\n", " ▬ ▬ Route wärmste Stationen (meiden)\n", "
'''\n", "m.get_root().html.add_child(folium.Element(legend_html))\n", "\n", "m\n" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 1000 }, "id": "qFo9dXKmHSWW", "outputId": "6668e998-2e16-4008-ac10-eb8a9532ce9f" }, "execution_count": 35, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "Top 5 Stationen (kühlste) – Route nach Nähe sortiert (~15.3 km Luftlinie):\n", " Köniz Gartenstadt Laterne (Score: 100.0%)\n", " Liebefeld Turnierstrasse Laterne (Score: 100.0%)\n", " Zollikofen 3m (Score: 66.7%)\n", " Ittigen Gantrischweg 1 Laterne (Score: 61.1%)\n", " Gümligen Dentenberg Laterne (Score: 66.7%)\n", "\n", "Top 5 Stationen (wärmste) – Route nach Nähe sortiert (~6.0 km Luftlinie):\n", " Emch+Berger (Zieglerstrasse) (Score: 22.2%)\n", " Hirschengraben (Score: 11.1%)\n", " Helvetiaplatz (Score: 0.0%)\n", " BernExpo Asphalt 0.5m (Score: 22.2%)\n", " Ittigen Worblentalstrasse (Score: 22.2%)\n" ] }, { "output_type": "execute_result", "data": { "text/plain": [ "" ], "text/html": [ "
Make this Notebook Trusted to load map: File -> Trust Notebook
" ] }, "metadata": {}, "execution_count": 35 } ] }, { "cell_type": "markdown", "source": [ "### 4.2 Stationstabelle\n", "Tabellarische Übersicht aller aktiven Stationen im Raum Bern, sortiert nach Comfort Score (absteigend). Zeigt Temperatur, Luftfeuchtigkeit, Humidex und Ampelklassifikation auf einen Blick." ], "metadata": { "id": "3IRtL1ALI3Tg" } }, { "cell_type": "code", "source": [ "# Ampel-Emoji für Lesbarkeit in der Tabelle\n", "def get_emoji(color):\n", " # visuelles Feedback für User: Ampelfarbe als Emoji\n", " mapping = {\"green\": \"🟢\", \"orange\": \"🟡\", \"red\": \"🔴\", \"gray\": \"⚫\"}\n", " return mapping.get(color, \"⚫\")\n", "\n", "# Tabelle aufbereiten: nur relevante Spalten, sauber benannt\n", "df_table = df_rtd_bern[[\n", " \"name\", \"temperature\", \"relativeHumidity\", \"humidex\", \"comfort_score\", \"color\"\n", "]].copy()\n", "\n", "# Zahlen auf 1 Dezimale runden (Lesbarkeit)\n", "df_table[[\"temperature\", \"relativeHumidity\", \"humidex\", \"comfort_score\"]] = \\\n", " df_table[[\"temperature\", \"relativeHumidity\", \"humidex\", \"comfort_score\"]].round(1)\n", "\n", "df_table[\"Ampel\"] = df_table[\"color\"].apply(get_emoji)\n", "\n", "# Spalten umbenennen für Lesbarkeit\n", "df_table = df_table.rename(columns={\n", " \"name\": \"Station\",\n", " \"temperature\": \"Temp (°C)\",\n", " \"relativeHumidity\": \"Luftf. (%)\",\n", " \"humidex\": \"Humidex\",\n", " \"comfort_score\": \"Comfort Score (%)\"\n", "}).drop(columns=[\"color\"])\n", "\n", "# Nach Comfort Score sortieren (beste zuerst)\n", "# NaN-Stationen (Ausreisser) landen am Ende\n", "df_table = df_table.sort_values(\n", " \"Comfort Score (%)\", ascending=False, na_position=\"last\"\n", ").reset_index(drop=True)\n", "\n", "# Stand: explizit als lokale Notebook-Zeit kennzeichnen (Colab = meist UTC)\n", "print(f\"HeatSafeBern – Aktuelle Hitzebelastung Bern\")\n", "print(f\"Stand (lokale Notebook-Zeit): {datetime.now().strftime('%d.%m.