31:T500,{ const container = yield htl.html`

`; const map = L.map(container, { scrollWheelZoom: false, minZoom: 6, maxZoom: 10 }); for (const a of ["LFTW", "LFBC", "LFBD"]) { const p = L.marker([airports[a].latitude, airports[a].longitude]) .addTo(map) .bindTooltip(airports[a].name, { permanent: a === "LFTW" }); } L.geoJSON(aoi).addTo(map); const bbox = L.geoJSON(south_of_france); L.geoJSON( bordeaux_flights.map((f) => f.feature({ category: (f) => f.callsign.slice(0, 4) }) ), { style: function (feature) { switch (feature.properties.category) { case "BENG": return { color: d3.schemeTableau10[0] }; case "DRAG": return { color: d3.schemeTableau10[1] }; case "MILA": return { color: d3.schemeTableau10[2] }; } } } ).addTo(map); map.fitBounds(bbox.getBounds(), { maxZoom: 9 }); map.setMaxBounds(L.latLngBounds(L.latLng(40, -7), L.latLng(50, 15))); L.tileLayer( "https://{s}.basemaps.cartocdn.com/rastertiles/voyager_labels_under/{z}/{x}/{y}.png", { attribution: "© OpenStreetMap contributors" } ).addTo(map); }32:T8c6,fire_map = (date) => { const map = Plot.plot({ width: map_width, margin: 10, //caption: `Situation on ${date}`, projection: { type: "transverse-mercator", domain: aoi }, color: { scheme: "reds" }, marks: [ Plot.geo(land, { clip: "frame", stroke: "#bab0ac", fill: "#eeeeee" }), Plot.density(fire, { y: "latitude", x: "longitude", weight: "brightness", fill: "density", filter: (x) => x.acq_date === date, bandwidth: 5, opacity: 0.7, scheme: "reds" }), Plot.dot(fire, { y: "latitude", x: "longitude", fill: "black", filter: (x) => x.acq_date === date, r: 1 }), Plot.dot(Object.values(airports), { x: "longitude", y: "latitude", filter: (f) => ["LFCH", "LFBC", "LFBD"].includes(f.icao) }), Plot.text(Object.values(airports), { x: "longitude", y: "latitude", text: "name", filter: (f) => ["LFCH", "LFBC", "LFBD"].includes(f.icao), height: (35 * map_width) / 700, fontSize: (20 * map_width) / 960, dy: -5, dx: -20, textAnchor: "end" }), Plot.geo( bordeaux_flights .filter( (f) => f.start.toISOString().slice(0, 10) === date.slice(0, 10) ) .map((f) => f.feature({ category: (f) => f.callsign.slice(0, 4) })), { stroke: function (feature) { switch (feature.properties.category) { case "BENG": return d3.schemeTableau10[0]; case "DRAG": return d3.schemeTableau10[1]; case "MILA": return d3.schemeTableau10[2]; case "PELI": return d3.schemeTableau10[3]; } }, strokeOpacity: 0.5, strokeWidth: 1.2 } ) ] }); return htl.html`

