Deep reinforcement learning (DRL) is promising for solving control problems in fluid mechanics, but it is a new field with many open questions. Possibilities are numerous and guidelines are rare concerning the choice of algorithms or best formulations for a given problem. Besides, DRL algorithms learn a control policy by collecting samples from an environment, which may be very costly when used with Computational Fluid Dynamics (CFD) solvers. Algorithms must therefore minimize the number of samples required for learning (sample efficiency) and generate a usable policy from each training (reliability). This paper aims to (a) evaluate three existing algorithms (DDPG, TD3, and SAC) on a fluid mechanics problem with respect to reliability and sample efficiency across a range of training configurations, (b) establish a fluid mechanics benchmark of increasing data collection cost, and (c) provide practical guidelines and insights for the fluid dynamics practitioner. The benchmark consists in controlling an airfoil to reach a target. The problem is solved with either a low-cost low-order model or with a high-fidelity CFD approach. The study found that DDPG and TD3 have learning stability issues highly dependent on DRL hyperparameters and reward formulation, requiring therefore significant tuning. In contrast, SAC is shown to be both reliable and sample efficient across a wide range of parameter setups, making it well suited to solve fluid mechanics problems and set up new cases without tremendous effort. In particular, SAC is resistant to small replay buffers, which could be critical if full-flow fields were to be stored.

Oil palm is one of Southeast Asia’s most common crops, and its expansion has caused substantial modification of natural habitats and put increasing pressure on biodiversity. Rising global demand for vegetable oil, coupled with oil palm’s high yield per unit area and the versatility of the palm oil product, has driven the expansion of oil palm agriculture in the region. Therefore, it is critical to identify management practices that can support biodiversity in plantations without exacerbating negative impacts on the environment. This study focuses on day-flying Lepidoptera (butterflies and moths), which contribute to the ecosystem functioning as pollinators, prey, and herbivore species. We assessed whether density and behaviours of day-flying Lepidoptera varied between different habitats within oil palm plantations and across seasons. We surveyed the density and behaviours of Lepidoptera communities in mature industrial oil palm plantations within the Biodiversity and Ecosystem Function in Tropical Agriculture (BEFTA) Programme sites, in Riau, Indonesia. We surveyed two distinct habitats within the plantations in March and September 2013: Edge habitats, which were bordered by plantation roads on one side, and Core habitats in the centre of oil palm planting blocks. We conducted analyses on the effect of habitat type and season on both the overall density and behaviour of Lepidoptera communities and, independently, on the most common species. In our surveys, we observed 1464 individuals across 41 species, with a significantly higher density in Edge than in Core habitats. While there was no significant difference between overall density in March and September surveys, there was an interaction between season and habitat, with density increasing more markedly in Edge than Core areas in September. There was also a significant effect of habitat and season on behavioural time budget for the community as a whole, with more active behaviours, such as foraging and mating, being recorded more frequently in Edge than Core habitats, and more commonly in September than March. The effect of habitat type, season, and their interaction differed between the six most common species. Our findings indicate that Lepidoptera abundance is affected by habitat characteristics in a plantation and can therefore be influenced by plantation management practices. In particular, our study highlights the value of road edges and paths in plantations for day-flying Lepidoptera. We suggest that increased non-crop vegetation in these areas, achieved through reduced clearing practices or planting of flowering plants, could foster abundant and active butterfly communities in plantations. These practices could form part of sustainability management recommendations for oil palm, such as those of the Roundtable on Sustainable Palm Oil.

We aimed to descriptively analyse the possible impact of the national COVID-19 interventions on the incidence of common infectious diseases in Denmark during spring and summer 2020. This observational study focused on national register data on infections caused by 16 different bacterial and viral pathogens. We included new cases registered between 1 January 2016 and 31 July 2020. The weekly number of new cases were analysed with respect to the COVID-19-related interventions introduced during 2020. We found a marked decrease in infections associated with droplet transmission coinciding with the COVID-19 interventions in spring and summer 2020. These included decreases in both viral and bacterial airway infections and also decreases in invasive infections caused by Streptococcus pneumoniae, Haemophilus influenzae and Neisseria meningitidis. There was also a reduction in cases associated with foodborne transmission during the COVID-19 lockdown period. We found no effect of the lockdown on infections by invasive beta-haemolytic streptococci group B, C and G, Staphylococcus aureus bacteraemia, Neisseria gonorrhoeae or Clostridioides difficile. In conclusion, we found that the widespread interventions such as physical distancing, less travel, hygiene measures and lockdown of schools, restaurants and workplaces together coincided with a marked decline in respiratory infections and, to a smaller extent, some foodborne-transmitted infections.