Li Yi
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Li Yi has a total of 23 co-authors. Here are the top co-authors by number of shared publications:
- King-Fai Li
Department of Environmental Sciences University of California, Riverside Riverside CA USA
3 shared publications - Jörg Bendix
Department of Geography Philipps‐Universität Marburg Marburg Germany
2 shared publications - Xianyao Chen
Frontier Science Center for Deep Ocean Multispheres and Earth System, Ocean University of China, Qingdao, China
2 shared publications Changshuo Chen
Geovis Technology Company Limited Beijing China
1 shared publication- Fei Huang
State Key Lab of Physical Oceanography/Frontier Science Center for Deep Ocean Multispheres and Earth System, and College of Oceanic and Atmospheric Sciences Ocean University of China Qingdao China
1 shared publication Ge Liu
1 shared publication
Latest Publications
- Open Access
- 15 September 2026
- Published by Elsevier in SSRN Electronic Journal
A Novel Objective Labeling Approach for Arctic Sea Fog Detection from Satellite Imagery Using Fully Convolutional Neural Networks
Sea fog significantly reduces visibility and poses serious risks to maritime activities in the Arctic Ocean. Convolutional neural networks, such as fully convolutional networks (FCN), are widely used for sea fog detection due to their ability to capture textural features from satellite imagery. However, their practical performance remains constrained by the lack of objective and efficient labeling strategies capable of producing region-level labels that encode the spatial continuity of fog patches. We propose a square expansion labeling method designed for convolution-based architectures, which exploits the spatial homogeneity of sea fog to transform one-dimensional CALIPSO observations into two-dimensional labels. Using this method, we develop a comprehensive dataset (2010–2018), where labels are derived from Cloud–Aerosol Lidar and Infrared Pathfinder Satellite Observations (CALIPSO) classification data, and corresponding multispectral inputs are obtained from Moderate Resolution Imaging Spectroradiometer (MODIS) imagery. Based on these objectively generated labels, the FCN-based model achieves a Probability of Detection (POD) of 0.935 and an overall Critical Success Index (CSI) of 0.801 on the test set, thereby demonstrating the feasibility and effectiveness of the proposed labeling strategy for convolutional neural network–based Arctic sea fog detection.
- 15 September 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessA Novel Objective Labeling Approach for Arctic Sea Fog Detection from Satellite Imagery Using Fully Convolutional Neural Networks
Sea fog significantly reduces visibility and poses serious risks to maritime activities in the Arctic Ocean. Convolutional neural networks, such as fully convolutional networks (FCN), are widely used for sea fog detection due to their ability to capture textural features from satellite imagery. However, their practical performance remains constrained by the lack of objective and efficient labeling strategies capable of producing region-level labels that encode the spatial continuity of fog patches. We propose a square expansion labeling method designed for convolution-based architectures, which exploits the spatial homogeneity of sea fog to transform one-dimensional CALIPSO observations into two-dimensional labels. Using this method, we develop a comprehensive dataset (2010–2018), where labels are derived from Cloud–Aerosol Lidar and Infrared Pathfinder Satellite Observations (CALIPSO) classification data, and corresponding multispectral inputs are obtained from Moderate Resolution Imaging Spectroradiometer (MODIS) imagery. Based on these objectively generated labels, the FCN-based model achieves a Probability of Detection (POD) of 0.935 and an overall Critical Success Index (CSI) of 0.801 on the test set, thereby demonstrating the feasibility and effectiveness of the proposed labeling strategy for convolutional neural network–based Arctic sea fog detection.
- Open Access
- 4 August 2026
- Published by Elsevier in SSRN Electronic Journal
SST-BTD as a discriminant of nighttime clear and sea fog/low stratus skies: A case study over the China's coastal region
The dual channel difference method is a standard technique for sea fog and low stratus (SFLS) detection using satellite data at night. However, the brightness temperature difference (BTD) of SFLS can deviate from the conventional fixed threshold in some cases, causing erroneous discrimination between SFLS and clear sea pixels. Using satellite data and radiative transfer simulations, we analyzed the distribution of these missed detection of SFLS in China’s coastal regions, as well as potential factors affecting the BTD of SFLS pixels. Our analysis identified sea surface temperature (SST) as a major factor contributing to the missed detection of SFLS in China’s coastal regions when a fixed threshold is used. To address this issue, we propose an SST-dependent dynamic BTD threshold method (SST-BTD). Compared to the CALIPSO observations, the SST-BTD method achieves a more accurate SFLS detection than the fixed threshold approach, increasing the Probability of Detection (POD) from 0.640 (fixed threshold) to 0.847, while maintaining a low False Alarm Ratio (FAR) of 0.037 and achieving a Critical Success Index (CSI) of 0.821, significantly reducing the missed SFLS pixels.
