HSHua Song

Hua Song

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Hua Song has a total of 22 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Substantial Diel Changes of Cloud Adjustments to Aerosols in Ship‐Tracks

    by Tianle Yuan, Hua Song, Robert Wood, Lazaros Oreopoulos, Kerry Meyer, William Smith, Ryan Eastman

    Clouds adjust their albedo, amount and liquid water path to anthropogenic aerosols, and produce a climate forcing. Such cloud adjustments are complex functions of both environmental conditions and interaction time. However, direct observations of temporal changes in cloud adjustments are limited. Here, we develop a method to show observational evidence of substantial diel changes in cloud adjustments within ship‐tracks. Polluted clouds experiencing extended nighttime exposure exhibit higher increases in cloud fraction while liquid‐water path adjustments follow a non‐monotonic evolution. The higher cloud fraction increase persists through daytime. Such diel variations likely stem from diel changes in solar heating and background cloud precipitation, and are important because nighttime changes persist and affect solar radiation during the day. Accounting for diel variations produces a stronger aerosol forcing, by . Our approach reveals how the temporal evolution of cloud adjustments can have important implications for aerosol‐cloud interactions.

  • Open Access

    Analyses of Virtual Ship‐Tracks Systematically Underestimate Aerosol‐Cloud Interactions Signals

    by Tianle Yuan, Hua Song, Lazaros Oreopoulos, Robert Wood, Kerry Meyer, Alice Crawford, William Smith, Ryan Eastman

    Ship‐tracks are important natural/opportunistic experiments to study aerosol‐cloud interactions (ACIs). However, detectable ship‐tracks are not produced in many instances. Virtual ship‐tracks have been conceived to expand the scale of ACIs analyses. Cloud responses in virtual ship‐tracks differ strongly from those of detected ones. Here we show that the current approach of virtual ship‐tracks can lead to systematic biases and errors and suggest necessary improvements. Errors in trajectory modeling introduce mismatches between areas actually affected by ship‐emissions and virtual ship‐track locations, that is, positional errors. Positional errors systematically underestimate ACI signals and the underestimate is severe as indicated by analysis of cloud droplet number concentration changes. The assumption of fixed ship‐track width also systematically diminishes resulting aerosol effects by more than 10%, which leads to a forcing difference of around 0.1 . We make suggestions to improve the simulation of virtual ship‐tracks so that their full potential for studying ACIs can be unleashed.


  • Open Access

    Detecting ship-produced NO2 plumes and shipping routes in TROPOMI data with a deep learning model

    by Tianle Yuan, Fei Liu, Lok N. Lamsal, Hua Song

    Ship emissions are important contributor to air pollution and the climate through interacting with clouds. They are the dominant anthropogenic source over the oceans. However, their magnitudes still have large uncertainty. Here we develop a deep learning model to detect ship-emitted NO2 plumes at the pixel level in TROPOMI tropospheric NO2 data. The trained model performs well and, when applied to a year of data, it finds major shipping routes, but misses several other routes. We show that high cloudiness in these shipping routes is the culprit because clouds block signals from reach the sensor. Indeed, detected shipping routes in this study complements shipping routes detected using ship-tracks that is only available in cloudy regions. Our method can find application in several areas such as improving ship emission estimates and compliance verifications. Our method will benefit from improved tropospheric NO2 retrievals since their quality is critical for plume detection.

  • Open Access

    Global Reduction in Ship-tracks from Sulfur Regulations for Shipping Fuel

    by Tianle Yuan, Hua Song, Robert Wood, Chenxi Wang, Lazaros Oreopoulos, Steven Platnick, Sophia von Hippel, Kerry Meyer, Siobhan Light, Eric Wilcox

    Ship-tracks are produced by ship-emitted aerosols interacting with marine low clouds. Here we apply deep learning models on satellite data to produce the first multi-year global climatology map of ship-tracks. We show that ship-tracks are at the nexus of cloud physics, maritime shipping, and fuel regulation. Our map captures major shipping lanes while missing others, reflecting the influences of background cloud and aerosol properties. Ship-track frequency is more than 10 times higher than expected from a previous survey. Interannual fluctuations in ship-track frequency reflect variations in cross-ocean trade, shipping activity, and fuel regulations. Fuel regulation can alter both detected ship-track density and pattern of shipping routes due to cost economics. The new fuel regulation, together with the COVID-19 pandemic, reduced ship-track frequency in 2020 to its lowest level in recent decades across the globe and may have ushered in a new era of low ship-track frequency. We estimate the aerosol indirect forcing induced by the fuel regulation to be between 0.02 and 0.22 Wm-2.

  • Open Access

    Automatically Finding Ship Tracks to Enable Large‐Scale Analysis of Aerosol‐Cloud Interactions

    by Tianle Yuan, Chenxi Wang, Hua Song, Steven Platnick, Kerry Meyer, Lazaros Oreopoulos

    Ship tracks appear as long winding linear features in satellite images and are produced by aerosols from ship exhausts changing low cloud properties. They are one of the best examples of aerosol‐cloud interaction experiments. However, manually finding ship tracks from satellite data on a large scale is prohibitively costly while a large number of samples are required to improve our understanding. Here we train a deep neural network to automate finding ship tracks. The neural network model generalizes well as it not only finds ship tracks labeled by human experts but also detects those that are occasionally missed by humans. It finds more ship tracks than all previous studies combined and produces a map of ship track distributions off the California coast that matches well with known shipping traffic. Our technique will enable studying aerosol effects on low clouds using ship tracks on a large scale, which will potentially narrow the uncertainty of the aerosol‐cloud interactions.


  • Intercomparisons of marine boundary layer cloud properties from the ARM CAP‐MBL campaign and two MODIS cloud products

    by Zhibo Zhang, Xiquan Dong, Baike Xi, Hua Song, Po‐Lun Ma, Steven J. Ghan, Steven Platnick, Patrick Minnis

    From April 2009 to December 2010, the Department of Energy Atmospheric Radiation Measurement (ARM) program carried out an observational field campaign on Graciosa Island, targeting the marine boundary layer (MBL) clouds over the Azores region. In this paper, we present an intercomparison of the MBL cloud properties, namely, cloud liquid water path (LWP), cloud optical thickness (COT), and cloud‐droplet effective radius (CER), among retrievals from the ARM mobile facility and two Moderate Resolution Imaging Spectroradiometer (MODIS) cloud products (Goddard Space Flight Center (GSFC)‐MODIS and Clouds and Earth's Radiant Energy System‐MODIS). A total of 63 daytime single‐layer MBL cloud cases are selected for intercomparison. Comparison of collocated retrievals indicates that the two MODIS cloud products agree well on both COT and CER retrievals, with the correlation coefficient R > 0.95, despite their significant difference in spatial sampling. In both MODIS products, the CER retrievals based on the 2.1 µm band (CER2.1) are significantly larger than those based on the 3.7 µm band (CER3.7). The GSFC‐MODIS cloud product is collocated and compared with ground‐based ARM observations at several temporal‐spatial scales. In general, the correlation increases with more precise collocation. For the 63 selected MBL cloud cases, the GSFC‐MODIS LWP and COT retrievals agree reasonably well with the ground‐based observations with no apparent bias and correlation coefficient R around 0.85 and 0.70, respectively. However, GSFC‐MODIS CER3.7 and CER2.1 retrievals have a lower correlation (R ~ 0.5) with the ground‐based retrievals. For the 63 selected cases, they are on average larger than ground observations by about 1.5 µm and 3.0 µm, respectively. Taking into account that the MODIS CER retrievals are only sensitive to cloud top reduces the bias only by 0.5 µm.