Datasets
The list of PFAS Known users was established manually by the FPP project using diverse sources. Methodological notes from the FPP project:
List of potential PFAS producers and users provided by Ian Cousins
Ian Cousins (Stockholm university) shared a research spreadsheet listing 108 potential producers and users in the world, including 25 in Europe.This list was the starting point to create a “Known PFAS users” category.
The list by INDITEX
The fashion group Inditex (Zara, Pull&Bear…) has developed a document named “The List by INDITEX” which classifies a dozen commercially available chemical families used for textile and leather manufacturing processes (such as dyestuffsor water & oils repellents), according to their level of compliance with Inditex’s own “product health standard” and “manufacturing restricted substance list”. Products including fluorocarbons are tagged with #7. We looked up fluorocarbon-containing products and included their manufacturers in our “PFAS know users” category.
Manufacturers of aqueous film-forming foams (AFFF)
We identified and located 23 type B (AFFF) firefighting foam manufacturing facilities using lists and information from the ECHA restriction report on firefighting foams, KEMI 2015, and IPEN 2018.
Google Maps
We entered the keyword “PTFE” in Google Maps and identified dozens of PTFE (Teflon) downstream users in Europe.
Serendipitous findings
Luck is part of research. Some locations were found by chance. For example, we were able to identify a few PFAS users while hovering over industrial zones and chemical parks in Google maps.
List of potential PFAS producers and users provided by Ian Cousins
Ian Cousins (Stockholm university) shared a research spreadsheet listing 108 potential producers and users in the world, including 25 in Europe.This list was the starting point to create a “Known PFAS users” category.
The list by INDITEX
The fashion group Inditex (Zara, Pull&Bear…) has developed a document named “The List by INDITEX” which classifies a dozen commercially available chemical families used for textile and leather manufacturing processes (such as dyestuffsor water & oils repellents), according to their level of compliance with Inditex’s own “product health standard” and “manufacturing restricted substance list”. Products including fluorocarbons are tagged with #7. We looked up fluorocarbon-containing products and included their manufacturers in our “PFAS know users” category.
Manufacturers of aqueous film-forming foams (AFFF)
We identified and located 23 type B (AFFF) firefighting foam manufacturing facilities using lists and information from the ECHA restriction report on firefighting foams, KEMI 2015, and IPEN 2018.
Google Maps
We entered the keyword “PTFE” in Google Maps and identified dozens of PTFE (Teflon) downstream users in Europe.
Serendipitous findings
Luck is part of research. Some locations were found by chance. For example, we were able to identify a few PFAS users while hovering over industrial zones and chemical parks in Google maps.
2024-07-03
2025-05-06
static
231
231
The European Pollutant Release and Transfer Register (EPRTR) lists most of the industrial sites in Europe. The FPP project used the list of activities established by Salvatore et al. (https://doi.org/10.1021/acs.estlett.2c00502) and transposed these into the NACE system to filter the sites of the EPRTR. The list of NACE codes used is in a csv file here: https://gitlab.com/pfas-data-hub/pdh-data/-/tree/main/data/raw/1__EPRTR/eprtr_nace.csv
EPRTR
()
2024-07-03
2025-05-06
dynamic
4982
4903
The EEA provides a list of 21,488 WWTP in Europe. This data was filtered according to the methodology from Salvatore et al. (2022) This means it was filtered “to include only “major” WWTPs, which have a design flow of ≥1 million gallons per day or an industrial pretreatment program.” We filtered the dataset to include WWTPs which treat the equivalent quantity of water, i.e. a minimum 3,700 cubic metres per day. We used the field uwwWasteWaterTreated, but this is not very satisfactory since it is more or less filled in different countries. This excludes almost all italian WWTPs for example. If you know of a more clever way to proceed, please reach out to us. For France this field (uwwWasteWaterTreated) clearly has daily values instead of yearly ones, so we corrected the threshold accordingly.