- 4 August 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessSST-BTD as a discriminant of nighttime clear and sea fog/low stratus skies: A case study over the China's coastal region
The dual channel difference method is a standard technique for sea fog and low stratus (SFLS) detection using satellite data at night. However, the brightness temperature difference (BTD) of SFLS can deviate from the conventional fixed threshold in some cases, causing erroneous discrimination between SFLS and clear sea pixels. Using satellite data and radiative transfer simulations, we analyzed the distribution of these missed detection of SFLS in China’s coastal regions, as well as potential factors affecting the BTD of SFLS pixels. Our analysis identified sea surface temperature (SST) as a major factor contributing to the missed detection of SFLS in China’s coastal regions when a fixed threshold is used. To address this issue, we propose an SST-dependent dynamic BTD threshold method (SST-BTD). Compared to the CALIPSO observations, the SST-BTD method achieves a more accurate SFLS detection than the fixed threshold approach, increasing the Probability of Detection (POD) from 0.640 (fixed threshold) to 0.847, while maintaining a low False Alarm Ratio (FAR) of 0.037 and achieving a Critical Success Index (CSI) of 0.821, significantly reducing the missed SFLS pixels.
- Open Access
- 21 January 2026
- Published by American Geophysical Union (AGU) in Journal of Geophysical Research: Atmospheres
- Vol. 131 (2)
Interannual Variability of Summertime Sea Fog Over North Pacific
The mid‐latitude North Pacific exhibits a sea fog frequency of 20%–40% in the summer (June–August). Here, we show that the interannual variability of the summer sea fog over the mid‐latitude North Pacific region is correlated with the Asian‐Pacific Oscillation (APO) driven by the seasonal heating of the Tibetan Plateau. During the positive phase of APO, the sea fog frequency is observed to be higher along with a stronger North Pacific Subtropical High, North Pacific Trough and South Asian High, while the Bering Sea Low and Okhotsk High are weaker. The stronger northward heat and water vapor fluxes in the positive phase of APO associated with the North Pacific Subtropical High lead to a higher air temperature and relative humidity over the mid‐latitude North Pacific region. Additionally, this northward warm advection is stronger in the upper boundary layer than the cooling by the sea surface temperature in the lower layer, creating a temperature inversion within the marine boundary layer and hence a higher sea fog frequency over the mid‐latitude North Pacific region. Our study shows that the teleconnection of the Tibetan Plateau can be extended to the mid‐latitude northern Pacific through the North Pacific Subtropical High.
- 21 January 2026
- Published by American Geophysical Union (AGU) in Journal of Geophysical Research: Atmospheres
- Vol. 131 (2)
Open AccessInterannual Variability of Summertime Sea Fog Over North Pacific
The mid‐latitude North Pacific exhibits a sea fog frequency of 20%–40% in the summer (June–August). Here, we show that the interannual variability of the summer sea fog over the mid‐latitude North Pacific region is correlated with the Asian‐Pacific Oscillation (APO) driven by the seasonal heating of the Tibetan Plateau. During the positive phase of APO, the sea fog frequency is observed to be higher along with a stronger North Pacific Subtropical High, North Pacific Trough and South Asian High, while the Bering Sea Low and Okhotsk High are weaker. The stronger northward heat and water vapor fluxes in the positive phase of APO associated with the North Pacific Subtropical High lead to a higher air temperature and relative humidity over the mid‐latitude North Pacific region. Additionally, this northward warm advection is stronger in the upper boundary layer than the cooling by the sea surface temperature in the lower layer, creating a temperature inversion within the marine boundary layer and hence a higher sea fog frequency over the mid‐latitude North Pacific region. Our study shows that the teleconnection of the Tibetan Plateau can be extended to the mid‐latitude northern Pacific through the North Pacific Subtropical High.
