Published March 11, 2025 | Version 3.0.0
Dataset Open

Global AIS-based Apparent Fishing Effort Dataset

Authors/Creators

Description

Overview

This dataset contains version 3.0 (March 2025 release) of the Global Fishing Watch apparent fishing effort dataset. Data is available for 2012-2024 and based on positions of >190,000 unique automatic identification system (AIS) devices on fishing vessels, of which up to ~96,000 are active in a given year. Fishing vessels are identified via a machine learning model, vessel registry databases, and manual review by GFW and regional experts. Vessel time is measured in hours, calculated by assigning to each AIS position the amount of time elapsed since the previous AIS position of the vessel. The time is counted as apparent fishing hours if the GFW fishing detection model - a neural network machine learning model - determines the vessel is engaged in fishing behavior during that AIS position. 

Data are spatially binned into grid cells that measure 0.01 or 0.1 degrees on a side; the coordinates defining each cell are provided in decimal degrees (WGS84) and correspond to the lower-left corner. Data are available in the following formats: 

  1. Daily apparent fishing hours by flag state and gear type at 100th degree resolution
  2. Monthly apparent fishing hours by flag state and gear type at 10th degree resolution
  3. Daily apparent fishing hours by MMSI at 10th degree resolution

The fishing effort dataset is accompanied by a table of vessel information (e.g. gear type, flag state, dimensions).

File structure

Fishing effort and vessel presence data are available as .csv files in daily formats. Files for each year are stored in separate .zip files. A README.txt and schema.json file is provided for each dataset version and contains the table schema and additional information. There is also a README-known-issues-v3.txt file outlining some of the known issues with the version 3 release.

Files are names according to the following convention:

  • Daily file format: 

    • [fleet/mmsi]-daily-csvs-[100/10]-v3-[year].zip

    • [fleet/mmsi]-daily-csvs-[100/10]-v3-[date].csv

  • Monthly file format: 

    • fleet-monthly-csvs-10-v3-[year].zip

    • fleet-monthly-csvs-10-v3-[date].csv

  • Fishing vessel format: fishing-vessels-v3.csv

  • README file format: README-[fleet/mmsi/fishing-vessels/known-issues]-v3.txt

File identifiers:

  • [fleet/mmsi]: Data by fleet (flag and geartype) or by MMSI

  • [100/10]: 100th or 10th degree resolution

  • [year]: Year of data included in .zip file

  • [date]: Date of data included in .csv files. For monthly data, [date]corresponds to the first date of the month

Examples: fleet-daily-csvs-100-v3-2020.zip; mmsi-daily-csvs-10-v3-2020-01-10.csv; fishing-vessels-v3.csv; README-fleet-v3.txt; fleet-monthly-csvs-10-v3-2024.zip; fleet-monthly-csvs-10-v3-2024-08-01.csv

Key documentation

  • For an overview of how GFW turns raw AIS positions into estimates of fishing hours, see this page.

  • The models used to produce this dataset were developed as part of this publication: D.A. Kroodsma, J. Mayorga, T. Hochberg, N.A. Miller, K. Boerder, F. Ferretti, A. Wilson, B. Bergman, T.D. White, B.A. Block, P. Woods, B. Sullivan, C. Costello, and B. Worm. "Tracking the global footprint of fisheries." Science 361.6378 (2018). Model details are available in the Supplementary Materials.

  • The README-known-issues-v3.txt file describing this dataset's specific caveats can be downloaded from this page. We highly recommend that users read this file in full.

  • The README-mmsi-v3.txt file, the README-fleet-v3.txt file, and the README-fishing-vessels-v3.txt files are downloadable from this page and contain the data description for (respectively) the fishing hours by MMSI dataset, the fishing hours by fleet dataset, and the vessel information file. These readmes contain key explanations about the gear types and flag states assigned to vessels in the dataset.

  • File name structure for the datafiles are available below on this page and file schema can be downloaded from this page. 

  • A FAQ describing the updates in this version and the differences between this dataset and the data available from the GFW Map and APIs is available here

Use Cases

The apparent fishing hours dataset is intended to allow users to analyze patterns of fishing across the world’s oceans at temporal scales as fine as daily and at spatial scales as fine as 0.1 or 0.01 degree cells. Fishing hours can be separated out by gear type, vessel flag and other characteristics of vessels such as tonnage. 

Potential applications for this dataset are broad. We offer suggested use cases to illustrate its utility. The dataset can be integrated as a static layer in multi-layered analyses, allowing researchers to investigate relationships between fishing effort and other variables, including biodiversity, tracking, and environmental data, as defined by their research objectives. 

A few example questions that these data could be used to answer:

  • What flag states have fishing activity in my area of interest?

  • Do hotspots of longline fishing overlap with known migration routes of sea turtles?

  • How does fishing time by trawlers change by month in my area of interest? Which seasons see the most trawling hours and which see the least?

Caveats

This global dataset estimates apparent fishing hours effort. The dataset is based on publicly available information and statistical classifications which may not fully capture the nuances of local fishing practices. While we manually review the dataset at a global scale and in a select set of smaller test regions to check for issues, given the scale of the dataset we are unable to manually review every fleet in every region. We recognize the potential for inaccuracies and encourage users to approach regional analyses with caution, utilizing their own regional expertise to validate findings. We welcome your feedback on any regional analysis at research@globalfishingwatch.org to enhance the dataset's accuracy.