2024-07-03
2025-05-06
dynamic
3895
3893
This data was collected manually during the FPP, using OSINT search. In some cases, a URL and / or a comment are present.
()
2024-07-03
2025-05-06
static
542
542
This list was manually collided by the FPP project. The sources are website of paper industry national trade association, as some of them have a page listing the paper mills in their country.
()
2024-07-03
2025-05-06
static
975
975
A dataset of only one row, one training center in Italy.
2024-07-03
2025-05-06
static
1
1
Training centers for municipal fire fighting units. In such centers AFFF are or were commonly used.
2024-07-03
2025-05-15
dynamic
247
247
Training center for fire-fighting in France. The list was established during the FPP.
2024-07-03
2025-05-06
static
5
5
Flanders fire fighting training centers. Obtained from a map that seems to be not available anymore, but the csv file with the locations is still online. The locations are given as polygons, we calculate the centroid of the polygons to have single coordinates
2024-07-03
2025-05-15
dynamic
814
814
In 2018, the European Commission’s Joint Research Centre collected emission data from over a hundred textile facilities across Europe. The FPP filed a "freedom of information" request and obtained PFAS emission data for the 15 facilities declaring PFAS discharges.
2024-07-03
2025-05-06
static
9
9
3
100.0%
9.0
6.0
The list of european facilities producing PFAS was established during the FPP, using OSINT techniques. A first list of 25 facilities was established, then each company was asked to comment before publication. The answers led to remove 5 locations from the list. From the remaining 20 facilities, 3 were no longer in use as of 2023.
We updated the list on May 19th 2025, adding a comment on the status of each plant after a search in grey literature. As far as we know this list was not established by anyone else, if you think there is a mistake or missing facilities please contact us.
Manual compilation during the FPP project
()
2024-07-03
2025-05-20
static
20
20
This dataset shows the data available on the PFAS map of the Flanders DOV.
The following layers are donwloaded:
pfas:pfas_analyseresultaten,
pfas:lantis_bodem_metingen,
pfas:lantis_gw_metingen_publiek,
pfas:lucht_gas_metingen,
pfas:lucht_zwevendstof_metingen,
waterbodems:pfas_meetpunten_fcs,
pfas:pfas_oppwater,
pfas:pfas_biota,
pfas:pfas_analyseresultaten,
pfas:lantis_bodem_metingen,
pfas:lantis_gw_metingen_publiek,
pfas:lucht_gas_metingen,
pfas:lucht_zwevendstof_metingen,
waterbodems:pfas_meetpunten_fcs,
pfas:pfas_oppwater,
pfas:pfas_biota,
2024-07-03
2025-05-17
dynamic
146208
19780
54
28.38%
30595.0
21433.0
2024-07-03
2025-05-06
static
983
402
22
82.1%
487.0
358.0
2024-07-03
2025-05-06
static
714
57
10
100.0%
494.0
52.0
Danske Region Sjælland
()
2024-07-03
2025-05-06
static
976
518
29
35.35%
190.0
53.0
Danske Region Midtjyllands
()
2024-07-03
2025-05-06
static
2744
974
33
62.97%
672.0
190.0
Danske Region Syddanmark
()
2024-07-03
2025-05-06
static
2952
1577
38
38.55%
652.0
223.0
Danske Region Nordjylland
()
2024-07-03
2025-05-06
static
1853
978
22
39.4%
436.0
141.0
2024-07-03
2025-05-06
static
225
90
22
76.89%
173.0
167.0
In 2009, the French Agency for Food, Environmental and Occupational Health & Safety (Anses) conducted a monitoring
campaign of PFAS in drinking water. The locations were not made public. The FPP project obtained the full dataset
with de-anonymised locations through a "freedom of information" request.