- Open Access
- 24 April 2023
- Published by American Geophysical Union (AGU) in Geophysical Research Letters
- Vol. 50 (8)
Adapting to a Foggy Future Along Trans‐Arctic Shipping Routes
Rapid retreat of Arctic sea ice extends the area of open ocean for new trans‐Arctic shipping routes. However, the projected routes may be too optimistic in terms of savings in shipping costs from shortened trans‐Arctic distances as they do not consider the increased sea fog frequency (SFF) over areas of the retreating sea ice. We show that delays due to sea fog can be 1–4 days, about 23%–27% along the Northwest Passage and 4%–11% along the Northern Sea Route than previous estimated. We design a route based on the projected sea‐ice extent and SFF. The new route can reduce the sailing time by 0.3–1 day by detouring the routes with lighter impacts of sea fog. More importantly the new route will lower the risk of catastrophic accidents compared to the shortest route and saves the additional costs due to unscheduled port calls.
- 24 April 2023
- Published by American Geophysical Union (AGU) in Geophysical Research Letters
- Vol. 50 (8)
Open AccessAdapting to a Foggy Future Along Trans‐Arctic Shipping Routes
Rapid retreat of Arctic sea ice extends the area of open ocean for new trans‐Arctic shipping routes. However, the projected routes may be too optimistic in terms of savings in shipping costs from shortened trans‐Arctic distances as they do not consider the increased sea fog frequency (SFF) over areas of the retreating sea ice. We show that delays due to sea fog can be 1–4 days, about 23%–27% along the Northwest Passage and 4%–11% along the Northern Sea Route than previous estimated. We design a route based on the projected sea‐ice extent and SFF. The new route can reduce the sailing time by 0.3–1 day by detouring the routes with lighter impacts of sea fog. More importantly the new route will lower the risk of catastrophic accidents compared to the shortest route and saves the additional costs due to unscheduled port calls.
- Open Access
- 20 February 2023
- Published by American Geophysical Union (AGU) in Earth and Space Science
- Vol. 10 (2)
Summer Marine Fog Distribution in the Chukchi–Beaufort Seas
We study the spatial and temporal variability of summer marine fog in the Chukchi–Beaufort region (175°E−150°W, 70°–86°N) using the in‐situ visibility measurements aboard the Chinese research fleet Xuelong (I and II) and a fog product we derive using the Vertical Feature Mask product of the spaceborne Cloud‐Aerosol Lidar and Infrared Pathfinder Satellite Observation (CALIPSO) observations. The Xuelong in‐situ observations show that the fog frequency in the Chukchi–Beaufort region has a maximum of ∼18% in the early morning and is less than 10% in the rest of the day. The latitudinal distribution of the Xuelong‐based in‐situ fog frequency further shows high fog occurrences at 74°N and 79°N, which are related to the local high fog occurrences near 72°–74°N and 76°–80°N in the central Chukchi–Beaufort region, as revealed by the longitude‐latitude pattern of the CALIPSO‐based spaceborne fog frequency distribution. The CALIPSO‐based fog frequency is also shown to be lower along the continental coastlines than in the Chukchi–Beaufort region. This longitude‐latitude distribution may be explained by a reduced fog formation due to the Pacific warm current flowing into the Arctic region through the Bering Strait in the summer as well as an enhanced fog formation in the Chukchi–Beaufort region when the southward flow of the Beaufort Gyre interacts with the Pacific warm current.
- 20 February 2023
- Published by American Geophysical Union (AGU) in Earth and Space Science
- Vol. 10 (2)
Open AccessSummer Marine Fog Distribution in the Chukchi–Beaufort Seas
We study the spatial and temporal variability of summer marine fog in the Chukchi–Beaufort region (175°E−150°W, 70°–86°N) using the in‐situ visibility measurements aboard the Chinese research fleet Xuelong (I and II) and a fog product we derive using the Vertical Feature Mask product of the spaceborne Cloud‐Aerosol Lidar and Infrared Pathfinder Satellite Observation (CALIPSO) observations. The Xuelong in‐situ observations show that the fog frequency in the Chukchi–Beaufort region has a maximum of ∼18% in the early morning and is less than 10% in the rest of the day. The latitudinal distribution of the Xuelong‐based in‐situ fog frequency further shows high fog occurrences at 74°N and 79°N, which are related to the local high fog occurrences near 72°–74°N and 76°–80°N in the central Chukchi–Beaufort region, as revealed by the longitude‐latitude pattern of the CALIPSO‐based spaceborne fog frequency distribution. The CALIPSO‐based fog frequency is also shown to be lower along the continental coastlines than in the Chukchi–Beaufort region. This longitude‐latitude distribution may be explained by a reduced fog formation due to the Pacific warm current flowing into the Arctic region through the Bering Strait in the summer as well as an enhanced fog formation in the Chukchi–Beaufort region when the southward flow of the Beaufort Gyre interacts with the Pacific warm current.