%Y %H:%M')} Uhr\")\n", "print(f\"Stationen: {len(df_table)} total, \"\n", " f\"{df_table['Ampel'].eq('⚫').sum()} keine Daten/Ausreisser\\n\")\n", "\n", "display(df_table)" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 1000 }, "id": "P6WtidqwJhTZ", "outputId": "2da14f3b-5241-4cbf-df6d-db77110b3496" }, "execution_count": 36, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "HeatSafeBern – Aktuelle Hitzebelastung Bern\n", "Stand (lokale Notebook-Zeit): 13.03.2026 17:20 Uhr\n", "Stationen: 31 total, 4 keine Daten/Ausreisser\n", "\n" ] }, { "output_type": "display_data", "data": { "text/plain": [ " Station Temp (°C) Luftf. (%) Humidex \\\n", "0 Liebefeld Turnierstrasse Laterne 12.4 48.1 10.7 \n", "1 Köniz Gartenstadt Laterne 12.4 48.1 10.7 \n", "2 Zollikofen 3m 12.9 48.2 11.3 \n", "3 Gümligen Dentenberg Laterne 12.9 47.1 11.3 \n", "4 Ittigen Gantrischweg 1 Laterne 13.1 46.8 11.4 \n", "5 Wabern Zentrum Laterne 13.3 45.0 11.5 \n", "6 Bern Merzenacker 13.2 46.2 11.5 \n", "7 Muri Blümlisalpstrasse 13.4 44.7 11.6 \n", "8 Muri Zentrum (Tavelweg) Laterne 13.4 44.4 11.6 \n", "9 Bremgartenwald 12.8 53.6 11.6 \n", "10 Vordere Länggasse 13.1 49.3 11.7 \n", "11 Bolligen Sternenkreisel Laterne 13.6 42.9 11.7 \n", "12 Muri Süd (Steinhübeliweg) Laterne 13.5 43.4 11.7 \n", "13 BernExpo Park 2.5m 13.6 44.2 11.8 \n", "14 BernExpo Asphalt 2.5m 13.5 44.8 11.8 \n", "15 Bern Lorrainepark 13.5 44.8 11.8 \n", "16 Wabern Gartentower Laterne 13.5 45.4 11.8 \n", "17 Wohlen Hinterkappelenring Laterne 13.5 45.3 11.8 \n", "18 Wabern Weyergut 13.4 45.5 11.8 \n", "19 Zollikofen Reichenbach Laterne 13.7 43.7 11.9 \n", "20 Bern Dalmaziquai 13.6 43.7 11.9 \n", "21 Galgenfeld Neubausiedlung 13.3 49.7 12.0 \n", "22 Ittigen Worblentalstrasse 13.8 44.2 12.1 \n", "23 BernExpo Asphalt 0.5m 13.7 45.7 12.1 \n", "24 Emch+Berger (Zieglerstrasse) 13.8 43.6 12.1 \n", "25 Hirschengraben 13.9 45.5 12.3 \n", "26 Helvetiaplatz 13.7 49.1 12.5 \n", "27 Eigerplatz 14.2 48.8 NaN \n", "28 Kehrsatz Breitägertenstrasse Laterne 12.2 47.4 NaN \n", "29 Köniz Blinzernplateau Strom-/Telefonmast 12.1 49.9 NaN \n", "30 Bern Hintere Länggasse 3.6 91.1 NaN \n", "\n", " Comfort Score (%) Ampel \n", "0 100.0 🟢 \n", "1 100.0 🟢 \n", "2 66.7 🟢 \n", "3 66.7 🟢 \n", "4 61.1 🟡 \n", "5 55.6 🟡 \n", "6 55.6 🟡 \n", "7 50.0 🟡 \n", "8 50.0 🟡 \n", "9 50.0 🟡 \n", "10 44.4 🟡 \n", "11 44.4 🟡 \n", "12 44.4 🟡 \n", "13 38.9 🟡 \n", "14 38.9 🟡 \n", "15 38.9 🟡 \n", "16 38.9 🟡 \n", "17 38.9 🟡 \n", "18 38.9 🟡 \n", "19 33.3 🟡 \n", "20 33.3 🟡 \n", "21 27.8 🔴 \n", "22 22.2 🔴 \n", "23 22.2 🔴 \n", "24 22.2 🔴 \n", "25 11.1 🔴 \n", "26 0.0 🔴 \n", "27 NaN ⚫ \n", "28 NaN ⚫ \n", "29 NaN ⚫ \n", "30 NaN ⚫ " ], "text/html": [ "\n", "
\n", "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
StationTemp (°C)Luftf. (%)HumidexComfort Score (%)Ampel
0Liebefeld Turnierstrasse Laterne12.448.110.7100.0🟢
1Köniz Gartenstadt Laterne12.448.110.7100.0🟢
2Zollikofen 3m12.948.211.366.7🟢
3Gümligen Dentenberg Laterne12.947.111.366.7🟢
4Ittigen Gantrischweg 1 Laterne13.146.811.461.1🟡
5Wabern Zentrum Laterne13.345.011.555.6🟡
6Bern Merzenacker13.246.211.555.6🟡
7Muri Blümlisalpstrasse13.444.711.650.0🟡
8Muri Zentrum (Tavelweg) Laterne13.444.411.650.0🟡
9Bremgartenwald12.853.611.650.0🟡
10Vordere Länggasse13.149.311.744.4🟡
11Bolligen Sternenkreisel Laterne13.642.911.744.4🟡
12Muri Süd (Steinhübeliweg) Laterne13.543.411.744.4🟡
13BernExpo Park 2.5m13.644.211.838.9🟡
14BernExpo Asphalt 2.5m13.544.811.838.9🟡
15Bern Lorrainepark13.544.811.838.9🟡
16Wabern Gartentower Laterne13.545.411.838.9🟡
17Wohlen Hinterkappelenring Laterne13.545.311.838.9🟡
18Wabern Weyergut13.445.511.838.9🟡
19Zollikofen Reichenbach Laterne13.743.711.933.3🟡
20Bern Dalmaziquai13.643.711.933.3🟡
21Galgenfeld Neubausiedlung13.349.712.027.8🔴
22Ittigen Worblentalstrasse13.844.212.122.2🔴
23BernExpo Asphalt 0.5m13.745.712.122.2🔴
24Emch+Berger (Zieglerstrasse)13.843.612.122.2🔴