2024-07-03
2025-05-06
static
450
0
10
39.11%
94.0
11.0
Schleswig-Holstein Landesamt für Umwelt (LfU)
()
2024-07-03
2025-05-06
static
28
24
8
100.0%
7.0
0.0
Rheinland-Pfalz Ministerium für Klimaschutz, Umwelt, Energie und Mobilität (MKUEM)
()
2024-07-03
2025-05-06
static
46
38
8
100.0%
0.0
0.0
Data from the Hessen authorities, communicated to the FPP. The raw excel contains a tab called Eluat, but I am not sure which matrix to consider, so leaving them out.
Hessisches Landesamt für Naturschutz, Umwelt und Geologie
()
2024-07-03
2025-05-06
static
434
0
17
100.0%
434.0
434.0
Bremen Die Senatorin für Klimaschutz, Umwelt, Mobilität, Stadtentwicklung und Wohnungsbau
()
2024-07-03
2025-05-06
static
200
14
19
100.0%
74.0
10.0
Removed on 21/02/2025 since it has neither city nor lat/lon
2024-07-03
2025-05-06
static
0
0
There is no difference in the data between surface water (river and lakes) and seawater, which is not great.
2024-07-03
2025-05-20
dynamic
214
202
17
100.0%
83.0
42.0
Datasets 47 and 48 come from www.elwasweb.nrw.de. Dataset 47 is Stehendes Gewässer (static surface water) Dataset 48 is Fließgewässer/Kanäle (running surface water)
To obtain or refresh the raw data:
go on https://www.elwasweb.nrw.de/elwas-web/index.xhtml Chose "Daten" > Oberflächengewässer > Auswertungen > Messstellen Chemie und Biologie Fachbereich: Chemie Auswertung: Einzelwerte Chemie, dann start
Messstellentyp: Fließgewässer (for dataset 48) Pick the widest time range possible Pick all Matrix Stoffgruppe: PFC, then Auswählen (this select all PFAS params) Pick all stations (Messstellen) click "Suchen"
you should see at least 26108 lines for fließgewässer (that was on 18/10/2024, the number should only go up), 1183 for stehende Gewässer
----- There is another CSV describing the params here: https://www.opengeodata.nrw.de/produkte/umwelt_klima/wasser/stoffliste/ We use it to add the cas id to the param names, so that they are recognised automatically during normalise.
2024-10-18
2025-05-21
dynamic
347
53
17
77.81%
249.0
89.0
Datasets 47 and 48 come from www.elwasweb.nrw.de. Dataset 47 is Stehendes Gewässer (static surface water) Dataset 48 is Fließgewässer/Kanäle (running surface water)
To obtain or refresh the raw data:
go on https://www.elwasweb.nrw.de/elwas-web/index.xhtml Chose "Daten" > Oberflächengewässer > Auswertungen > Messstellen Chemie und Biologie Fachbereich: Chemie Auswertung: Einzelwerte Chemie, dann start
Messstellentyp: Fließgewässer (for dataset 48) Pick the widest time range possible Pick all Matrix Stoffgruppe: PFC, then Auswählen (this select all PFAS params) Pick all stations (Messstellen) click "Suchen"
you should see at least 26108 lines for fließgewässer (that was on 18/10/2024, the number should only go up), 1183 for stehende Gewässer
----- There is another CSV describing the params here: https://www.opengeodata.nrw.de/produkte/umwelt_klima/wasser/stoffliste/ We use it to add the cas id to the param names, so that they are recognised automatically during normalise.
2024-10-18
2025-05-21
dynamic
8299
558
18
66.73%
4817.0
1838.0
2024-10-21
2025-05-06
static
64
30
18
34.38%
22.0
22.0
The data was obtained by press request during the Forever Pollution Project Since they are all samplings in the output water of the Zentrale Abwasserbehandlungsanlage (ZARA) of the Gendorf Chemie Park, the same lat / lon was used for all points (loc was determined using Google Maps and a photo of the company website)
The data was received as a printed report during FPP, then manually typed in a spreadsheet. The number format of the values is not the same on all lines, likely typing errors. I'm waiting for a image of the printed data to come back to the source and correct the data. (20/11/2024)
2024-10-21
2025-05-06
static
0
0
Istituto di Ricerca sulle Acque – CNR
()
2024-07-03
2025-05-06
static
9
9
10
100.0%
9.0
9.0
This data is from the ARPA of the Veneto Region. The file is periodically updated, but always contains all the
historical data.