Caveats relating to known sources of inaccuracy as well as interpretation pitfalls to avoid are described in the README-known-issues-v3.txt file available for download from this page. We highly recommend that users read this file in full. The issues described include: 

  • Data from 2024 should be considered provisional, as vessel classifications may change as more data from 2025 becomes available. 

  • MMSI is used in this dataset as the vessel identifier. While MMSI is intended to serve as the unique AIS identifier for an individual vessel, this does not always hold in practice.

  • The Maritime Identification Digits (MID), the first 3 digits of MMSI, are the only source of information on vessel flag state when the vessel does not appear on a registry. The MID may be entered incorrectly, obscuring information about an MMSI’s flag state.

  • AIS reception is not consistent across all areas and changes over time.

Alternative ways to access

  1. Query using SQL in the Global Fishing Watch public BigQuery dataset: global-fishing-watch.fishing_effort_v3

  2. Download the entire dataset from the Global Fishing Watch Data Download Portal (https://globalfishingwatch.org/data-download/datasets/public-fishing-effort)

Files

fishing-vessels-v3.csv

Files (26.3 GB)

Name Size Download all
md5:b5ba27cedd5426c0bcb8e6009e911cf0
114.8 MB Preview Download
md5:6480e047e898c365623a87cf2463f057
4.3 kB Preview Download
md5:623e8f35cfae5b79201941079f7eb369
112.5 MB Preview Download
md5:1b38df9465727636e170074855d053ce
389.5 MB Preview Download
md5:eb8930eee742f3ca8c5a910ed87e34c9
477.7 MB Preview Download
md5:d546cf28355a02352df009376fcc216d
585.3 MB Preview Download
md5:c06de397e3e075fd94eae5a90e144e15
807.1 MB Preview Download
md5:ff164d5f59ac31c4a8d4132b45dfac88
1.2 GB Preview Download
md5:c81b213da78fdeb26ab75223b1e08a56
1.5 GB Preview Download
md5:c5c42b6e954ab27add6dddf876e73bfb
1.6 GB Preview Download
md5:311b093b2abae011ad68c6d77fffd82e
1.8 GB Preview Download
md5:53227a541f8d5f51de1c60cdc7726b7b
2.1 GB Preview Download
md5:ed3695b3074e61f30ee812e642ef6cc9
2.4 GB Preview Download
md5:61fd7e282e194814c0586aa90778fdab
3.4 GB Preview Download
md5:6fae93f0126bf90f21d6f3a799d58001
3.3 GB Preview Download
md5:4e4c019416ec5a1e422e8804fe4d1169
1.5 kB Preview Download
md5:1e2a628eb249ff62a38cb0315a526957
19.0 MB Preview Download
md5:0965d473e9bf06b5e0567f3efad658fe
33.3 MB Preview Download
md5:525c250e8a942db79a56afcbdb9dc922
38.5 MB Preview Download
md5:f64ce2f0098de84ae21bbd5b0e80d937
49.4 MB Preview Download
md5:06c8d023793bb5633e4e27b4f6abcd62
66.9 MB Preview Download
md5:bd9c22c65fc49f72f54e7eaa4849b833
85.6 MB Preview Download
md5:1d35b8f0d914546d0ee67ac50cfd8e47
102.7 MB Preview Download
md5:f5e9c75c6e14f33c43e1ce213e2b592d
107.4 MB Preview Download
md5:630844627e5814edc9a63a99ba67f087
115.9 MB Preview Download
md5:7815df77783ed7064bdf1fefe11320a6
119.4 MB Preview Download
md5:ef02529bb34d0cc75c62b740dc63d04b
127.7 MB Preview Download
md5:c7ddb54e8cf0b89aa30a9fea588e6253
139.0 MB Preview Download
md5:872b8b44e8833b9dff637bc659f85584
131.7 MB Preview Download
md5:9e9c5cb7bcc35ae228ed0e14712b8fb6
1.7 kB Preview Download
md5:920f3434b53af9f3423f2087947b595a
56.0 MB Preview Download
md5:1b9927f3abea83a74e76300abddee7ee
164.6 MB Preview Download
md5:38b7f72aa20822a4ffd9cc4d0a7bb2fc
200.1 MB Preview Download
md5:88e0213e828a833ce01968e82ee5141e
229.2 MB Preview Download
md5:998e5c54971e3eb4e2665a7474b45217
300.0 MB Preview Download
md5:b30ba3a8c40804dc417b8279d4d7e7fb
376.8 MB Preview Download
md5:49c034fa6f0f236945ec1eea48da3fa5
435.6 MB Preview Download
md5:08848cf233385a09be7e12d1a7355d79
464.5 MB Preview Download
md5:7ef8d7705d3d5123a0c29401f3422541
478.8 MB Preview Download
md5:e72dad47710b01cad3dadebbe7df953c
516.2 MB Preview Download
md5:c762f968715d75ea5f29b2800f4f96c5
629.3 MB Preview Download
md5:7b55cca87029903c9becd09b11810455
739.2 MB Preview Download
md5:51b0988ac6258482c1666c113c93f004
745.7 MB Preview Download
md5:32523eb3d34b2655d1fc7355a75b08b8
1.1 kB Preview Download
md5:3c8eb8822bfd459f4fe48135be6e5078
10.1 kB Preview Download
md5:be86244c910248e1aebf052789498a1e
8.0 kB Preview Download
md5:cfc37e00ead71adb68849a168f225867
11.1 kB Preview Download
md5:f0c95507d4449a3d2edb443bd54c30da
3.4 kB Preview Download

Additional details

Related works

Is new version of
Publication: 10.1126/science.aao5646 (DOI)