25Hirschengraben13.945.512.311.1🔴
26Helvetiaplatz13.749.112.50.0🔴
27Eigerplatz14.248.8NaNNaN⚫
28Kehrsatz Breitägertenstrasse Laterne12.247.4NaNNaN⚫
29Köniz Blinzernplateau Strom-/Telefonmast12.149.9NaNNaN⚫
30Bern Hintere Länggasse3.691.1NaNNaN⚫
\n", "
\n", "
\n", "\n", "
\n", " \n", "\n", " \n", "\n", " \n", "
\n", "\n", "\n", "
\n", " \n", " \n", " \n", "
\n", "\n", "
\n", "
\n" ], "application/vnd.google.colaboratory.intrinsic+json": { "type": "dataframe", "variable_name": "df_table", "summary": "{\n \"name\": \"df_table\",\n \"rows\": 31,\n \"fields\": [\n {\n \"column\": \"Station\",\n \"properties\": {\n \"dtype\": \"string\",\n \"num_unique_values\": 31,\n \"samples\": [\n \"Eigerplatz\",\n \"Bern Lorrainepark\",\n \"BernExpo Asphalt 0.5m\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"Temp (\\u00b0C)\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 1.8132368728297048,\n \"min\": 3.6,\n \"max\": 14.2,\n \"num_unique_values\": 16,\n \"samples\": [\n 12.4,\n 12.9,\n 13.4\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"Luftf. (%)\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 8.407372596625809,\n \"min\": 42.9,\n \"max\": 91.1,\n \"num_unique_values\": 26,\n \"samples\": [\n 53.6,\n 45.5,\n 48.1\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"Humidex\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 0.40333368651535023,\n \"min\": 10.7,\n \"max\": 12.5,\n \"num_unique_values\": 12,\n \"samples\": [\n 12.3,\n 12.1,\n 10.7\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"Comfort Score (%)\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 22.414523934067468,\n \"min\": 0.0,\n \"max\": 100.0,\n \"num_unique_values\": 12,\n \"samples\": [\n 11.1,\n 22.2,\n 100.0\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"Ampel\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 4,\n \"samples\": [\n \"\\ud83d\\udfe1\",\n \"\\u26ab\",\n \"\\ud83d\\udfe2\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n }\n ]\n}" } }, "metadata": {} } ] }, { "cell_type": "markdown", "source": [ "## 5. Historische Analyse – Beste Trainingszeit\n", "Auswertung historischer Zeitreihendaten (Sommer 2025, Juli–August) für mehrere Berner Stationen. Ziel: Identifikation der durchschnittlich kühlsten Tageszeit als indikative Empfehlung für Outdoor-Aktivitäten. Der Humidex-Tagesverlauf wird über sechs geografisch diversifizierte Stationen ermittelt." ], "metadata": { "id": "UdeP01k-KBgu" } }, { "cell_type": "code", "source": [ "#Daten laden\n", "\n", "# Historische Analyse: Sommer 2025 (Juli–August)\n", "# Für den PoC werden Sommerdaten 2025 verwendet (verfügbar gemäss API-Test)\n", "# HIST_TO = 01.09 damit der gesamte August enthalten ist (Ende exklusiv)\n", "HIST_FROM = \"2025-07-01T00:00:00Z\"\n", "HIST_TO = \"2025-09-01T00:00:00Z\"\n", "\n", "# Fest definierte Stationen – geografisch diversifiziert über den Raum Bern\n", "# Bewusste Auswahl statt df_rtd_valid.head(): head() würde die aktuell kühlsten\n", "# Stationen nehmen – diese wechseln je nach Tageszeit und wären nicht reproduzierbar.