Severe air pollution plagues Arequipa, Peru due to anthropogenic and natural emissions. Persistent volcano emission in the vicinity of Arequipa makes it among the largest SO₂ sources in the world. Since volcano plumes mostly exist in the free troposphere and stratosphere where horizontal transport acts rather quickly, previous studies mostly focused on their global scale impacts. Whether these plumes can affect near surface air quality has not attracted much research attention. This study uses WRF-Chem simulations to reveal that in the presence of northerly/northwesterly winds and favorable mountain meteorology, the plume from volcano Sabancaya (elevation 5960 m, ~80 km north of Arequipa) can be brought down to near surface of Arequipa through two steps of transport and dispersion processes: 1) With northerly/northwesterly winds, the free troposphere plume from Sabancaya is transported southward and intercepted by mountain Chachani located between Sabancaya and Arequipa and subsequently transported downward to Arequipa by nighttime downslope winds linked to large-amplitude lee-side mountain gravity waves. Often the plume reaches down to be close to the boundary layer over Arequipa. 2) In the following day, convective boundary layer growth brings the above-boundary-layer plume to near the surface through vertical mixing processes, thus exacerbating ambient air pollution in Arequipa. A mechanism on how volcano plumes above 6 km height cause air pollution over the lower-lying Arequipa city is therefore revealed for the first time. The mountain dynamic effect in inducing the large-amplitude mountain lee waves is further illustrated by an idealized simulation excluding mountain’s thermal effect.

Planetary boundary layer (PBL) structure over the ocean and the model capability to simulate such structure are less well-understood than their counterparts over land. In this study, observations and WRF simulations are examined to study the boundary layer structure over the Southern Ocean, focusing on the coupling between the oceanic boundary layer and the cloud layer above. Based on the lower tropospheric vertical profiles and cross-sections, three cloud-boundary layer coupling modes are identified including a coupled mode with a weak positive surface heat flux (type 1), and two decoupled modes in the presence of either a negative surface heat flux driving a shallow stable boundary layer (type 2) or a strong positive surface heat flux (type 3). Numerical simulations are conducted for representative cases of each mode using the conventional YSU PBL scheme without and with the cloud-induced top-down mixing option (referred to as YSUtopdown), as well as the MYNN and the MYNN eddy-diffusivity mass-flux scheme (MYNN-EDMF) that adopts a holistic treatment of mixed-layer thermals and shallow convective clouds. The MYNN-EDMF scheme offers the best representation of the decoupled type 3 mode where its capability to simulate different vertical extents of local mixing and nonlocal mass flux is found to be essential. Two key parameters in MYNN-EDMF dictating shallow cloud formation are also identified. The YSUtopdown scheme develops deeper boundary layer than the YSU scheme and exhibits more consistency with observations for the coupled type 1 mode. For the decoupled type 2 mode, all four schemes perform similarly well.

Detailed convective boundary layer (CBL) structure and the impact factors over the Tibetan Plateau has not been clearly understood, particularly for the level of neutral stability (zn), at which statically unstable lower CBL begins to transit into slightly stable upper CBL. Substantial uncertainties still exist in numerical models with different planetary boundary layer (PBL) schemes to reproduce such detailed structure. In this study, detailed CBL structure and processes over the Tibetan Plateau are examined using multi-year radiosonde data and large-eddy simulation (LES), particularly focusing on the impact of surface heating and entrainment on zn. The results indicated that the values of zn spatially ranged within 0.16–0.38zi on the plateau, with zi representing the CBL depth, and zn was higher in the southwestern region and lower in the southeastern region. Surface-/entrainment-induced large-scale thermals (corresponding to nonlocal fluxes) tended to suppress/elevate zn, due to warm turbulence penetrating into the upper/lower CBL, whereas small-scale eddies (corresponding to local fluxes) played an opposite role on modifying zn. The LES results suggested that zn increased before 08:00 Local Time (about 80 minutes after sunrise) because surface-induced small eddies dominated during the early stage of CBL growth and zn decreased afterwards as large-scale surface-induced thermals became more active. These improved understanding provides guidance for further improvement of PBL schemes.