Experimental analysis of soil moisture response to rainfall in a typical grassland hillslope under different vegetation treatments.

The responses of soil moisture to rainfall are of great significance for watershed hydrological modeling. However, few studies have been done to investigate these responds on hillslope in a typical semi-arid grassland region. This study used high temporal resolution soil moisture data to explore the soil moisture dynamics, response conditions and its controls of 0-40 cm soil profile in the upslope (14°), midslope (9°), and downslope (4°) of a typical grassland inland river basin under bare ground (BG), stubble (SG), and natural grassland (CK) treatments. The results showed that soil water content and water storage increased in the downslope direction, and all showed as BG > SG > CK. The dry and wet changes in fast-changing layer (5 cm) and active layer (10 cm) were rapid, while soil moisture below 20 cm was relatively stable and fluctuated only in heavy or continuous rainfalls. The soil moisture response process varied greatly under different rainfall, rainfall intensity and antecedent soil moisture conditions, which explained 41.1% of the total difference. The rainfall replenishment threshold and the required initial soil profile water content of soil moisture response in 5 cm, 10 cm and 20 cm soil layers were 5.8 mm, 8.0 mm, 11.4 mm and 8.7 vol%, 9.4 vol%, 10.8 vol%, respectively. Soil properties, vegetation characteristics and topography could explain 38.8%, 14.5% and 5.6% of the soil moisture variation on the hillslope. In addition, under the comprehensive influence of environmental factors, changes in soil moisture of the upslope were significantly affected by soil sand content, the differences in the midslope were mainly due to soil clay content and belowground biomass, whereas the vegetation characteristics were the main factors in the downslope. This study can contribute to the further understanding of slope-scale ecohydrological processes and hydrological simulation of semi-arid grassland watersheds.

Dynamic predictability and activity-location contexts in human mobility.

Human travelling behaviours are markedly regular, to a large extent predictable, and mostly driven by biological necessities and social constructs. Not surprisingly, such predictability is influenced by an array of factors ranging in scale from individual preferences and choices, through social groups and households, all the way to the global scale, such as mobility restrictions in response to external shocks such as pandemics. In this work, we explore how temporal, activity and location variations in individual-level mobility-referred to as predictability states-carry a large degree of information regarding the nature of mobility regularities at the population level. Our findings indicate the existence of contextual and activity signatures in predictability states, suggesting the potential for a more nuanced approach to estimating both short-term and higher-order mobility predictions. The existence of location contexts, in particular, serves as a parsimonious estimator for predictability patterns even in the case of low resolution and missing data.

Contrasting social and non-social sources of predictability in human mobility.

Social structures influence human behavior, including their movement patterns. Indeed, latent information about an individual's movement can be present in the mobility patterns of both acquaintances and strangers. We develop a "colocation" network to distinguish the mobility patterns of an ego's social ties from those not socially connected to the ego but who arrive at a location at a similar time as the ego. Using entropic measures, we analyze and bound the predictive information of an individual's mobility pattern and its flow to both types of ties. While the former generically provide more information, replacing up to 94% of an ego's predictability, significant information is also present in the aggregation of unknown colocators, that contain up to 85% of an ego's predictive information. Such information flow raises privacy concerns: individuals sharing data via mobile applications may be providing actionable information on themselves as well as others whose data are absent.

Facile fabrication of composited solid phase microextraction thin membranes for sensitive detections of trace hydroxylated polycyclic aromatic hydrocarbons in human urine.

Solid phase microextraction (SPME) has potential to be used for the high-performance enrichments of hydroxylated polycyclic aromatic hydrocarbons (OH-PAHs), which are important biomarkers of PAH exposure. By choosing suitable adsorbent, it is conducive to fabricate new-type of SPME device for improved extraction efficiencies towards OH-PAHs. In this study, a novel method of surface solvent evaporation has been proposed to fabricate SPME thin membrane, integrating the advantages of polydimethylsiloxane (PDMS) and different porous adsorbents. The powdery metal organic framework (Uio66-NH2), porous polymer (powdery polymer aerogel, PPA) and ordered mesoporous carbon (OMC) have been chosen as typical adsorbents and fabricated as thin membranes successfully, indicating the universality of the proposed method for membrane fabrication. Comparing the extraction efficiencies of three prepared membranes towards OH-PAHs, the OMC-PDMS membrane has demonstrated best enrichment efficiencies. The OMC-PDMS membrane was used for the enrichments of trace OH-PAHs in human urine of both smokers and nonsmokers, combining with liquid chromatographic tandem mass spectrometry (LC-MS/MS). The detection limits were in the range of 0.15-0.39 ng L-1, and satisfactory recoveries were found to be 82.1%-115%. It can be seen that the universal strategy to fabricate SPME membrane is helpful to achieve broad-spectrum or selective enrichments of target analytes from complex matrix by simple modulation of membrane components.

Tuning Fairness by Balancing Target Labels.

The issue of fairness in machine learning models has recently attracted a lot of attention as ensuring it will ensure continued confidence of the general public in the deployment of machine learning systems. We focus on mitigating the harm incurred by a biased machine learning system that offers better outputs (e.g., loans, job interviews) for certain groups than for others. We show that bias in the output can naturally be controlled in probabilistic models by introducing a latent target output. This formulation has several advantages: first, it is a unified framework for several notions of group fairness such as Demographic Parity and Equality of Opportunity; second, it is expressed as a marginalization instead of a constrained problem; and third, it allows the encoding of our knowledge of what unbiased outputs should be. Practically, the second allows us to avoid unstable constrained optimization procedures and to reuse off-the-shelf toolboxes. The latter translates to the ability to control the level of fairness by directly varying fairness target rates. In contrast, existing approaches rely on intermediate, arguably unintuitive, control parameters such as covariance thresholds.