${date}

${Plot.legend({ color: { domain: ["BENGALE", "DRAGON", "MILAN", "PELICAN"], range: d3.schemeTableau10 }, columns: 4, width: 350, style: { fontSize: "8pt", margin: "2ex" } })}${map}`; }33:T482,On the following map, we mapped intervals when `PELICAN` aircraft have been seen by receivers in the OpenSky Network: - in Toulouse area, the receiver is located on the scientific campus to the South of the city; - in Marseille area, one receiver is located around Marseille, and the other on Salon-de-Provence Air Base; - surprisingly, no receiver saw any of these aircraft in Bordeaux area, but the quality of the receivers there and the low altitude flown by Canadair can explain that. When a receiver sees an aircraft, we have no information about its location, but we can keep in mind that the lower it is, the closer to the receiver. Smaller dots are associated to lower minimum altitudes recorded: aircraft are (sometimes?) detected in Marseille area when they land or depart from their base at Nîmes-Garons airport. In Toulouse, they fly usually around 6000 ft (approx. 2000 m), which suggests they flew very close to the city when they were detected. An interesting pattern is also visible for the `PELIC34` aircraft which seems to have flown back to its base in Nîmes on July 17th evening, before leaving for Bordeaux again on the next day.34:[{"id":"7bb920ab998710c0e94aaa8b61292f960df75e11e8a869e91ae6d7279b88cbc8d809be3d0d2a64e6402a439bd1e1bed902b3a38c8ea396e6a8e4d8653f80e5e4","href":"https://static.observableusercontent.com/files/7bb920ab998710c0e94aaa8b61292f960df75e11e8a869e91ae6d7279b88cbc8d809be3d0d2a64e6402a439bd1e1bed902b3a38c8ea396e6a8e4d8653f80e5e4","name":"trajectories.json.gz","type":"application/gzip","lastModified":1679666642847.823,"size":11811727},{"id":"721d38838fb8477fefd7f85ade3021b8237228267cb19aed74b6bdbdcfdae930e01870c3a5200cd5949f763aede2e1a6046718918ba8b194154070d38f419039","href":"https://static.observableusercontent.com/files/721d38838fb8477fefd7f85ade3021b8237228267cb19aed74b6bdbdcfdae930e01870c3a5200cd5949f763aede2e1a6046718918ba8b194154070d38f419039","name":"fire_nrt_J1V-C2_337798.json","type":"application/json","lastModified":1679564864611.054,"size":1256970}] 2f:{"title":"Firefighting activities of civil security aircraft in France","cells":[{"id":0,"value":"# Firefighting activities of civil security aircraft in France\n\n*In this page, we replay the firefighting aircraft movements during extreme fire events in mid-July 2022. Data is based on transponder data collected by the [OpenSky Network](www.opensky-network.org) and satellite images published by NASA.*","pinned":false,"hidden":false,"mode":"md"},{"id":117,"value":"Civil security aircraft are used for missions, such as firefighting, life and rescue, and medical transportation. In France, four different types of aircraft are used for firefighting activities. We can identify them by the following callsigns:\n- `PELICAN` operated by Bombardier Canadair CL-415 aircraft;\n- `MILAN` operated by De Havilland Dash-8 Q400 MR aircraft;\n- `BENGALE` operated by Beechcraft B200 aircraft;\n- `DRAGON` operated by Airbus helicopters EC145 and H145.\n\n## Fleet description","pinned":false,"hidden":false,"mode":"md"},{"id":1573,"value":"

French Canadair aircraft, photo by Joachim Bertrand, CC BY-NC-ND 2.0 license (source)

\n","pinned":false,"hidden":false,"mode":"html"},{"id":124,"value":"**Canadair** (`PELICAN`) are the iconic aircraft used in the fight against fires. Currently, the French Civil Security owns 12 aircraft (CL-415), all based at Nîmes-Garons airport (`LFTW`). They can fly up to 350 km/h and have a water dropping capacity of 6,000 litres. The Canadair is an amphibious aircraft that can land on water and take off again. It has the ability to refuel in flight, which is a significant time saver compared to other aircraft that are required to return to an airbase for refuelling. To do this, it must skim the surface of a body of water, which can be a lake, a river, or the sea. This operation is called “scooping” and takes only 9 to 12 seconds to fill its two tanks of 3,000 litres each, collecting 6 tons of water.","pinned":false,"hidden":false,"mode":"md"},{"id":1144,"value":"\n

Sécurité Civile Dash water bomber aircraft, image by Stéphane Hernaut, CC-BY-SA-3.0 license (source)

\n","pinned":false,"hidden":false,"mode":"html"},{"id":129,"value":"The French Civil Security is also equipped with **Dash-8 Q400** (`MILAN`) as **water bombers**. The Dash has a tank that can hold 10,000 litres of water or fire retardant (ammonium polyphosphate combined with iron oxide). Unlike the Canadair, it must land on a runway to fill its tank, but compensates for this constraint with a higher speed (700 km/h). Versatile, the Dash aircraft can also transport passengers or equipment.\nIn the same way as Canadairs, Dash-8 can rely on the 22 French \"Pélicandromes\" deployed on aerodromes to quickly refuel firefighting aircraft.\n\n","pinned":false,"hidden":false,"mode":"md"},{"id":1153,"value":"

`F-ZBFJ` Sécurité Civile Beechcraft B200 King Air, copyright Joseph Debattista (source)

