Using satellite data to calculate maritime shipping’s carbon footprint

International shipping carries over half of global freight and accounts for about 11% of total energy use in the transportation sector. Yet until recently, few data were readily available to assess the industry’s carbon footprint. As policy makers try to evaluate maritime shipping’s contribution to climate change, shippers work to optimize energy use in their supply chains, and consumers demand more information about the carbon footprint of products, scientists have reacted by redoubling their efforts to characterize the scale and geographical distribution of carbon dioxide (CO₂) emissions from ships.

At first, researchers relied upon sales statistics for international bunker fuel (a type of fuel oil almost exclusively consumed by ships) from the International Energy Agency to estimate CO₂ emissions. But developing countries, some of which are deeply involved in international shipping, are not required to report and verify bunker sales, which decreases the accuracy of such estimates. Later analyses developed a so-called bottom-up approach: calculating CO₂ emissions from each vessel per hour, identifying hourly activities per year, summing them, and spatially allocating them across shipping lanes. This approach requires substantial data input and is highly sensitive to changes in certain assumptions, such as speed, from which engine power output is estimated. Since ship power demand increases in rough proportion to the cube of the speed, even a slight decrease in speed translates into a significant reduction in power output and thus in emissions.

During the global economic downturn of 2009–2010, decline in international trade coincided with deliveries of a record number of new ships, which had been ordered when trade was booming. In reaction, many shipping companies cut steaming speeds, which increases travel times and therefore reduces the number of total trips per year—by as much as half, potentially—to absorb excess ship capacity and conserve fuel. Consequently, fuel consumption and CO₂ emissions from shipping plummeted to a level so low that the International Maritime Organization (IMO), the governing body of international shipping, felt it necessary to update its estimate of shipping emissions earlier this year.

Another weakness in the bottom-up approach is that it depends on voluntary ship reporting to spatially allocate emissions across shipping routes. That information is important to understanding differences in ship carbon intensity across the world. But any voluntary reporting scheme is vulnerable to large uncertainties.

The ICCT is interested in improving our ability to assess ships’ carbon footprint. So we teamed up with researchers at the University of College London on a novel way around those limitations using the Satellite Automatic Identification System (S-AIS). S-AIS is a tracking system used to identify and locate vessels. The IMO International Convention for the Safety of Life at Sea, passed in the aftermath of the sinking of the Titanic, now requires that every ship larger than 300 gross tons be equipped with an S-AIS transponder that transmits ship identification, speed, and location every few seconds. This almost continuous stream of data enables researchers to calculate ship speed and therefore CO₂ emissions with greater resolution and confidence. The data also in offer an improved picture of shipping routes across the globe (see the videos below). And the initial results of this research, reported here and here, indicate that ship speeds are substantially lower than previously thought and that in-use ship efficiency varied greatly even within individual ship types.

Ship movement in August 2011, as recorded by the Satellite Automatic Identification System (S-AIS). The velocity of the dots in the video reflects ship speed. The onboard S-AIS transponder electronically transmitted ship information, location, and speed to satellite, which monitors and tracks ship locations for safety purposes.

Like every data set, S-AIS has room to improve. Data transmission is sometimes interrupted by on-shore signals, making ship movement near coasts harder to track reliably. Ships sometimes turn off their S-AIS transponders; activity in Northeast Asia is suspiciously low. But researchers and other stakeholders working on improving data quality and coverage. For example, Exact Earth, an S-AIS data provider, launched another satellite last year to better capture AIS signals. S-AIS can also be complemented by shore-based AIS data and crosschecked against bunker sales data and other bottom-up data to generate more robust estimates of emissions from shipping. As these data sets become further refined and combined in ever more productive ways, that improved ability to assess shipping’s carbon footprint is moving within reach.