\n", "# Fest kodierte IDs garantieren, dass immer dieselben Lagen verglichen werden:\n", "# Innenstadt (Wärmeinsel), Wohnquartier, Grünfläche, Stadtrand Nord/Süd\n", "HIST_STATIONS = [\n", " {\"stationId\": 11070, \"name\": \"Hirschengraben\"}, # Innenstadt / Wärmeinsel\n", " {\"stationId\": 11046, \"name\": \"Bern Merzenacker\"}, # Grünfläche / Park\n", " {\"stationId\": 11037, \"name\": \"Eigenplatz\"}, # Stadtmitte Referenz\n", " {\"stationId\": 11060, \"name\": \"Vordere Länggasse\"}, # Wohnquartier\n", " {\"stationId\": 11080, \"name\": \"Wabern Weyergut\"}, # Südlicher Stadtrand\n", " {\"stationId\": 11117, \"name\": \"Ittigen Papiermühle\"}, # Nördlicher Rand / Aare-Nähe\n", "]\n", "\n", "# Pflichtfelder: ohne diese drei Spalten ist Humidex-Berechnung nicht möglich\n", "HIST_REQUIRED_COLS = [\"dateObserved\", \"temperature\", \"relativeHumidity\"]\n", "\n", "all_timeseries = []\n", "\n", "for station in HIST_STATIONS:\n", " # Timeseries-Endpunkt liefert CSV mit zeitlich hochauflösende Messwerte pro Station\n", " # stationId als URL-Parameter, Zeitraum als ISO-8601 UTC\n", " url = (f\"{URL_TIMESERIES}?stationId={station['stationId']}\"\n", " f\"&timeFrom={HIST_FROM}&timeTo={HIST_TO}\")\n", " try:\n", " resp = requests.get(url, headers={\"Accept\": \"text/csv\"}, timeout=20)\n", " # raise_for_status stoppt sofort bei HTTP-Fehlern (z.B. 404, 500)\n", " resp.raise_for_status()\n", "\n", " # Content-Type prüfen: sicherstellen dass wirklich CSV geliefert wurde\n", " # verhindert stilles Scheitern, falls API plötzlich JSON zurückgibt\n", " content_type = resp.headers.get(\"content-type\", \"\")\n", " if \"csv\" not in content_type and \"text\" not in content_type:\n", " print(f\" {station['name']}: unerwarteter Content-Type ({content_type})\")\n", " continue\n", "\n", " # delimiter=\";\" weil Smart Urban Heat Map API Semikolon-getrennte CSVs liefert\n", " ts = pd.read_csv(io.StringIO(resp.text), delimiter=\";\")\n", "\n", " # Pflichtfelder-Check: fehlende Spalten würden später zu Crash führen\n", " # besser hier abfangen mit verständlicher Fehlermeldung\n", " missing_cols = [c for c in HIST_REQUIRED_COLS if c not in ts.columns]\n", " if missing_cols:\n", " print(f\" {station['name']}: fehlende Spalten {missing_cols}\")\n", " continue\n", "\n", " if len(ts) > 0:\n", " # Station-Metadaten anhängen, damit nach pd.concat noch erkennbar ist,\n", " # welche Zeile von welcher Station stammt\n", " ts[\"stationId\"] = station[\"stationId\"]\n", " ts[\"stationName\"] = station[\"name\"]\n", " all_timeseries.append(ts)\n", " print(f\" ✓ {station['name']}: {len(ts)} Datenpunkte\")\n", " else:\n", " print(f\" {station['name']}: keine historischen Daten\")\n", "\n", " except Exception as e:\n", " print(f\" {station['name']}: Fehler – {e}\")\n", "\n", "# Transparenz: wie viele Stationen haben tatsächlich Daten geliefert?\n", "# falls < 6, trotzdem weiterrechnen – Mittel bleibt aussagekräftig\n", "print(f\"\\nGeladen: {len(all_timeseries)} / {len(HIST_STATIONS)} Stationen mit historischen Daten\")" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "33K_itCgSdRN", "outputId": "93520e91-1102-4ebc-ba8f-d2b3736f662e" }, "execution_count": 37, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ " ✓ Hirschengraben: 8091 Datenpunkte\n", " ✓ Bern Merzenacker: 8572 Datenpunkte\n", " ✓ Eigenplatz: 8476 Datenpunkte\n", " ✓ Vordere Länggasse: 8575 Datenpunkte\n", " ✓ Wabern Weyergut: 8485 Datenpunkte\n", " ✓ Ittigen Papiermühle: 6799 Datenpunkte\n", "\n", "Geladen: 6 / 6 Stationen mit historischen Daten\n" ] } ] }, { "cell_type": "code", "source": [ "# Guard: ohne Daten ist die Analyse nicht möglich\n", "if not all_timeseries:\n", " raise ValueError(\"Keine historischen Daten geladen – Analyse abgebrochen.