","pinned":false,"hidden":false,"mode":"html"},{"id":137,"value":"\nUnlike the Canadair and Dash 8, the **Beechcraft B200** (`BENGALE`) is not a water bomber aircraft. It is primarily used for **reconnaissance and surveillance** of areas, **coordination** of aerial operations and transportation of people or equipment. The B200 can fly up to 500 km/h.","pinned":false,"hidden":false,"mode":"md"},{"id":312,"value":"## Data analysis over a week in July 2022","pinned":false,"hidden":false,"mode":"md"},{"id":1430,"value":"The summer of 2022 was devastating for wildfires in France, particularly in the department of Gironde (Bordeaux). The first fire started on **Tuesday, July 12th**, around 3 pm in La Teste-de-Buch. One hour later, a second fire broke out in the town of Landiras. These two unprecedented fires caused the evacuation of 36,750 residents and ravaged 20,800 hectares of forest for nearly two weeks.\n\nThe figures below illustrates the day-by-day progression of these fires during the peak of intensity. Trajectory data come from ADS-B positions collected and shared by the [OpenSky Network](https://opensky-network.org/). Data related to fire positions is provided by the NASA [FIRMS](https://firms.modaps.eosdis.nasa.gov/) (Fire Information for Resource Management System) service. Fire data is provided as point coordinates associated with a brightness value that we plot together with a density map below:\n","pinned":false,"hidden":false,"mode":"md"},{"id":1840,"value":"Plot.plot({\n width,\n //margin: 10,\n caption: `Evolution of fire outbreaks according to the NASA FIRMS service`,\n projection: {\n type: \"transverse-mercator\",\n domain: aoi\n },\n fx: {\n tickFormat: (d) => d3.utcFormat(\"%B %d\")(d3.utcParse(\"%Y-%m-%d\")(d)),\n padding: 0\n },\n facet: { data: fire, x: (d) => d.acq_date },\n color: { scheme: \"reds\" },\n marks: [\n Plot.geo(land, {\n clip: \"frame\",\n stroke: \"#bab0ac\",\n fill: \"#eeeeee\"\n }),\n Plot.density(fire, {\n y: \"latitude\",\n x: \"longitude\",\n weight: \"brightness\",\n fill: \"density\",\n bandwidth: 5,\n opacity: 0.7,\n scheme: \"reds\"\n }),\n Plot.dot(fire, {\n y: \"latitude\",\n x: \"longitude\",\n fill: \"black\",\n r: 1\n })\n ]\n})","pinned":false,"hidden":false,"mode":"ojs"},{"id":1838,"value":"\nAll aircraft are based in Nîmes airport near Marseilles, but the local base during that week was Bordeaux airport, with some layover occurrences found at Cazaux Air Base for `BENGALE` aircraft (in blue). Also, only few `DRAGON` helicopters flew to the North of the area, but didn't seem to take an active role in the Gironde fires.","pinned":false,"hidden":false,"mode":"md"},{"id":689,"value":"$31","pinned":false,"hidden":false,"mode":"ojs"},{"id":99,"value":"Plot.plot({\n color: { scheme: \"tableau10\", legend: true },\n caption: \"Recorded movements by Sécurité Civile's aircraft\",\n marks: [\n Plot.barX(stats, {\n x: \"start\",\n x2: \"stop\",\n y: \"callsign\",\n fill: (f) => f.callsign.slice(0, 4),\n tip: \"flight_id\"\n }),\n Plot.ruleX(\n d3.utcDay.range(\n new Date(\"2022-07-12 00:00Z\"),\n new Date(\"2022-07-20 00:00Z\")\n ),\n { strokeOpacity: 0.2 }\n )\n ],\n x: { type: \"utc\", round: true, nice: d3.utcDay },\n y: { inset: 10, label: null },\n marginLeft: 80,\n width,\n height: 300\n})","pinned":false,"hidden":false,"mode":"ojs"},{"id":1982,"value":"fire_map(\"2022-07-13\")","pinned":true,"hidden":false,"mode":"ojs"},{"id":1441,"value":"fire_map(\"2022-07-14\")","pinned":false,"hidden":false,"mode":"ojs"},{"id":1455,"value":"On July 14th, 2,900 hectares had already burned in La Teste-de-Buch, in the Arcachon district, and the situation remained unfavorable. In Landiras, in the Langon district, the situation was very unfavorable with several reactivations of the fire. Three other fires were controlled in the north of Blayais. Canadair planes continued to drop water until nightfall.","pinned":false,"hidden":false,"mode":"md"},{"id":1460,"value":"fire_map(\"2022-07-15\")","pinned":false,"hidden":false,"mode":"ojs"},{"id":1452,"value":"\nOn July 15th, in La Teste-de-Buch, the fire reached Lake Cazaux and progressed towards the south and the left flank, towards the homes along the road leading to the lake. 10,000 people were evacuated. In Landiras, 4,200 hectares had already burned, and the fire was progressing. Firefighters, supported by military units and aerial means, were more than ever engaged to contain these large-scale fires. Nearly **1,000 firefighters, 3 Canadair planes, and 1 Dash plane** were mobilized.