Our project is jointly supported by the Belmont Forum and the National Science Foundation (NSF). This international collaboration aims to improve our understanding of the interaction between sea ice loss and fog over the Arctic Ocean. Our US team particularly focuses on developing Arctic fog detection using existing satellite measurements to better understand how the climatology of Arctic fog might have changed under global change. Over the funding period, we successfully achieved two major accomplishments: We developed an Arctic fog detection algorithm (called the ∂T method) based on satellite-based infrared spectral measurements made by NASA's Moderate Resolution Imaging Spectroradiometer (MODIS) and we also developed an Arctic fog detection algorithm based on satellite-based lidar measurements made by NASA's Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observation (CALIPSO).  We validated both algorithms by independent data, including ship-based visibility measurements made on the Chinese research vessel Xuelong. 

 Infrared-based and Lidar-based fog detection algorithms

Fog is a class of "very low" stratus cloud whose base touches the surface (land or ocean). Although it is physically similar to cloud, fog plays a different role on the radiation budget. Despite its importance, fog is often not considered in climate studies related to vertical cloud distribution. Therefore, in the NSF proposal, we proposed to devise a satellite-based Arctic fog observation using infrared measurements. Satellite-based lidar measurements may also be used to detect fog but such measurements have been available only for a decade; satellite-based infrared measurements, in contrast, can be used to extend the fog retrieval to a longer period.

World Meteorology Organization (WMO) defines fog based on a horizontal visibility of less than 1 km. This definition works well for surface observations but can hardly be used in satellite-based observations. From satellite perspectives, one has to detect fog from a top-down approach. We developed a completely different method using a new quantity, the sea-surface/cloud-top temperature difference (denoted by ∂T). ∂T is defined as the difference between the cloud-top brightness temperature detected by MODIS and the sea-surface temperature obtained from Optimum Interpolation Sea Surface Temperature (OISST). ∂T was first introduced by the NOAA scientist, Ellrod in 2003, to detect night-time fog/stratus over conterminous US.

The MODIS ∂T method is an indirect method: It relies on the inference of the temperature inversion in the fog layer. In contrast, the CALIPSO fog detection algorithm is a direct method: the lidar measurement directly provide the cloud-base height to detect fog.  Thus, we further developed the CALIPSO fog detection algorithm to provide an independent fog dataset that will be crucial for a long-term validation of the ∂T method.

We validated the CALIPSO fog detection against some ground truth. Surface observations of fog over the Arctic Ocean are extremely rare. We used the fog reports from two academic survey cruise expeditions, named Xuelong (literally means Snow Dragon) I and II, during the summers from 2016 to 2020. The cruise typically departs from Shanghai, China, in July and returns in September. The time, location, and horizontal visibility recorded along the cruise track are used to compare with our CALIPSO-based fog frequencies. Following the WMO definition, we define a fog event when the in-situ horizontal visibility is less than 1 km.

Projected Arctic sea fog in future climate scenarios

The societal impacts of Arctic sea fog under global change is one of our primary focuses.  Ship companies can realize the greatest advantage, because present travels of more than 20,000 km from Far East to Northwest Europe via the Suez Canal can be reduced to about 10,000 km and the averaged sailing times can be shortened from 20 days in the 1990s to 11 days either if the Northern Sea Route (NSR) via north of the Russian Federation through the Arctic, or the Northwest Passage (NWP) via the Canadian Arctic Archipelago are used. 

However, sea ice is not the only meteorological variables that poses safety issue on the journey. For example, increased fog formation over newly formed open-water area along the sea ice cracks severely reduces the visibility; the vessel may have to slow down or even stop when the visibility is too low, which could delay the shipping plan and lead to significant economic loss.

We developed a new cost function based on some pre-determined correlation between the cruising speeds with the local sea-fog/sea-ice conditions to optimize the time spent on a trans-Arctic route.

Along the NWP, the fastest path is again more poleward than the shortest path.  It runs across Baffin Bay and the Alaskan coastline to avoid high-fog regions. Although the fastest path is longer than the shortest path by 117 km, the voyage time of the fastest path is shortened by up to 1 day in both RCP8.5 and RCP4.5 because of the less-fog path.  

Last Modified: 11/28/2022
Modified by: Ka-Kit Tung