\")\n", "\n", "# Alle Stationen zu einem DataFrame zusammenführen\n", "# ignore_index=True, damit der Index eindeutig bleibt (jede Station startet bei 0)\n", "df_hist = pd.concat(all_timeseries, ignore_index=True)\n", "\n", "# BOM/Whitespace-Schutz: manche CSV-Exporte enthalten ein Byte-Order-Mark (\\ufeff)\n", "# am Anfang der ersten Spalte → würde \"dateObserved\" zu \"\\ufeffdateObserved\" machen\n", "# und den Pflichtfelder-Check still zum Scheitern bringen\n", "df_hist.columns = df_hist.columns.str.replace(\"\\ufeff\", \"\", regex=False).str.strip()\n", "\n", "# Zeitstempel parsen: utc=True weil API UTC-Zeitstempel liefert\n", "# errors=\"coerce\" wandelt unparsbare Werte in NaT um statt Crash\n", "df_hist[\"dateObserved\"] = pd.to_datetime(df_hist[\"dateObserved\"], utc=True, errors=\"coerce\")\n", "# Zeilen ohne gültigen Zeitstempel können nicht einer Tagesstunde zugeordnet werden\n", "df_hist = df_hist.dropna(subset=[\"dateObserved\"])\n", "\n", "# Lokalzeit Bern (UTC+2 im Sommer) für intuitive Stundenachse\n", "# ohne Konvertierung würde 06:00 Uhr lokaler Morgen als 04:00 UTC erscheinen\n", "df_hist[\"hour\"] = df_hist[\"dateObserved\"].dt.tz_convert(\"Europe/Zurich\").dt.hour\n", "\n", "# Numerische Typen erzwingen: CSVs können BOM oder Komma-Dezimal enthalten,\n", "# was pandas als String einliest – to_numeric konvertiert sauber, Rest wird NaN\n", "df_hist[\"temperature\"] = pd.to_numeric(df_hist[\"temperature\"], errors=\"coerce\")\n", "df_hist[\"relativeHumidity\"] = pd.to_numeric(df_hist[\"relativeHumidity\"], errors=\"coerce\")\n", "\n", "# Zeilen mit ungültigen Messwerten entfernen BEVOR Humidex berechnet wird\n", "# sonst entstehen viele NaN-Humidex, die das Stundenmittel still verzerren\n", "n_before = len(df_hist)\n", "df_hist = df_hist.dropna(subset=[\"temperature\", \"relativeHumidity\"])\n", "n_dropped = n_before - len(df_hist)\n", "if n_dropped > 0:\n", " print(f\"Zeilen mit ungültigen Messwerten entfernt: {n_dropped}\")\n", "\n", "# Humidex vektorisiert berechnen: np.vectorize wendet calculate_humidex\n", "# elementweise auf alle Zeilen an (schneller als axis=1 apply)\n", "# calculate_humidex gibt NaN zurück bei rh<=0 oder rh>100 (seltene Ausreisser)\n", "# pandas groupby ignoriert NaN bei mean/std/count → kein Einfluss auf Ergebnis\n", "df_hist[\"humidex\"] = np.vectorize(calculate_humidex)(\n", " df_hist[\"temperature\"].to_numpy(),\n", " df_hist[\"relativeHumidity\"].to_numpy()\n", ")\n", "# Transparenz: wie viele Humidex-NaNs sind trotz gültiger temp/RH entstanden?\n", "n_humidex_nan = df_hist[\"humidex\"].isna().sum()\n", "if n_humidex_nan > 0:\n", " print(f\"Humidex-NaNs nach Berechnung (rh<=0 oder rh>100): {n_humidex_nan}\")\n", "\n", "# Stundenmittel über alle Stationen und alle Tage im Zeitraum\n", "# Warum Humidex und nicht Comfort Score?\n", "# Der Comfort Score ist eine relative Momentaufnahme: er normalisiert Stationen\n", "# gegeneinander innerhalb eines einzigen RTD-Snapshots (100% = kühlste Station\n", "# *gerade jetzt*). Über die Zeit ist er nicht vergleichbar – um 06:00 Uhr wäre\n", "# eine Station \"100%\", um 14:00 Uhr eine andere, obwohl es insgesamt viel heisser\n", "# ist. Der Humidex ist ein absoluter Wert und macht Tagesstunden direkt\n", "# vergleichbar: ein niedrigerer Ø Humidex um 06:00 Uhr bedeutet unabhängig vom\n", "# Snapshot wirklich kühlere Bedingungen als um 14:00 Uhr.