\n","pinned":false,"hidden":false,"mode":"md"},{"id":1471,"value":"fire_map(\"2022-07-16\")","pinned":false,"hidden":false,"mode":"ojs"},{"id":1464,"value":"On July 16th, despite the significant ground and air means engaged (1,200 firefighters, 3 Canadair planes, and 1 Dash plane), the fires were not yet contained. The situation was stabilized in La Teste-de-Buch, but the situation was very unfavorable in Landiras (6,500 hectares burned), with winds pushing the fires towards the southwest.","pinned":false,"hidden":false,"mode":"md"},{"id":1473,"value":"fire_map(\"2022-07-17\")","pinned":false,"hidden":false,"mode":"ojs"},{"id":1466,"value":"On July 17th, more than **1,200 firefighters, 4 Canadair planes, and 2 Dash planes** were still mobilized. In La Teste-de-Buch, the fire threatened the campsites of the Dune of Pilat again due to numerous reactivations, supported by strong gusts of wind. In Landiras, the situation was contained on the left front flank of the fire. However, the situation remained unfavorable on the right front flank.","pinned":false,"hidden":false,"mode":"md"},{"id":1475,"value":"fire_map(\"2022-07-18\")","pinned":false,"hidden":false,"mode":"ojs"},{"id":1468,"value":"\nOn July 18th, nearly 1,700 firefighters from all over France, supported by significant aerial means (**6 Canadair planes and 3 Dash planes**), were fighting these two unprecedented fires. This day was marked by very unfavorable weather conditions, which contributed to the progression of the fire and led to new evacuations on both fire sites. 4,700 hectares of forest burned in La Teste-de-Buch, with the fire progressing to the north and south. 12,000 hectares burned in Landiras, with the fire progressing on different fronts despite the creation of firebreaks. In the late afternoon, a new fire with an area of 70 hectares broke out in Vensac, in the Médoc. The situation remained very unfavorable and uncertain.","pinned":false,"hidden":false,"mode":"md"},{"id":258,"value":"$32","pinned":false,"hidden":false,"mode":"ojs"},{"id":1267,"value":"## So where are the Canadair?\n\nThe maps above show no records of the Canadair/`PELICAN` trajectories. The dataset contains however records sent by those aircraft, but no positional information.\n\nEach aircraft is identified by an hexadecimal number (the ICAO identifier) which uniquely identifies a transponder. Transponders can then be mapped to a register of tail numbers of aircraft. Recent transponders are ADS-B compliant, a technology where aircraft regularly broadcast their position. Older transponders only reply to requests by secondary surveillance radars.\n\nIt seems that Canadair aircraft owned by French Sécurité Civile fall into this category. We find records associated with their transponder identifiers, mapped with `PELICAN` callsigns for legibility reasons. The only information we have is the transponder identifier, altitude data points (in feet), and which receiver of the OpenSky Network saw the aircraft.","pinned":false,"hidden":false,"mode":"md"},{"id":1758,"value":"Inputs.table(flights.find((f) => f.callsign.slice(0, 4) === \"PELI\").data, {\n columns: [\n \"timestamp\",\n \"icao24\",\n \"callsign\",\n \"latitude\",\n \"longitude\",\n \"altitude\",\n \"groundspeed\",\n \"track\",\n \"vertical_rate\",\n \"serials\"\n ],\n width: { timestamp: \"20%\" },\n sort: \"timestamp\",\n layout: \"auto\",\n format: {\n icao24: (elt) => html`

