Concepts:Chokepoint: the full list of chokepoints we cover and associated additional information. Vessel Count: we use real-time data on vessel movements—Automatic Identification System (AIS) signals of vessels— to calculate the yearly average number of ships passing through the chokepoint, over the timeframe 2019-2024. These yearly averages are available by 5 main ship categories (container, dry bulk, general cargo, ro-ro and tanker) plus the total. Top Industries: dominant traded industries based on the volume of goods estimated to flow through the chokepoint. Note that this is not based on official statistics.Variables: portid: chokepoint unique id. portname: chokepoint name. lat: latitude of the chokepoint location.lon: longitude of the chokepoint location.vessel_count_total: yearly average number of all ships transiting through the chokepoint. Estimated using AIS data between 2019-2024. The total is calculated over the sum of vessel_count_container, vessel_count_dry_bulk, vessel_count_general_cargo, vessel_count_roro and vessel_count_tanker.vessel_count_container: yearly average number of containers transiting through the chokepoint. Estimated using AIS data between 2019-2024. vessel_count_dry_bulk: yearly average number of dry bulk carriers transiting through the chokepoint. Estimated using AIS data between 2019-2024. vessel_count_general_cargo: yearly average number of general cargo ships transiting through the chokepoint. Estimated using AIS data between 2019-2024. vessel_count_roro: yearly average number of Ro-Ro ships transiting through the chokepoint. Estimated using AIS data between 2019-2024. vessel_count_tanker: yearly average number of tankers transiting through the chokepoint. Estimated using AIS data between 2019-2024. industry_top1: first dominant traded industries based on the volume of goods estimated to flow through the chokepoint.industry_top2: second dominant traded industries based on the volume of goods estimated to flow through the chokepoint.industry_top3: third dominant traded industries based on the volume of goods estimated to flow through the chokepoint.How to Cite?These datasets are based on raw AIS data from the United National Global Platform and estimates by the PortWatch team based on the methodology described in the paper. The recommended citation is: “Sources: UN Global Platform; IMF PortWatch (portwatch.imf.org).”

To help countries identifying the main systemic ports of interest, we have classified ports into those that are globally systemic (i.e., important for global supply-chains), regionally systemic (e.g., essential for regional trade integration), and domestically systemic (e.g., important for domestic supply-chains).In particular, global systemically important ports (G-SIPs) are the world’s largest ports that cumulatively handle more than 50 percent of global trade value. Regional systemically important ports (R-SIPs) are the ports that handle at least five percent of the imports of at least four countries. Domestic systemically important ports (D-SIPs) are the ports that cumulatively handle more than 80 percent of a country’s imports.Concepts:Port: the full list of ports we cover and associated additional information. Vessel Count: we use real-time data on vessel movements—Automatic Identification System (AIS) signals of vessels— to calculate the yearly average number of ships flowing through the port, over the timeframe 2019-2024. These yearly averages are available by 5 main ship categories (container, dry bulk, general cargo, ro-ro and tanker) plus the total. Top Industries: dominant traded industries based on the volume of goods estimated to flow through the port. Note that this is not based on official statistics. The 13 economic industries are based on the International Convention on the Harmonized Commodity Description and Coding System (HS Convention) and align to the 21 HS Sections as per the table below.NameHS SectionAnimal & Animal Products1Vegetable Products2Prepared Foodstuffs & Beverages3+4Mineral Products5Chemical & Allied Industries6Plastics, Rubber, Leather7+8Wood & Wood Products9+10Textiles & Footwear11+12Stone & Glass13+14Metals15Machinery & Electrical Equipment16+18Vehicles & Equipment17Miscellaneous19+20+21Port's Share of Economy's Maritime Trade: based on AIS-derived imports and exports at the port-level with respect to country totals during 2019-24.Variables: portid: port unique id. portname: port name. lat: latitude of the port location.lon: longitude of the port location.vessel_count_total: yearly average number of all ships calling at the port. Estimated using AIS-derived portcalls data between 2019-2023. The total is calculated over the sum of vessel_count_container, vessel_count_dry_bulk, vessel_count_general_cargo, vessel_count_roro and vessel_count_tanker.vessel_count_container: yearly average number of containers calling at the