Search

Remote sensing data play a critical role in improving numerical weather prediction (NWP). However, the physical principles of radiation dictate that data voids frequently exist in physical space (e.g., subcloud area for satellite infrared radiance or no-precipitation region for radar reflectivity). Such data gaps impair the accuracy of initial conditions derived from data assimilation (DA), which has a negative impact on NWP. We use the barotropic vorticity equation to demonstrate the potential of deep learning augmented data assimilation (DDA), which involves reconstructing spatially complete pseudo-observation fields from incomplete observations and using them for DA. By training a convolutional autoencoder (CAE) with a long simulation at a coarse “forecast” resolution (T63), we obtained a deep learning approximation of the “reconstruction operator,” which maps spatially incomplete observations to a model state with full spatial coverage and resolution. The CAE was applied to an incomplete streamfunction observation (∼30% missing) from a high-resolution benchmark simulation and demonstrated satisfactory reconstruction performance, even when only very sparse (1/16 of T63 grid density) observations were used as input. When only spatially incomplete observations are used, the analysis fields obtained from ensemble square root filter (EnSRF) assimilation exhibit significant error. However, in DDA, when EnSRF takes in the combined data from the incomplete observations and CAE reconstruction, analysis error reduces significantly. Such gains are more pronounced with sparse observation and small ensemble size because the DDA analysis is much less sensitive to observation density and ensemble size than the conventional DA analysis, which is based solely on incomplete observations. Significance Statement Data assimilation plays a critical role in improving the skills of modern numerical weather prediction by establishing accurate initial conditions. However, unobservable regions are common in observation data, particularly those derived from remote sensing. The nonlinear relationship between data from observable regions and the physical state of unobservable regions may impede DA efficiency. As a result, we propose that deep learning be used to improve data assimilation in such cases by reconstructing a spatially complete first guess of the physical state with deep learning and then applying data assimilation to the reconstructed field. Such deep learning augmentation is found effective in improving the accuracy of data assimilation, especially for sparse observation and small ensemble size.

The Pearl River delta (PRD) region has undergone rapid urbanization since the 1980s, which has had significant effects on the sea-breeze circulation in this region. Because the sea breeze plays an important role in pollutant transportation and convective initiation in the PRD region, it is meaningful to study the effects of urbanization on the sea breeze. In this study, three numerical experiments were conducted from 2 June to 31 August 2010 with land-use data from 1988, 1999, and 2010. For each simulation, characteristics of the sea breeze such as the start time, end time, intensity, height, pumping ability, and inland penetration distance were quantified. By comparing the characteristics of the sea breeze in these simulations, its response to urbanization was quantified. The results show that urbanization enhances the duration, height, and intensity of the sea breeze but blocks its inland penetration. One physical mechanism is proposed to dynamically elucidate the response of the sea breeze to urbanization. Because the urban area in the PRD region is concentrated near the coast, urbanization imposes a positive heating gradient on the coastal region and a negative heating gradient on the region farther inland. The positive heating gradient may intensify the sea breeze, and the negative heating gradient may prevent the sea breeze from propagating farther inland.

The Pearl River delta (PRD) region has experienced rapid economic development since the 1980s and has become one of the world’s largest industrial zones and metropolitan areas. Previous studies have shown that the sea-breeze circulation can contribute to pollutant transportation and convective initiation, so it is useful to study the dynamic structure of the sea-breeze circulation in the PRD region. Many researchers have focused on the effects of environmental factors, such as topography, urbanization, and background wind, on the sea breeze, but most focused only on case studies and did not quantify the characteristics of the sea-breeze circulation climatologically. In this study, a sea-breeze identification metric was defined to identify sea-breeze events from WRF simulation data of 2012 and quantify their characteristics, including their start time, end time, strength, height, frequency, pumping ability, and inland-penetrating distance. The results indicate that this method works well to identify and quantify the sea-breeze events of 2012. It is found that the solenoid term, the largest positive contributor to vorticity acceleration, is mostly modulated by the temperature gradient. Therefore, the frontogenesis of the sea-breeze front is discussed in this study. The result shows that offshore background wind that increases frontogenesis is favorable to the development of the sea breeze, but it also prevents it from propagating vertically and horizontally.

The Pearl River Delta (PRD) region, located in the southern part of Guangdong Province in China, is one of the most rapidly developing regions in the world. The evolution of local and regional sea-breeze circulation (SBC) is believed to be responsible for forming meteorological conditions for high air-pollution episodes in the PRD. To understand better the impacts of urbanization and its associated urban heat island (UHI) on the local- and regional-scale atmospheric circulations over PRD, a number of high-resolution numerical experiments, with different approaches to treat the land surface and urban processes, have been conducted using the fifth-generation Pennsylvania State University–National Center for Atmospheric Research Mesoscale Model (MM5). The results show that an accurate urban land-use dataset and a proper urban land-use parameterization are critical for the mesoscale model to capture the major features of the observed UHI effect and land–sea-breeze circulations in the PRD. Stronger UHI in the PRD increases the differential temperature gradient between urbanized areas and nearby ocean surface and hence enhances the mesoscale SBC. The SBC front consequently penetrates farther inland to overcome the prevailing easterly flow in the western part of inland Hong Kong. Additional sensitivity studies indicate that further industrial development and urbanization will strengthen the daytime SBC as well as increase the air temperature in the lowest 2 km of the atmosphere.