${elt}`,\n    callsign: (elt) => html`${elt}`,\n    timestamp: (elt) => html`${d3.utcFormat(\"%Y-%m-%d %H:%M:%S\")(elt)}`\n  }\n})","pinned":false,"hidden":false,"mode":"ojs"},{"id":1765,"value":"$33","pinned":false,"hidden":false,"mode":"md"},{"id":1650,"value":"Plot.plot({\n  marks: [\n    Plot.dot(\n      pelican\n        .filter((x) => x.serials !== 91739)\n        .map(\n          (x) =>\n            new Object({\n              ...x,\n              area:\n                x.serials === 1433801924 ? \"Toulouse area\" : \"Marseille area\"\n            })\n        ),\n      {\n        x: \"start\",\n        y: \"callsign\",\n        fill: \"area\",\n        r: \"altitude\"\n      }\n    ),\n    Plot.ruleX(\n      d3.utcDay.range(\n        new Date(\"2022-07-12 00:00Z\"),\n        new Date(\"2022-07-20 00:00Z\")\n      ),\n      { strokeOpacity: 0.2 }\n    )\n  ],\n  x: { type: \"utc\" },\n  y: { label: null },\n  color: { legend: true, type: \"categorical\" },\n  r: { legend: true },\n  marginLeft: 100,\n  width,\n  height: 300\n})","pinned":false,"hidden":false,"mode":"ojs"},{"id":267,"value":"## Boring stuff","pinned":false,"hidden":false,"mode":"md"},{"id":1798,"value":"aoi = new Object({\n  // turf = require(\"@turf/turf@6.5.0/turf.min.js\")\n  // turf.rewind(turf.bboxPolygon([-2, 44, 0, 45.6]), { reverse: true })\n  type: \"Feature\",\n  //rewind: true,\n  bbox: [-2, 44, 0, 45.6],\n  properties: {},\n  geometry: {\n    type: \"Polygon\",\n    coordinates: [\n      [\n        [-2, 44],\n        [-2, 45.6],\n        [0, 45.6],\n        [0, 44],\n        [-2, 44]\n      ]\n    ]\n  }\n})","pinned":false,"hidden":false,"mode":"ojs"},{"id":1805,"value":"south_of_france = new Object({\n  type: \"Feature\",\n  bbox: [-1.5, 43, 7, 45.5],\n  geometry: {\n    type: \"Polygon\",\n    coordinates: [\n      [\n        [-1.5, 43],\n        [-1.5, 45.5],\n        [7, 45.5],\n        [7, 43],\n        [-1.5, 43]\n      ]\n    ]\n  }\n})","pinned":false,"hidden":false,"mode":"ojs"},{"id":801,"value":"map_width = 550","pinned":true,"hidden":false,"mode":"ojs"},{"id":39,"value":"flights = d3.sort(\n  await Traffic.fromArrayBuffer(\n    await FileAttachment(\"trajectories.json.gz\").arrayBuffer()\n  ),\n  // sort flights by duration (descending)\n  (x) => -x.duration\n)","pinned":false,"hidden":false,"mode":"ojs"},{"id":1217,"value":"land = d3.json(\n  \"https://www.data.gouv.fr/fr/datasets/r/d993e112-848f-4446-b93b-0a9d5997c4a4\"\n)","pinned":false,"hidden":false,"mode":"ojs"},{"id":841,"value":"fire = FileAttachment(\"fire_nrt_J1V-C2_337798.json\")\n  .json()\n  .then((entries) => entries.filter((x) => x.acq_date !== \"2022-07-12\"))","pinned":false,"hidden":false,"mode":"ojs"},{"id":1598,"value":"bordeaux_flights = flights.filter((f) => f.intersects(aoi))","pinned":false,"hidden":false,"mode":"ojs"},{"id":86,"value":"stats = bordeaux_flights.map((f) =>\n  f.rollup({\n    start: (f) => f.start,\n    stop: (f) => f.stop,\n    callsign: \"callsign\",\n    icao24: \"icao24\",\n    flight_id: op.max(\"flight_id\"),\n    altitude_min: op.min(\"altitude\"),\n    serials: (f) =>\n      f\n        .entries()\n        .filter((elt) => elt.serials[0] !== undefined)\n        .reduce((cumul, elt) => cumul.add(...elt.serials), new Set())\n  })\n)","pinned":false,"hidden":false,"mode":"ojs"},{"id":1337,"value":"pelican = [].concat(\n  ...flights\n    .filter((f) => f.callsign.slice(0, 4) === \"PELI\")\n    .map(\n      (flight) =>\n        new Flight(\n          aq.from(\n            [].concat(\n              ...flight\n                .entries()\n                .map((x) =>\n                  x.serials.map((s) => new Object({ ...x, serials: s }))\n                )\n            )\n          )\n        )\n    )\n    .map((f) =>\n      f.data\n        .groupby(\"serials\")\n        .rollup({\n          start: op.min(\"timestamp\"),\n          stop: op.max(\"timestamp\"),\n          callsign: op.max(\"callsign\"),\n          altitude: op.min(\"altitude\")\n        })\n        .objects()\n    )\n)","pinned":false,"hidden":false,"mode":"ojs"},{"id":6,"value":"import { Flight, Traffic } from \"@xoolive/traffic-js\"","pinned":false,"hidden":false,"mode":"ojs"},{"id":88,"value":"import { aq, op } from \"@uwdata/arquero\"","pinned":false,"hidden":false,"mode":"ojs"},{"id":759,"value":"import { airports } from \"@openaviation/airports\"","pinned":false,"hidden":false,"mode":"ojs"}],"files":"$W34","stdlib":"1","titling":"auto"}