port. Estimated using AIS-derived portcalls data between 2019-2024.vessel_count_dry_bulk: yearly average number of dry bulk carriers calling at the port. Estimated using AIS-derived portcalls data between 2019-2024.vessel_count_general_cargo: yearly average number of general cargo ships calling at the port. Estimated using AIS-derived portcalls data between 2019-2024.vessel_count_roro: yearly average number of Ro-Ro ships calling at the port. Estimated using AIS-derived portcalls data between 2019-2024.vessel_count_tanker: yearly average number of tankers calling at the port. Estimated using AIS-derived portcalls data between 2019-2024.industry_top1: first dominant traded industries based on the volume of goods estimated to flow through the port.industry_top2: second dominant traded industries based on the volume of goods estimated to flow through the port.industry_top3: third dominant traded industries based on the volume of goods estimated to flow through the port.share_country_maritime_import: based on AIS-derived imports volume at the port-level with respect to country totals during 2019-24.share_country_maritime_export: based on AIS-derived exports volume at the port-level with respect to country totals during 2019-24.How to Cite?These datasets are based on raw AIS data from the United National Global Platform and estimates by the PortWatch team based on the methodology described in the paper. The recommended citation is: “Sources: UN Global Platform; IMF PortWatch (portwatch.imf.org).”

We use real-time data on vessel movements—Automatic Identification System (AIS) signals of vessels—as our primary data source.Source:Arslanalp, S., Koepke, R., & Verschuur, J. Tracking Trade from Space: An Application to Pacific Island Countries. IMF Working Paper No. 2021/225. https://www.imf.org/en/Publications/WP/Issues/2021/08/20/Tracking-Trade-from-Space-An-Application-to-Pacific-Island-Countries-464345Concepts:Ports: Full list of ports we cover and associated additional information can be found here. Port Calls: a port call is defined when a ship enters the port boundary. Port calls with a turnaround time of less than 5 hours and no draft change between the current and next port are excluded to filter out vessels in transit (i.e., vessels that visit a port for reasons other then loading and discharging cargo, such as anchoring, refueling or provisioning). Trade Estimates: as described in the paper, we use the vessel information (length, width, draft, capacity, block coefficient) to estimate the payload (or utilization rate) of the vessel when entering and leaving the port boundary. The change in the vessel payload (in percentage points) multiplied by the vessel’s deadweight tonnage (the maximum carrying capacity) in metric tons is the resulting trade flow (either import or export) in metric tons.Ship Categories: all the indicators are available by 5 main ship categories: container, dry bulk, general cargo, ro-ro and tanker.Variables: date: all port call dates are expressed in Coordinated Universal Time (UTC), a standard used to set all time zones around the world. year: as extracted from date. month: month 1-12 extracted from date. day: day 1-31 extracted from date. portid: port id. Full list of ports and associated additional information can be found here. portname: port name. country: country the port resides in. ISO3: ISO 3-letter country code of the port. portcalls_container: number of container ships entering the port at this date. portcalls_dry_bulk: number of dry bulk carriers entering the port at this date. portcalls_general_cargo: number of general cargo ships entering the port at this date. portcalls_roro: number of ro-ro ships entering the port at this date. portcalls_tanker: number of tankers entering the port at this date. portcalls_cargo: total number of ships (excluding tankers) entering the port at this date. This is the sum of portcalls_container, portcalls_dry_bulk, portcalls_general_cargo and portcalls_roro. portcalls: total number of ships entering the port at this date. This is the sum of portcalls_container, portcalls_dry_bulk, portcalls_general_cargo, portcalls_roro and portcalls_tanker. import_container: total import volume (in metric tons) of all container ships entering the port at this date. import_dry_bulk: total import volume (in metric tons) of all dry bulk carriers entering the port at this date. import_general_cargo: total import volume (in metric tons) of all general cargo ships entering the port at this date. import_roro: total import volume (in metric tons) of all ro-ro ships entering the port at this date. import_tanker: total import volume (in metric tons) of all tankers entering the port at this date. import_cargo: total import volume (in metric tons) of all ships (excluding tankers) entering the port at this date. This is the sum of import_container, import_dry_bulk, import_general_cargo