Search

Skillful subseasonal forecasts are crucial for various sectors of society but pose a grand scientific challenge. Recently, machine learning-based weather forecasting models outperform the most successful numerical weather predictions generated by the European Centre for Medium-Range Weather Forecasts (ECMWF), but have not yet surpassed conventional models at subseasonal timescales. This paper introduces FuXi Subseasonal-to-Seasonal (FuXi-S2S), a machine learning model that provides global daily mean forecasts up to 42 days, encompassing five upper-air atmospheric variables at 13 pressure levels and 11 surface variables. FuXi-S2S, trained on 72 years of daily statistics from ECMWF ERA5 reanalysis data, outperforms the ECMWF's state-of-the-art Subseasonal-to-Seasonal model in ensemble mean and ensemble forecasts for total precipitation and outgoing longwave radiation, notably enhancing global precipitation forecast. The improved performance of FuXi-S2S can be primarily attributed to its superior capability to capture forecast uncertainty and accurately predict the Madden-Julian Oscillation (MJO), extending the skillful MJO prediction from 30 days to 36 days. Moreover, FuXi-S2S not only captures realistic teleconnections associated with the MJO but also emerges as a valuable tool for discovering precursor signals, offering researchers insights and potentially establishing a new paradigm in Earth system science research. This paper introduces FuXi-S2S, a machine-learning model that outperforms conventional numerical weather prediction models at subseasonal timescales globally, extending the skillful Madden–Julian Oscillation prediction form 30 days to 36 days. [ABSTRACT FROM AUTHOR]

In the face of an aging population and declining birth rates, urban Chinese families are grappling with a new challenge of simultaneously caring for both elderly parents and children. This study employs a three-generational analytical framework and a mixed research method to examine the intergenerational distribution of care resources and its consequences for urban Chinese families. The findings reveal intergenerational competition for care resources within families, with a general tendency to prioritize children and only respond to crisis situations for elderly parents; this leads to a care deficit for older individuals. Furthermore, social care resources are ineffective in bridging this care gap, thereby inducing anxiety within families as they strive to balance the care needs of both elderly parents and children. The study argues that insufficient support from care policies exacerbates the resource squeeze on the elderly side by the children's side within families, which calls for enhancing the policy support system to help families balance eldercare, childcare, and work commitments. [ABSTRACT FROM AUTHOR]

NaTi2(PO4)3 (NTP) is considered a promising anode material for sodium-ion batteries due to its open and stable framework structure. However, its poor conductivity prevents its direct application as an anode material. In this paper, a dual strategy of La3+ doping and carbon coating is proposed to improve the conductivity of NTP. The results show that the NTP-La0.06@C electrode doped with 6.00% La and 5.00 wt% of carbon coating has excellent electrochemical performance, and the coulombic efficiency is up to 95.31% in the first cycle. The initial specific capacity is 103.05 mAh·g−1 with a capacity retention of 97.30% after 1000 cycles at 10C rate. Even after 2000 cycles, there is still a specific capacity of 98.62 mAh·g−1 with a capacity retention of 92.71%, and as well as the NTP-La0.06@C has a higher Na+-insert diffusion coefficient of 3.19×10−11 cm2·s−1 than the pristine NTP@C of 2.35×10−11 cm2·s−1 during charging. Therefore, the La3+ doping and carbon coating strategy shows great potential in improving the performance of NaTi2(PO4)3 anode materials. [ABSTRACT FROM AUTHOR]

A solar heat pump based on the photovoltaic photothermal (PV/T) module is a new technology that can improve the photovoltaic efficiency and recovery of waste heat in photovoltaic conversion. The comprehensive efficiency of a system can thus be greatly improved. At present, there is little research on the simulation of a solar heat pump based on the Simulink 2018 software. In this paper, PV/T modules are modeled and simulated using the Simulink software based on the typical meteorological parameters in Beijing city during summer and winter conditions. Considering the system characteristics of a simultaneous operation, the models of all components such as the compressor, condenser, evaporator, expansion valve, and heat storage tanks contained in a system are established based on the system's working principle, energy conservation equations, and some empirical formulas and then simulated. As PV/T modules are the key influencing factor for system performance, the model and simulation process are introduced in detail in this paper. The results show that the surface temperature of PV/T modules under different operating conditions are all significantly lower than that of ordinary photovoltaic panels. The average temperature of PV/T modules is only 15.2 °C, which is 20.2 °C lower than that of ordinary panels under summer conditions, and the average temperature of PV/T modules is only 0.5 °C, 17.8 °C lower than that of ordinary panels under winter conditions. The average photoelectric efficiency can be improved by 15.4% and 8.9%. And, the temperature change amplitude of PV/T modules is lower, which weakens the temperature fluctuation of the modules. The photovoltaic efficiency is therefore further increased. As a result, the power generation and photovoltaic efficiency are both improved significantly. [ABSTRACT FROM AUTHOR]