\n", "# groupby(hour) → für jede der 24 Stunden: Mittelwert, Streuung, Anzahl Messwerte\n", "# std wird später als Konfidenzband im Plot verwendet\n", "# sort_values(\"hour\") sicherstellt monoton steigende x-Achse für fill_between\n", "# Hinweis: Stunden mit wenigen Messwerten (n klein) könnten den Mittelwert leicht\n", "# verzerren. Für den PoC wird keine Mindestschwelle gesetzt – n wird pro Stunde\n", "# mitgeloggt und kann bei Bedarf als Filterkriterium ergänzt werden.\n", "hourly_mean = (\n", " df_hist.groupby(\"hour\")[\"humidex\"]\n", " .agg([\"mean\", \"std\", \"count\"])\n", " .rename(columns={\"mean\": \"humidex_mean\", \"std\": \"humidex_std\", \"count\": \"n\"})\n", " .reset_index()\n", " .sort_values(\"hour\")\n", ")\n", "\n", "# Kühlste/wärmste Stunde für Annotation und Empfehlung\n", "coolest_hour = hourly_mean.loc[hourly_mean[\"humidex_mean\"].idxmin()]\n", "warmest_hour = hourly_mean.loc[hourly_mean[\"humidex_mean\"].idxmax()]\n", "# n_stations dynamisch: falls einzelne Stationen keine Daten lieferten,\n", "# stimmt das Label im Plot trotzdem (nicht fix \"6 Stationen\")\n", "n_stations = df_hist[\"stationName\"].nunique()\n", "\n", "print(f\"Kühlste Tagesstunde : {int(coolest_hour['hour']):02d}:00 Uhr (Ø Humidex {coolest_hour['humidex_mean']:.1f})\")\n", "print(f\"Wärmste Tagesstunde : {int(warmest_hour['hour']):02d}:00 Uhr (Ø Humidex {warmest_hour['humidex_mean']:.1f})\")\n", "print(f\"\\nDatengrundlage: {len(df_hist):,} valide Messwerte, {n_stations} Stationen, Sommer 2025 (Jul–Aug)\")\n", "print(f\"\\nMesswerte pro Stunde (n) – Übersicht:\")\n", "print(hourly_mean[[\"hour\", \"n\"]].to_string(index=False))" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "60ekbrC3TiSF", "outputId": "55ed25ed-f0c0-49ec-925a-95b0b4ab70f2" }, "execution_count": 38, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "Humidex-NaNs nach Berechnung (rh<=0 oder rh>100): 2\n", "Kühlste Tagesstunde : 06:00 Uhr (Ø Humidex 17.8)\n", "Wärmste Tagesstunde : 16:00 Uhr (Ø Humidex 28.2)\n", "\n", "Datengrundlage: 48,998 valide Messwerte, 6 Stationen, Sommer 2025 (Jul–Aug)\n", "\n", "Messwerte pro Stunde (n) – Übersicht:\n", " hour n\n", " 0 2085\n", " 1 2050\n", " 2 2056\n", " 3 2070\n", " 4 2063\n", " 5 2054\n", " 6 2022\n", " 7 2051\n", " 8 2028\n", " 9 2048\n", " 10 2021\n", " 11 1998\n", " 12 2025\n", " 13 2022\n", " 14 2023\n", " 15 2008\n", " 16 2047\n", " 17 2036\n", " 18 2020\n", " 19 2052\n", " 20 2044\n", " 21 2057\n", " 22 2058\n", " 23 2058\n" ] } ] }, { "cell_type": "code", "source": [ "import matplotlib.pyplot as plt\n", "\n", "fig, ax = plt.subplots(figsize=(12, 5))\n", "\n", "# Konfidenzband ±1 Std.-Abw. zeigt die Streuung zwischen Stationen und Tagen\n", "# alpha=0.15 damit die Kurve im Vordergrund klar lesbar bleibt\n", "# fillna(0): falls eine Stunde nur 1 Messwert hat, ist std=NaN → fill_between\n", "# würde Warnung werfen; 0 ist korrekt (kein Band bei n=1)\n", "ax.fill_between(\n", " hourly_mean[\"hour\"],\n", " hourly_mean[\"humidex_mean\"] - hourly_mean[\"humidex_std\"].fillna(0),\n", " hourly_mean[\"humidex_mean\"] + hourly_mean[\"humidex_std\"].fillna(0),\n", " alpha=0.15, color=\"steelblue\", label=\"±1 Std.-Abw.