and import_roro. import: total import volume (in metric tons) of all ships entering the port at this date. This is the sum of import_container, import_dry_bulk, import_general_cargo, import_roro and import_tanker. export_container: total export volume (in metric tons) of all container ships entering the port at this date. export_dry_bulk: total export volume (in metric tons) of all dry bulk carriers entering the port at this date. export_general_cargo: total export volume (in metric tons) of all general cargo ships entering the port at this date. export_roro: total export volume (in metric tons) of all ro-ro ships entering the port at this date. export_tanker: total export volume (in metric tons) of all tankers entering the port at this date. export_cargo: total export volume (in metric tons) of all ships (excluding tankers) entering the port at this date. This is the sum of export_container, export_dry_bulk, export_general_cargo and export_roro. export: total export volume (in metric tons) of all ships entering the port at this date. This is the sum of export_container, export_dry_bulk, export_general_cargo, export_roro and export_tanker.How to Cite? These datasets are based on raw AIS data from the United National Global Platform and estimates by the PortWatch team based on the methodology described in the paper. The recommended citation is: “Sources: UN Global Platform; IMF PortWatch (portwatch.imf.org).”About AIS Data The UN has made available satellite-based AIS data through the UN Global Platform (UNGP) to national and international agencies that are members to the UN-CEBD (UN, 2021). The platform contains live data and global archive data from December 1, 2018. AIS data at the UNGP are provided by Spire, which collects AIS messages from two different satellite constellations, with more than 65 AIS equipped satellites. Spire complements this information with data collected by FleetMon through terrestrial receivers. There are several challenges with using AIS data. First, ships can turn off their AIS transponder to avoid being detected. Strictly speaking, this is not legal and is mostly limited to fishing vessels conducting illegal fishing or oil tankers circumventing international sanctions, usually in international waters. It is not common for container and other cargo ships (which is the focus of our study) entering a country’s port. In fact, in most jurisdictions, port authorities make it mandatory for ships entering a port to keep their AIS transponders on at all times for the safety all vessels in the port. Second, the AIS data do not have information about the ship’s carrying capacity (i.e., deadweight tonnage) and maximum draft. To fill this gap, we use ship registry databases from FleetMon and IHS Markit (the latter is available from the UNGP), with information for around 120,000 vessels. Finally, a potentially more serious challenge with AIS data is that some information is entered manually and, hence, may have human errors. This is expected as AIS was intended originally for safety at sea, not for producing statistics. For our purposes, the key issue is that the crew may not always update the draft information after a ship leaves the port. The draft is the vertical distance between the waterline and the bottom of the ship’s hull and is a measure of the payload of the vessel. However, our algorithm uses techniques to address this issue. Particularly, the missing information can be backtracked or imputed in most cases, given the wealth of information in the AIS data (Arslanalp, Koepke, Verschuur, 2021). The AIS was originally developed by the International Maritime Organization (IMO) in 2004 as an outcome of amendments to the International Convention SOLAS (Safety of Life at Sea) in 2002. It is a self-reporting system, which allows vessels to periodically broadcast their identity, navigation, position data and other characteristics. The AIS has been made compulsory for all international commercial ships with gross tonnage of 300 or more tons (i.e., virtually all commercial ships) and all passenger ships regardless of size. There are three main types of information in AIS messages. AIS broadcasts voyage-related information (including ship location, speed, course, heading, rate of turn, destination, draft, and estimated arrival time), static information (including ship ID, ship type, ship size and dimensions), and dynamic information. Dynamic information such as the positional aspects (latitude and longitude) is automatically transmitted, depending on the vessels’ speed and course. The signals can be picked up by satellite or terrestrial receivers. For ships in open seas, however, the signals can only be picked up by satellite receivers as terrestrial receivers typically cover only about 15–20 nautical miles from the coast. For island states, satellite data tend to be much more reliable as the coverage of terrestrial receivers can be low (or nonexistent) for these smaller countries. Terrestrial receivers are useful for congested ports where congestion may make it difficult for satellites to capture all emitted messages. Additional information on AIS data can be found in Arslanalp et al. (2019), Verschuur et al. (2020), and the UN’s AIS Handbook. References: Arslanalp, S., Koepke, R., & Verschuur, J. Tracking Trade from Space: An Application to Pacific Island Countries. IMF Working Paper No. 2021/225. https://www.imf.org/en/Publications/WP/Issues/2021/08/20/Tracking-Trade-from-Space-An-Application-to-Pacific-Island-Countries-464345 AIS Handbook https://unstats.un.org/wiki/display/AIS/AIS+Handbook