Some columns are in microgram/l, we manually convert them / merge them with the ng/l columns.
Some substances were ignored because I couldn't find the corresponding CAS-ID:
Cl-PFPECA (0,1) ;
Cl-PFPECA (0,2) ;
Cl-PFPECA (0,3) ;
Cl-PFPECA (0,4) ;
Cl-PFPECA (1,0) ;
Cl-PFPECA (1,1) ;
Cl-PFPECA (2,0) ;
Cl-PFPECA (2,1) ;
Cl-PFPECA (3,0) ;
Cl-PFPECA (4,0)
ARPA Veneto
()
2024-07-03
2025-05-06
static
17265
2134
33
70.92%
9208.0
5552.0
This dataset has an api in the Socrata format, so we use a dedicated python API that makes everything very simple. The data contains data about all substances, so we filter on the CAS_IDS that we have on our list. Strangely the PFNS is listed with cas_id 474511-07-4, which seems to be a mistake. We corrected it to 68259-12-1.
2024-07-03
2025-05-15
dynamic
4169
260
23
5.18%
54.0
6.0
Note on data treatment: in the raw file some lines had interted lat / lon. We took the min of both values as the lon, the max as the lat to correct this.
ARPA Sardegna
()
2024-07-03
2025-05-06
static
175
87
1
50.86%
25.0
25.0
2024-07-03
2025-05-06
static
69
23
17
88.41%
9.0
0.0
2024-07-03
2025-05-06
static
622
94
2
83.92%
110.0
107.0
2024-07-03
2025-05-06
static
934
19
17
7.71%
17.0
1.0
2024-07-03
2025-05-06
static
15
8
2
66.67%
10.0
10.0
2024-07-03
2025-05-06
static
16
13
25
18.75%
2.0
1.0
2024-07-03
2025-05-06
static
6381
4434
41
54.18%
3457.0
3399.0
Sistema Nacional de Informação de Recursos Hídricos (SNIRH) Departamento de Recursos Hídricos da Agência Portuguesa do Ambiente
()
2024-07-03
2025-05-06
static
31
22
1
0.0%
0.0
0.0
Sistema Nacional de Informação de Recursos Hídricos (SNIRH) Departamento de Recursos Hídricos da Agência Portuguesa do Ambiente
()
2024-07-03
2025-05-06
static
11
8
1
100.0%
11.0
11.0
2024-07-03
2025-05-06
static
1
1
29
100.0%
0.0
0.0
2024-07-03
2025-05-06
static
49
49
11
42.86%
11.0
1.0
Water quality monitoring data from the UK Environment Agency. The data is updated from time to time. Some points have name of locations, some are marked redacted, in which case the localisation is rounded (easting and northings rounded before lat/lon are calculated).
2024-07-03
2025-05-21
dynamic
10313
1269
12
100.0%
7822.0
2211.0
The FPP project obtained PFAS data in Scotland surface water for the year 2018 through a "freedom of information request" to the Scottish Environment Protection Agency (SEPA).
Scotland Scottish Environment Protection Agency (SEPA)
()
2024-07-03
2025-05-06
static
61
22
9
75.41%
6.0
0.0
Watershed sampling campaign
()
2024-07-03
2025-05-06
static
3
2
19
100.0%
3.0
3.0
This Data comes from a FOI to the UK Environment Agency. Sadly, the data does not indicate the dates of the samplings, so they are discarded for now
No sampling date provided
2024-07-03
2025-05-06
static
0
0
Data form the UK Water Industry Research, from the program "Chemical Investigations Programme Phase 2" The data was manually downloaded from the WIR website (free account needed) I extracted one file per year 2015-2020, with no restrictions on substances, stations or companies. The data is then filtered on all parameter names that are in our synonym list.