\"\n", ")\n", "\n", "# Mittlerer Humidex-Verlauf als Hauptkurve\n", "# marker=\"o\" macht einzelne Stundenwerte sichtbar (24 Punkte)\n", "ax.plot(\n", " hourly_mean[\"hour\"],\n", " hourly_mean[\"humidex_mean\"],\n", " color=\"steelblue\", linewidth=2.5, marker=\"o\", markersize=4,\n", " label=f\"Ø Humidex ({n_stations} Stationen)\"\n", ")\n", "\n", "# Kühlste Stunde als vertikale Linie + Annotation\n", "# xytext x-Position: min(cool_h+1.2, 20) verhindert dass Text bei späten Stunden\n", "# (z.B. 22/23 Uhr) aus dem Plotbereich läuft\n", "cool_h = int(coolest_hour[\"hour\"])\n", "cool_v = coolest_hour[\"humidex_mean\"]\n", "ax.axvline(cool_h, color=\"green\", linestyle=\"--\", linewidth=1.5, alpha=0.8)\n", "ax.annotate(\n", " f\"Kühlste Zeit\\n{cool_h:02d}:00 Uhr\\nHumidex {cool_v:.1f}\",\n", " xy=(cool_h, cool_v),\n", " xytext=(min(cool_h + 1.2, 20), cool_v + 1.5),\n", " fontsize=9, color=\"green\",\n", " arrowprops=dict(arrowstyle=\"->\", color=\"green\", lw=1.2)\n", ")\n", "\n", "# Humidex-Referenzlinien als Kontext für den Leser\n", "# Hinweis: diese Schwellen stammen aus der Humidex-Skala (Environment Canada),\n", "# nicht aus unserer Ampellogik (die auf relativem Comfort Score basiert)\n", "# x-Position der Labels an max(hour) gekoppelt → robust falls Stundenbereich\n", "# sich ändert (z.B. nur Tagesstunden analysiert)\n", "x_max = hourly_mean[\"hour\"].max()\n", "ax.axhline(27, color=\"orange\", linestyle=\":\", linewidth=1, alpha=0.7)\n", "ax.axhline(35, color=\"red\", linestyle=\":\", linewidth=1, alpha=0.7)\n", "ax.text(x_max - 0.3, 27.3, \"Gewisse Belastung (27)\", fontsize=7.5, color=\"orange\", ha=\"right\")\n", "ax.text(x_max - 0.3, 35.3, \"Starke Belastung (35)\", fontsize=7.5, color=\"red\", ha=\"right\")\n", "\n", "# set_xticks(range(0,24)) erzwingt alle 24 Stunden auf der x-Achse\n", "# ohne dies würden nur die im Datensatz vorhandenen Stunden erscheinen\n", "# Achsenbeschriftung \"Europe/Zurich\" statt \"UTC+2\": UTC+2 gilt nur im Sommer (CEST),\n", "# Europe/Zurich ist die korrekte IANA-Zeitzone inkl. DST-Handling\n", "ax.set_xlabel(\"Tageszeit (Lokalzeit Bern, Europe/Zurich)\", fontsize=11)\n", "ax.set_ylabel(\"Ø Humidex\", fontsize=11)\n", "ax.set_title(\n", " \"Humidex-Tagesverlauf – Sommer 2025 (Jul–Aug)\\n\"\n", " f\"Mittel über {n_stations} geografisch diversifizierte Stationen in Bern\",\n", " fontsize=12, fontweight=\"bold\"\n", ")\n", "ax.set_xticks(range(0, 24))\n", "ax.set_xticklabels([f\"{h:02d}:00\" for h in range(24)], rotation=45, ha=\"right\", fontsize=8)\n", "ax.legend(fontsize=9)\n", "ax.grid(True, alpha=0.3)\n", "plt.tight_layout()\n", "plt.show()\n", "\n", "# Abschliessende Empfehlung als klare Aussage\n", "# Hinweis: Im Produkt würde diese Information als einfacher Badge/Satz erscheinen\n", "# (\"Empfehlung: 06:00–08:00 Uhr – historisch kühlste Zeit\") – die Grafik dient\n", "# im PoC als methodische Untermauerung dieser Aussage, nicht als direkte UI\n", "print(f\"\\n Empfehlung: Outdoor-Training idealerweise um {cool_h:02d}:00 Uhr \"\n", " f\"(niedrigster Ø Humidex im Sommer 2025)\")" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 541 }, "id": "rUC1XmnDZSBe", "outputId": "2bc0cefb-13bd-4b2f-dde4-f4727ff70c19" }, "execution_count": 39, "outputs": [ { "output_type": "display_data", "data": { "text/plain": [ "
" ], "image/png": "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\n" }, "metadata": {} }, { "output_type": "stream", "name": "stdout", "text": [ "\n", " Empfehlung: Outdoor-Training idealerweise um 06:00 Uhr (niedrigster Ø Humidex im Sommer 2025)\n" ] } ] }, { "cell_type": "markdown", "source": [ "## 6. Zusammenfassung – HeatSafeBern Systemoutput\n", "Kompakte Übersicht der wichtigsten Ergebnisse aus RTD-Snapshot\n", "und historischer Analyse" ], "metadata": { "id": "tmn-aFYZaiaV" } }, { "cell_type": "code", "source": [ "from datetime import datetime\n", "\n", "# Ampel-Verteilung aus RTD-Snapshot (inkl. gray = Ausreisser/keine