The country layer captures the network of ports and routes that facilitate a country’s maritime trade. The includes all the exporting (either in country or in transport connected country), transhipment, and importing (at ports of bilateral trade partner) ports that facilitate a country’s maritime exports, and the importing (either in country or in transport connected country), transhipment, and exporting (at ports of bilateral trade partner) ports that facilitate for a country’s maritime imports. The Oxford Maritime Transport (OxMarTrans) model predicts the allocation of maritime trade flows (based on bilateral trade data) on the maritime transport network, including the port and route taken, to determine the dependency between ports and trade flow. In other words, it captures how trade between origin and destination is most likely being shipped across the global maritime transport network, including the port used for exporting, transhipment (if required) and importing. The resulting network consist of >2.1 million unique port-country pair combinations across >25,000 unique country pairs and 13 economic sectors, which capture what share of maritime trade between two countries going through specific ports and on specific routes. The base year considered is 2022.The 13 economic sectors are based on the International Convention on the Harmonized Commodity Description and Coding System (HS Convention) and align to the 21 HS Sections as per the table below.NameHS SectionAnimal & Animal Products1Vegetable Products2Prepared Foodstuffs & Beverages3+4Mineral Products5Chemical & Allied Industries6Plastics, Rubber, Leather7+8Wood & Wood Products9+10Textiles & Footwear11+12Stone & Glass13+14Metals15Machinery & Electrical Equipment16+18Vehicles & Equipment17Miscellaneous19+20+21The spillover simulator shows the amount of country’s imports or exports that is at-risk of being affected because of a disruption at a selected port in value terms. This could be because the port was exporting/importing a country’s imports/exports, or because of transhipments. This impacts can be analysed in aggregate terms (across all commodity sectors) or per commodity sector. In addition, the amount of trade at-risk is expressed as a fraction of an economy’s Gross Domestic Product (GDP) to capture how impactful trade disruptions are in relative terms for respective country economies. It should be noted that in most cases trade is not directly lost, but merely delayed in case of smaller disruptions (<7 days). However, in case of larger disruptions (>30 days) and limited possibilities to reroute goods, trade bottlenecks could occur, potentially resulting in severe supply shortages. GDP data comes from the October 2023 WEO (NGDPD = Gross domestic product, current prices; Year 2022).Source: Verschuur, J., Koks, E.E. & Hall, J.W. Ports’ criticality in international trade and global supply-chains. Nat Commun 13, 4351 (2022). https://doi.org/10.1038/s41467-022-32070-0Variables:from_portid = port id. Full list of ports can be found here.from_portname = port name. from_country = country the port resides in.from_iso3 = ISO 3-letter country code of the port.to_country = country for which we compute trade that is at-risk of being affected because of a disruption at from_portname.to_iso3 = ISO 3-letter country code of the country for which we compute trade that is at-risk of being affected because of a disruption at from_portname.industry = one of the following: - Animal & Animal Products- Vegetable Products- Prepared Foodstuffs & Beverages- Mineral Products- Chemical & Allied Industries- Plastics, Rubber, Leather- Wood & Wood Products- Textiles & Footwear- Stone & Glass - Metals- Machinery & Electrical Equipment- Vehicles & Equipment- Miscellaneous- Total (sum of all of the above)hs_section = corresponding HS section(s)unit = all values are expressed in US Dollars.scale = all values are expressed in units.daily_export_value_at_risk = daily to_country’s exports that is at-risk of being affected because of a disruption at from_portname.daily_import_value_at_risk = daily to_country’s imports that is at-risk of being affected because of a disruption at from_portname.