We then proceed to 3 corrections: 1: The data has 1 location (Stoke St George) where the coordinates are in EPSG 27700 (all the rest being in lat / lon directly). 2: The data has various locations where for some rows, the lat or the lon is 0, although other rows at the same location have the values. We take the values from complete rows. 3: Olivia.Mair@defra.gov.uk sent us a correction request by email on 2023-02-23, so we manually correct the position of 1 water treatment plant.
2024-07-03
2025-05-21
dynamic
49533
2116
11
98.98%
15359.0
438.0
2024-07-03
2025-05-06
static
95
95
16
100.0%
95.0
83.0
Malta Environment and Resources Authority
()
2024-07-03
2025-05-06
static
18
6
2
100.0%
18.0
18.0
Swiss Federal Office for the Environment
()
2024-07-03
2025-05-06
static
146
124
15
100.0%
146.0
146.0
2024-07-03
2025-05-06
static
125
0
3
100.0%
125.0
125.0
2024-07-03
2025-05-06
static
25
0
9
100.0%
3.0
0.0
Service public de Wallonie (SPW)
()
2024-07-03
2025-05-06
static
75
56
46
45.33%
34.0
34.0
Service public de Wallonie (SPW)
()
2024-07-03
2025-05-06
static
2596
55
20
84.05%
530.0
22.0
Service public de Wallonie (SPW)
()
2024-07-03
2025-05-06
static
87
67
20
73.56%
27.0
14.0
Service public de Wallonie (SPW)
()
2024-07-03
2025-05-06
static
97
43
8
13.4%
1.0
0.0
2024-07-03
2025-05-15
dynamic
1448
1170
41
56.42%
796.0
674.0
Data from the drinking water public register. The matrix was determined as follow: if libtypeeau is 'EAU DISTRIBUEE DESINFECTEE' or 'EAU DISTRIBUEE SANS DESINFECTION', then we consider it to be drinking water otherwise we look at the field inae.
2024-07-03
2025-05-21
dynamic
39632
1965
59
6.46%
1534.0
254.0
Removed on 21/02/2025 since it has neither city nor lat/lon
2024-07-03
2025-05-06
static
0
0
Source data is table 71 of the "Endbericht Projekt Popmon II", available here: https://www.ages.at/download/sdl-eyJ0eXAiOiJKV1QiLCJhbGciOiJIUzI1NiJ9.eyJpYXQiOjE2MDk0NTkyMDAsImV4cCI6NDA3MDkwODgwMCwidXNlciI6MCwiZ3JvdXBzIjpbMCwtMV0sImZpbGUiOiJmaWxlYWRtaW4vQUdFU18yMDIyLzZfRk9SU0NIVU5HL1dpc3Nlbi1Ba3R1ZWxsL0xlYmVuc21pdHRlbHNpY2hlcmhlaXQvMjAyMS9FbmRiZXJpY2h0X1BPUE1PTl9JSS5wZGYiLCJwYWdlIjoyMzkwfQ.fAmu0prHOgign6SL5Pr1w7pfOiGzD49QE9fN1cylOPA/Endbericht_POPMON_II.pdf
AGES
()
2024-07-03
2025-05-06
static
10
0
7
100.0%
10.0
9.0
2024-07-03
2025-05-21
dynamic
126
13
11
42.06%
31.0
4.0
Data from the UK Environment Agency. So far we filter the list of determinands codes, but this is not great, I asked their support for a better list.
2024-08-30
2025-05-06
dynamic
13694
2814
48
80.66%
3797.0
391.0
This is data from the NRW , like datasets 47 and 48, but here the data is directly published via a CSV file. The CSV files come zipped, but in a weird way: the zipped and uncompressed files are the same size and I cannot automate the unzipping at all. For now, I just added the CSV files, but this is a bit sad, since it mean it wont update alone. I will try to contact them to solve it.