Daten)\n", "n_green = (df_rtd_bern[\"color\"] == \"green\").sum()\n", "n_orange = (df_rtd_bern[\"color\"] == \"orange\").sum()\n", "n_red = (df_rtd_bern[\"color\"] == \"red\").sum()\n", "n_gray = (df_rtd_bern[\"color\"] == \"gray\").sum()\n", "\n", "# Kühlste Station aus validen RTD-Stationen\n", "coolest_station = df_rtd_valid.loc[df_rtd_valid[\"comfort_score\"].idxmax()]\n", "\n", "# Beste Trainingszeit aus historischer Analyse\n", "best_hour = int(coolest_hour[\"hour\"])\n", "best_humidex = coolest_hour[\"humidex_mean\"]\n", "\n", "# stations_raw = Roh-API-Output vor jeglicher Bereinigung/Filterung\n", "# Guard verhindert NameError falls Zelle isoliert ausgeführt wird\n", "total_raw = len(stations_raw) if \"stations_raw\" in dir() else \"?\"\n", "\n", "print(\"=\" * 55)\n", "print(\" HEATSAFEBERN – ZUSAMMENFASSUNG\")\n", "print(f\" Stand: {datetime.now().strftime('%d.%m.%Y %H:%M')} Uhr\")\n", "print(\"=\" * 55)\n", "print()\n", "print(f\"Stationen analysiert : {len(df_rtd_bern)} (von {total_raw} total)\")\n", "print(f\"Geo-Filter : {GEO_RADIUS_KM} km Radius um Bern\")\n", "print()\n", "print(\"Ampel-Verteilung:\")\n", "print(f\" 🟢 Grün : {n_green:>2} Stationen\")\n", "print(f\" 🟡 Orange : {n_orange:>2} Stationen\")\n", "print(f\" 🔴 Rot : {n_red:>2} Stationen\")\n", "print(f\" ⚫ Grau : {n_gray:>2} Stationen (Ausreisser / keine Daten)\")\n", "print()\n", "print(f\"Kühlste Station : {coolest_station['name']}\")\n", "print(f\" Humidex : {coolest_station['humidex']:.1f}\")\n", "print(f\" Comfort Score : {coolest_station['comfort_score']:.1f}%\")\n", "print()\n", "# cool_route = Top-5 Stationen nach Comfort Score, geografisch via\n", "# Nearest-Neighbour sortiert – nicht nach Score-Ranking, sondern nach Nähe\n", "print(\"Route Top-5 Stationen (kühlste), nach Nähe sortiert (Luftlinie):\")\n", "for i, (_, s) in enumerate(cool_route.iterrows(), 1):\n", " print(f\" {i}. {s['name']} ({s['comfort_score']:.1f}%)\")\n", "print()\n", "print(f\"Beste Trainingszeit : {best_hour:02d}:00 Uhr\")\n", "print(f\" Ø Humidex {best_humidex:.1f} – basierend auf Sommer 2025,\")\n", "print(f\" {n_stations} Stationen, {len(df_hist):,} Messwerte\".replace(\",\", \"'\"))\n", "print()\n", "print(\"=\" * 55)" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "v28VxhFPawJP", "outputId": "7f8d70fb-d9f8-4f49-b40a-e82f240b50fd" }, "execution_count": 40, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "=======================================================\n", " HEATSAFEBERN – ZUSAMMENFASSUNG\n", " Stand: 13.03.2026 17:21 Uhr\n", "=======================================================\n", "\n", "Stationen analysiert : 31 (von 64 total)\n", "Geo-Filter : 10 km Radius um Bern\n", "\n", "Ampel-Verteilung:\n", " 🟢 Grün : 4 Stationen\n", " 🟡 Orange : 17 Stationen\n", " 🔴 Rot : 6 Stationen\n", " ⚫ Grau : 4 Stationen (Ausreisser / keine Daten)\n", "\n", "Kühlste Station : Liebefeld Turnierstrasse Laterne\n", " Humidex : 10.7\n", " Comfort Score : 100.0%\n", "\n", "Route Top-5 Stationen (kühlste), nach Nähe sortiert (Luftlinie):\n", " 1. Köniz Gartenstadt Laterne (100.0%)\n", " 2. Liebefeld Turnierstrasse Laterne (100.0%)\n", " 3. Zollikofen 3m (66.7%)\n", " 4. Ittigen Gantrischweg 1 Laterne (61.1%)\n", " 5. Gümligen Dentenberg Laterne (66.7%)\n", "\n", "Beste Trainingszeit : 06:00 Uhr\n", " Ø Humidex 17.8 – basierend auf Sommer 2025,\n", " 6 Stationen' 48'998 Messwerte\n", "\n", "=======================================================\n" ] } ] } ] }