With the increasing demand for logistics in modern society, how to achieve low-cost and efficient logistics delivery has become an urgent research topic. A hybrid evolutionary JAYA algorithm (H-JAYA) based on global optimization was designed to address the complex path planning problem of electric vehicles. This algorithm introduces a reverse learning mechanism to calculate the current optimal and worst individuals, while using differential perturbation mechanism and sine cosine operator to update the individual's position. In addition, the study used the H-JAYA algorithm to construct a corresponding mathematical model for the optimization problem of electric vehicle paths. The results showed that in the three examples, the H-JAYA algorithm tested the optimal curve convergence speed, and it tended to stabilize after about 30 iterations. Meanwhile, in the RCDP5001 example, the total cost of the H-JAYA algorithm reached the lowest value of 623 yuan. The H-JAYA algorithm has significant advantages in solving the distribution path problem of electric vehicles, and can be well applied to practical logistics distribution, providing effective technical support for modern e-commerce logistics planning. [ABSTRACT FROM AUTHOR]

High‐resolution and accurate prediction of near‐surface weather parameters based on numerical weather prediction (NWP) models is essential for many downstream and real‐world applications. Traditional dynamical or statistical downscaling methods are insufficient to derive high‐resolution data from operational NWP forecasts, making it essential to devise new approaches. In recent years, an increasing number of researchers have explored the implementations of deep learning (DL) based models for spatial downscaling, motivated by the similarity between the super‐resolution (SR) problem in computer vision (CV) and downscaling. Furthermore, while transformer‐based models have become state‐of‐the‐art models for many SR tasks, they are rarely applied for downscaling of weather forecasts or climate projections. This study adapted transformer‐based models such as SwinIR and Uformer to downscale the temperature at 2 m (T2m$$ {T}_{2\mathrm{m}} $$) and wind speed at 10 m (WS10m$$ W{S}_{10\mathrm{m}} $$) over Eastern Inner Mongolia, encompassing the area from 39.6–46°N latitude and 111.6–118°E longitude. We used high‐resolution forecast (HRES) data from the European Centre for Medium‐range Weather Forecast (ECMWF) with a spatial resolution of 0.1° as the input and gridded observation data from the China Meteorological Administration (CMA) Land Data Assimilation System (CLDAS) at a spatial resolution of 0.01° as the target. Given that the models use observation data rather than a coarse‐grained version of forecast data as the target, they accomplish both bias correction and spatial downscaling. The results demonstrate that the performance of SwinIR and Uformer is superior to that of two convolutional neural network (CNN) based models (UNet and RCAN). Additionally, we introduced a novel module to extract features of varying resolution from the high‐resolution topography data and applied a multiscale feature fusion module to merge features of different scales, contributing to further enhancement of Uformer's performance. [ABSTRACT FROM AUTHOR]

• The relationship between magnetic field intensity and frost suppression effect is not linear. • As the temperature of the cold surface decreases, the suppressing freezing ability of the magnetic field becomes worse. • The initial freezing time of the droplet decreases as its volume increases from 20 μL to 50 μL. In this study, the effect of magnetic field on droplet freezing was investigated based on a visualization system. Possible influencing factors on droplet freezing, such as magnetic field intensity, cold surface temperature, and droplet volume, were considered in this study. During the two stages of droplet supercooling and freezing process, the changes of the droplet internal temperature, the formation of initial crystal nucleus, and the freezing time were observed. Experimental results show that in comparison with the non-magnetic surface, the supercooling degree of droplets was increased in the range of 150–300 mT magnetic field intensity, thereby prolonging the freezing time of droplets. Particularly, the droplet was best suppressed on a cold surface with 250 mT magnetic field intensity. The freezing time of droplets was prolonged by 132.0 %. However, the droplets were accelerated to freeze on the cold surface with magnetic field intensities of 350 mT and 400 mT. Droplet freezing on cold surfaces was also affected by surface temperature and droplet volume. As the surface temperature of magnetic cooling dropped from −10 °C to −20 °C, the ability of the magnetic field to inhibit droplet freezing would be gradually weakened. With the increase of droplet volume from 20 μL to 50 μL, the initial freezing time of droplets was shortened. These results are related to the influence of magnetic field on the nucleation process and hydrogen bond in the water molecular cluster. [ABSTRACT FROM AUTHOR]