The data of the messtellen is available in a separate CSV. The coordinates are in UTM 32, with the last 2 digits often replaced by XX, not sure why (maybe safety reasons) I replaced these XX by 50: this is not perfect, but the difference is max 20 meters, so they still provide value on the map.
2024-10-18
2025-05-21
dynamic
1860
541
23
63.92%
1036.0
668.0
The data was communicated to us by the journalist of Radio France who published the investigation. They did their own sampling campaign.
2024-10-21
2025-05-06
static
89
0
25
42.7%
18.0
3.0
This data was given to us by the NGO Pesticide Action Network (PAN), who conducted their own study. Some lines had no coordinates and were removed. We don't have the precise date, but they confirmed the samplings were done in 2024.
2024-11-06
2025-05-06
static
34
34
1
100.0%
34.0
33.0
Data from the drinking water monitoring from the Spanish Basque country (Euskadi). Data is freely available by download, only data from 2023 on has PFAS data. The X and Y coordinate are switched in their data, if at some point they correct it we will need to update the normaliser or the extractor.
Dpto. de Salud del Gobierno Vasco
()
2024-11-14
2025-05-21
dynamic
250
13
20
6.4%
5.0
0.0
The data is manually donwloaded from https://www.gkd.bayern.de/en/downloadcenter/wizard The GDK confirmed the data is updated once a year and there is no API. So we can update it manually once a year. Select all time period, individual values, then a program (multiple program fails, so one by one): Groudwater - chemie, Rivers - chemie (biota program), Rivers - chemie (water program), lakes -chemie we get 4 zips, 2 of them will uncompress in the same folder "fluesse - chemie", but that is not a problem. Once unzipped, The files are single files per sampling, which is as far as I know, unique among our data, but not hard to parse. Each file has 6 to 8 info lines, a blank line, then 2 data lines, one with headers, then with data, csv style. Some exceptions files have no data
2024-11-18
2025-05-21
dynamic
3879
1352
32
74.94%
1017.0
253.0
This data comes from FOI requests done by Watershed Investigation in the UK to various british institutions.
()
2025-01-24
2025-05-21
dynamic
1002
398
2
100.0%
1002.0
1000.0
Data obtained by a FOI request to the Nature England. The coordinates are provided with UK Grid Reference with 2 only digits, providing an accuracy of only around 7 km.
2025-04-02
2025-05-06
static
60
58
16
100.0%
60.0
60.0
Data obtained by a FOI request to the Nature England. The coordinates are provided with UK Grid Reference with 2 only digits, providing an accuracy of only around 7 km.
2024-07-03
2025-05-06
static
86
40
41
100.0%
86.0
86.0
Data had a weird codeStation starting with St Found a doc here https://www.aquaref.fr/system/files/RDI-RSP_2019_LotC_Screening_environnemental.pdf with the Code Stations and the Sandre station number, Bingo! 4 stations are missing, excluding them for now
2025-02-21
2025-05-21
dynamic
45
45
38
88.89%
40.0
37.0
Data downloaded from the Landesanstalt für Umwelt Baden-Württemberg, more precisely from the "Daten- und Kartendienst der LUBW 4.0", here: https://umweltdaten.lubw.baden-wuerttemberg.de/repositories/yUm3pYRGyaP7hFRow7Rf/workbooks/Fliessgewaesserguete,8SMZrw9xObs2ChqTSHk1/worksheets/Uebersicht,-R5O52_zepkftwFbe4hS?workbookHash=JWPlghdPks8hQOwGr51Qnxnbdrs_t_5qYEnUcrymF5IfViFO Dataset 140 is about surface water (also contains sediment and suspended matter data) There is no API, it's a manual download, so it could be nice to update it from time to time.