The spillover simulator allows users to visualize the potential spillover effects of port disruptions. Users are able to fill in a disruption duration (# of days), which is used to scale the output of the spillover analysis (linearly). The simulator consists of three different types of visualizations, which distinguish the spillover impacts on three different spatial scales: the port layer, the country trade layer, and the supply-chain layer. This methodology is based on Verschuur, Koks and Hall (2022).The port layer describes the network of port-to-port connections given vessels movements between two respective ports. In other words, it captures what the last port of call was for the vessels arriving at the port of interest, and the next port of call is for the vessels leaving the port of interest. The weight on the network is the total carrying capacity of vessels between port pairs, which captures the total amount of freight that could be handled on a route.The network is based on a satellite-derived dataset of vessel movements of around 120,000 vessels between 2019 and 2023, from which a port-to-port transport network was constructed. This network includes the loaded capacity (payload multiplied by the carrying capacity of the vessel) of vessels on a given route. The result of the simulator shows the amount of maritime capacity being at-risk of facing delays or disruptions due to port disruptions that affect outgoing vessel movements as a percentage of the average annual capacity on that given route. Any form of route adjustments are not taken into consideration.Source: Verschuur, J., Koks, E.E. & Hall, J.W. Ports’ criticality in international trade and global supply-chains. Nat Commun 13, 4351 (2022). https://doi.org/10.1038/s41467-022-32070-0How to cite? These dataset combine data from the journal article published by researchers affiliated with Oxford University and calculations by the PortWatch team. The recommended citation is: “Sources: University of Oxford; IMF PortWatch (portwatch.imf.org).”Variables: from_portid = origin port id. Full list of ports can be found here. from_portname = origin port name. from_country = country the origin port resides in. from_iso3 = ISO 3-letter country code of the origin port. to_portid = destination port id. Full list of ports can be found here. to_portname = destination port name. to_country = country the destination port resides in. to_iso3 = ISO 3-letter country code of the destination port. average_transit_days = average number of transit days for the route from origin to destination port. daily_capacity_at_risk = the loaded capacity (payload multiplied by the carrying capacity of the vessel) of vessels on a given route. Values are expressed in Metric Tons.relative_capacity_at_risk = the daily_capacity_at_risk as a percentage of the annual average loaded capacity (payload multiplied by the carrying capacity of the vessel) of vessels on a given route. Values are expressed in Percent.

SourcesVerschuur, J., Koks, E.E., Li, S. et al. Multi-hazard risk to global port infrastructure and resulting trade and logistics losses. Commun Earth Environ 4, 5 (2023). https://doi.org/10.1038/s43247-022-00656-7Verschuur, J., Koks, E.E. & Hall, J.W. Systemic risks from climate-related disruptions at ports. Nat. Clim. Chang. 13, 804–806 (2023). https://doi.org/10.1038/s41558-023-01754-wConcepts:Climate risk: expressed as either the annual expected physical asset damages (that is, the damage to port infrastructure) reflecting natural hazards for each port, or as the annual expected number of downtime per port because of expected natural disasters.Variables:portid = port id. Full list of ports can be found here.portname = port name. country = country of the port.ISO3= ISO 3-letter country code of the port.lat = latitude location of the port.lon = longitude location of the port.unit = depending on the measure this will be either USD per year (for measure = Physical asset damages) or Days per year (for measure = Port downtime).measure = Physical asset damages or Port downtimevalue = climate risk valuehazard = type of hazard. This can be one of the following: fluvial flooding, pluvial flooding, coastal flooding, cyclones, earthquakes, operational thresholds.How to cite? These dataset combine data from the journal article published by researchers affiliated with Oxford University and calculations by the PortWatch team. The recommended citation is: “Sources: University of Oxford; IMF PortWatch (portwatch.imf.org).”