2025-03-14
2025-05-06
static
6123
209
21
85.6%
2319.0
671.0
Data downloaded from the Landesanstalt für Umwelt Baden-Württemberg, more precisely from the
"Daten- und Kartendienst der LUBW 4.0", here: https://umweltdaten.lubw.baden-wuerttemberg.de/repositories/uXwTw947n7bRQG4iOuXI/workbooks/Grundwasserguete,ElW90OFKM4Yp6SuaXMUC/worksheets/Messwerttabelle,uLSUz2jIT4XNn9Ftr2gN?workbookHash=wsOVd6LYUWvdy4StJyAkPlyCa0zk9K_o144ymskuOsX3QztF&predecessor=ElW90OFKM4Yp6SuaXMUC%2Chash%3DwsOVd6LYUWvdy4StJyAkPlyCa0zk9K_o144ymskuOsX3QztF&worksheetPredecessor=j_4Qviq9qY3F9f5s1S9m&transferConditionMode=MERGE
Dataset 141 is about groundwater.
There is no API, it's a manual download, so it could be nice to update it from time to time.
Surprisingly, the lat and lon of the stations are not given in the download, but I found the data of all german
stations for groundwater here: https://umwelt.info/de/artikel/karte_grundwasser
I joined using the Messtelle-nummer and got almost all the data points.
2025-03-14
2025-05-21
dynamic
3462
796
16
47.0%
836.0
197.0
This data results from the French National PFAS Plan. Industrial facilities were targeted based on their activity, and asked to conduct self evaluation of their PFAS emission (sending samples to certified labs).
The results are published regularly by each French regions, in different format, although they clearly come from the same database, which makes the data extraction process really tedious. The data also contained the flow per day, allowing to calculate the emitted PFAS per day. This is not on our map since it doesn't fit our data format. But you can find it in the analysis done by Le Monde here: https://www.lemonde.fr/les-decodeurs/article/2025/04/01/visualisez-les-380-sites-industriels-qui-rejettent-l-essentiel-des-pfas-en-france_6589454_4355770.html
DISCLAIMER: there are data problems. If you see something incorrect, please check first in the source file of the region that published the data (click the source link, on this page there will be a link to a file with the data)
Technical steps:
1) we listed the webpages where the region publish the link to their data (as the link changes with each update). This is done in a public spreadsheet here: https://docs.google.com/spreadsheets/d/15soDzQqpIAeSarbXWpEHjVv-uQavPewmlbOfPqHk4QE/edit?gid=0#gid=0
2) For each region, we find the link, download the data, read it although the format differs, merge it. See function process_df.
3) We can then merge the data into a big dataset.
4) We found installation data here: https://www.data.gouv.fr/fr/datasets/base-des-installations-classees-icpe/ Although I must say I am not sure that the data is regularly updated. We use it to try to find the lat / lon of each installation. it works for 2200 / 2700 installations, which seems not bad. Contact us if you see mistakes.
DISCLAIMER: there are data problems. If you see something incorrect, please check first in the source file of the region that published the data (click the source link, on this page there will be a link to a file with the data)
Technical steps:
1) we listed the webpages where the region publish the link to their data (as the link changes with each update). This is done in a public spreadsheet here: https://docs.google.com/spreadsheets/d/15soDzQqpIAeSarbXWpEHjVv-uQavPewmlbOfPqHk4QE/edit?gid=0#gid=0
2) For each region, we find the link, download the data, read it although the format differs, merge it. See function process_df.
3) We can then merge the data into a big dataset.
4) We found installation data here: https://www.data.gouv.fr/fr/datasets/base-des-installations-classees-icpe/ Although I must say I am not sure that the data is regularly updated. We use it to try to find the lat / lon of each installation. it works for 2200 / 2700 installations, which seems not bad. Contact us if you see mistakes.
()
2025-04-07
2025-05-29
dynamic
20984
2017
50
57.39%
11458.0
8484.0