SourcesVerschuur, J., Koks, E.E., Li, S. et al. Multi-hazard risk to global port infrastructure and resulting trade and logistics losses. Commun Earth Environ 4, 5 (2023). https://doi.org/10.1038/s43247-022-00656-7Verschuur, J., Koks, E.E. & Hall, J.W. Systemic risks from climate-related disruptions at ports. Nat. Clim. Chang. 13, 804–806 (2023). https://doi.org/10.1038/s41558-023-01754-wConcepts:Trade risk: shows for the selected country the trade risk of country imports or exports at ports globally. This is derived by combining the country trade flows at the port-level with the annual expected downtime. For example, if a port handles US$ 1 billion of a country’s imports every year, and the expected annual downtime is 5.5 days per year, then the trade at-risk indicator is US$ 15 million per year (=1*5.5/365).Climate scenarios: The Climate Scenarios are available for present-day (roughly based on data for the last 30-40 years) and three RCP future scenarios for 2050.The RCPs provide trajectories of greenhouse gas (GHG) concentrations for the 21st century, as adopted by the International Panel on Climate Change (IPCC). The RCPs – originally RCP2.6, RCP4.5, RCP6, and RCP8.5 – are labelled after a possible range of GHG concentrations by 2100 (2.6, 4.5, 6, and 8.5 W/m2, respectively). PortWatch uses three RCP scenarios: RCP2.6, RCP4.5 and RCP8.5. Based on the latest global climate models, these RCP scenarios result in a global mean temperature increase at the end of the century (2081-2100) compared to pre-industrial (1850-1900) of 1.8 (1.3 - 2.4) degrees (RCP2.6), 2.7 (2.1 - 3.5) degrees (RCP4.5) and 4.4 (3.3 - 5.7) degrees (RCP8.5). The higher RCP values mean higher GHG emissions and therefore higher global temperatures and more pronounced effects of climate change.The lower RCP values, on the other hand, are more desirable for humans but require more stringent climate change mitigation efforts to achieve them. The RCPs are not considered a forecast but a plausible range of GHG concentrations. For more details, please see: https://ar5-syr.ipcc.ch/topic_futurechanges.php.industry: The 13 economic sectors are based on the International Convention on the Harmonized Commodity Description and Coding System (HS Convention) and align to the 21 HS Sections as per the table below.NameHS SectionAnimal & Animal Products1Vegetable Products2Prepared Foodstuffs & Beverages3+4Mineral Products5Chemical & Allied Industries6Plastics, Rubber, Leather7+8Wood & Wood Products9+10Textiles & Footwear11+12Stone & Glass13+14Metals15Machinery & Electrical Equipment16+18Vehicles & Equipment17Miscellaneous19+20+21Variables:from_country = country of interest.from_iso3 = ISO 3-letter country code.to_portid = port id. Full list of ports can be found here.to_portname = port name. to_country = country of the port.to_iso3 = ISO 3-letter country code of the port.to_lat = latitude location of the port.to_lon = longitude location of the port.scenario = either present or one of the RCP scenarios. industry = one of the following:- Animal & Animal Products- Vegetable Products- Prepared Foodstuffs & Beverages- Mineral Products- Chemical & Allied Industries- Plastics, Rubber, Leather- Wood & Wood Products- Textiles & Footwear- Stone & Glass - Metals- Machinery & Electrical Equipment- Vehicles & Equipment- Miscellaneous- Total (sum of all of the above)hs_section = corresponding HS section(s)rank = goes from 1 to 100 and rank the ports for a specific country of interest based on the trade value at risk. The rank is only available for the industry Total (sum of all commodities).days_downtime_at_port = annual expected downtime in days for the specific port.trade_value_at_risk = country trade flows handled by the port.unit = all values are expressed in US Dollars.How to cite? These dataset combine data from the journal article published by researchers affiliated with Oxford University and calculations by the PortWatch team. The recommended citation is: “Sources: University of Oxford; IMF PortWatch (portwatch.imf.org).”

For the supply-chain layer, the output of the country trade layer is coupled to a multi-regional input-output table (EORA MRIO). A MRIO captures how industrial production and consumption is dependent on sector input from within the domestic economy and from other economies (the MRIO covers 170 economies globally). By coupling this to the country trade network, we can embed the maritime transport of flows within the tables, thereby capturing how these trade flows are used in the economy. This therefore does not only capture how a supply-chain is exposed to downtime because of trade flowing through a certain port (direct links), but also indirectly how you are exposed as a firm because you rely on firms upstream in your supply-chain that trade through a port prone to downtime (the firms you depend on, 1st order suppliers, or firms that the firms that you depend on depend on, 2nd order suppliers, etc). In other words, as a supply-chain, you may be indirectly exposed to port disruptions without your direct input or consumption goods flowing through an at-risk port (e.g. you are exposed to closure of a raw materials exporting port without you directly using raw materials in your production process).To understand the supply-chain dependencies between global supply-chains and a port, the Hypothetical Extraction Method (HEM) is used. The HEM hypothetically removes the share of trade flows between countries that go through a certain port from the MRIO and re-evaluates the industrial production globally without these flows, without any adaptation in the economic system. The difference between the original production and the HEM production is the dependency of various supply-chains on a specific port. The base year considered in 2022, which is the latest available MRIO table available.Within the portal, the absolute amount of industry production and final consumption at-risk can be visualized (all in value terms), as well as the relative amount of production and final consumption (as a fraction of an economy’s total production and consumption). This can be done in aggregate terms (across all commodity sectors) or per 13 commodity sector. Similar as with the country trade layer, the absolute numbers are indicative numbers only, and represents the relative exposure of specific supply-chains to a port disruption, which allows identifying at-risk supply chains across countries and sectors. The absolute values should however be treated with care. The 13 commodity sectors are based on the International Convention on the Harmonized Commodity Description and Coding System (HS Convention) and align to the 21 HS Sections as per the table below.NameHS SectionAnimal & Animal Products1Vegetable Products2Prepared Foodstuffs & Beverages3+4Mineral Products5Chemical & Allied Industries6Plastics, Rubber, Leather7+8Wood & Wood Products9+10Textiles & Footwear11+12Stone & Glass13+14Metals15Machinery & Electrical Equipment16+18Vehicles & Equipment17Miscellaneous19+20+21Source: Verschuur, J., Koks, E.E. & Hall, J.W. Ports’ criticality in international trade and global supply-chains. Nat Commun 13, 4351 (2022). https://doi.org/10.1038/s41467-022-32070-0How to cite? These dataset combine data from the journal article published by researchers affiliated with Oxford University and calculations by the PortWatch team. The recommended citation is: “Sources: University of Oxford; IMF PortWatch (portwatch.imf.org).”Variables:from_portid = port id. Full list of ports can be found here.from_portname = port name. from_country = country the port resides in.from_iso3 = ISO 3-letter country code of the port.to_country = country for which we compute industry production and final consumption that are at-risk of being affected because of a disruption at from_portname.to_iso3 = ISO 3-letter country code of the country for which we compute industry production and final consumption that are at-risk of being affected because of a disruption at from_portname.industry = one of the following: - Animal & Animal Products- Vegetable Products- Prepared Foodstuffs & Beverages- Mineral Products- Chemical & Allied Industries- Plastics, Rubber, Leather- Wood & Wood Products- Textiles & Footwear- Stone & Glass - Metals- Machinery & Electrical Equipment- Vehicles & Equipment- Miscellaneous- Total (sum of all of the above)hs_section = corresponding HS section(s)daily_consumption_at_risk = daily to_country’s consumption that is at-risk of being affected because of a disruption at from_portname.daily_industryoutput_at_risk = daily to_country’s industry output that is at-risk of being affected because of a disruption at from_portname.unit = all values are expressed in US Dollars.scale = all values are expressed in units.

Sources:GDACS platform: https://www.gdacs.org/ Concepts:GDACS was created in 2004 as a cooperation framework between the United Nations and the European Commission, to address significant gaps in information collection and analysis in the early phase of major sudden-onset disasters. GDACS provides real-time access to disaster information systems and currently reports are issued for earthquakes and possible subsequent tsunamis, tropical cyclones, floods and volcanic eruptions. The platform includes Disaster Alerts, a virtual On-Site Operations Coordination Centre (OSOCC) to cooperate and exchange disaster-related real time information, and maps and satellite imagery. GDACS information and data are used by many governments, disaster response organizations, and researchers and are publicly available through the GDACS platform.The data provided by GDACS is utilized in the following manner: Scan GDACS data. We scan the GDACS data on an automated and daily basis to retrieve information on active disruptions​. From each disaster report, we extract three types of information: basic information, the polygon of the impacted area, and the affected population, that is the population that is in a certain proximity of the disaster. The basic information includes information on name, location, type, event severity, countries that are impacted, etc. To determine the severity of each disaster, GDACS produces a score. The score varies between 0 to 3, for which disasters with a score between 0 to 1 are marked green, 1 to 2 orange, and 2 to 3 red. For calculation of the severity score of each disaster type (earthquakes, tsunamis, tropical cyclones, floods, volcano, or droughts) different criteria are taken into consideration. For example, for details on how alert scores are calculated for Tropical Cyclones, please visit: https://www.gdacs.org/Knowledge/models_tc.aspx. As of now, we only consider disasters with a severity score of 2 to 3 (red).Intersect disaster impact area with PortWatch ports boundaries and chokepoints boundaries. For each disaster with a severity score above 2 (a disaster in red category), we intersect the extracted impact area from GDACS with the PortWatch ports and chokepoints. The PortWatch ports and chokepoints with boundaries within the disaster impact area are marked as disrupted ports. PortWatch alerts. We send out an email alert combining the information about the disaster disrupted ports and the satellite based data​ which provides real-time information on port calls and import and export activity through the disrupted ports. Click here to subscribe to our email alerts and other updates about the PortWatch platform.PortWatch disruption pages. We include the disruption in our disruption monitor. Users can access details on each disruption with detailed analysis on spillovers at connected ports and countries, along with the unfolding of the impact on the real-time data.Variables:eventid= unique id of the event.eventtype = one of earthquakes (EQ), wild fires (WF), tropical cyclones (TC), floods (FL), volcano (VO), or droughts (DR) or other - e.g. geopolitical tensions - (OT).eventname= short name for event.htmlname = descriptive version of the eventname. htmldescription = description of event.alertlevel = refer to GDACS alert levels. They are based on a risk matrix that considers the likelihood of societies being unable to cope with a disaster at the national level. The final score also takes into account the affected country's coping capacity, which is based on the INFORM Index. This index measures a country's ability to deal with disasters through organized activities, government efforts, and infrastructure. See https://www.gdacs.org/Knowledge/models_eq.aspx for more information.country = affected country(ies). Derived based on the intersected ports.fromdate = start date of the event.todate = end date of the event.lat = latitude coordinate of the centroid of the event polygon.long = longitude coordinate of the centroid of the event polygon.affectedports = list of ports that intersect with the disruption polygon.n_affectedports = number of ports that intersect with the disruption polygon.affectedpopulation = exposed population based on GDACS assessment.How to cite? These dataset combine data from the journal article published by researchers affiliated with Oxford University and calculations by the PortWatch team. The recommended citation is: “Sources: University of Oxford; IMF PortWatch (portwatch.imf.org).”