Sea surface temperature (SST) is one of the most important parameters in the global ocean-atmosphere system, and its long-term changes will have a significant impact on global climate and ecosystems. Accurate prediction of SST, therefore, especially the improvement of long-term predictive skills is of great significance for fishery farming, marine ecological protection, and planning of maritime activities. Since the effective and precise description of the long-range dependence between input and output requires higher model prediction ability, it is an extremely challenging task to achieve accurate long-term prediction of SST. Inspired by the successful application of the Transformer and its variants in natural language processing similar to time series prediction, we introduce it to the SST prediction in the China Sea. The model TransDtSt-Part is developed by embedding the temporal information in the classic Transformer, combining attention distillation and partial stacked connection, and performing generative decoding. High-resolution satellite-derived data from the National Oceanic and Atmospheric Administration is utilized, and long-term SST predictions with day granularity are achieved under univariate and multivariate patterns. With Root Mean Square Error and Mean Absolute Error as metrics, the TransDtSt-Part outperforms all competitive baselines in five oceans (i.e., subareas of Bohai, Yellow Sea, East China Sea, Taiwan Strait, and South China Sea) and six prediction horizons (i.e., 30, 60, 90, 180, 270, and 360 days). Experimental results demonstrate that the performance of the innovative model is encouraging and promising for the long-term prediction of SST.
With the assistance of the evaluation algorithms based on the well-performed backpropagation neural network (BPNN), we quantitatively analyze the importance of the structural parameters of the supported helical microfiber (HMF) temperature sensor. The relative output intensities of HMF sensor at different temperatures are predicted by the BPNN with the HMF’s structural parameters as the input variables. The best-forecasted performance is obtained by the BPNN with one hidden layer of ten neurons. Compared with the actual values, the root-mean-square error (RMSE) and the correlation coefficient of the predicted values are 9.7 $\times \,\,10^{-3}$ dB and 99.84%, respectively. Based on the BPNN with precise prediction, the backward stepwise elimination and the holdback input randomization methods are used to quantitatively discuss the influence of the structural parameters on the output intensity of the HMF. The relative importance from high to low is the helical length (~38%), microfiber diameter (~27%), helical angle (~25%), and cone angle (~10%). The importance of four geometric parameters obtained by the two methods is ranked the same. Quantitative analysis of structural parameters relying on the well-predicted BPNN can give basic information on the structural characteristics of the HMF sensor, which helps to optimize the structure design of the optical sensors based on micro/nanofiber and provides a powerful guarantee for its practical application.
We distinctly reveal the difference in the exciton generation processes in phosphorescent organic light-emitting devices with an exciplex-type co-host and a single host. Excitons in the co-host consisting of 4,4,4-tris(N-carbazolyl)-triphenylamine and 1,3,5-tris(N-phenylbenzimidazol-2-yl)benzene are created via efficient energy transfer from the exciplex to the phosphorescent dopant. In contrast, excitons in the single host of 4,4,4-tris(N-carbazolyl)-triphenylamine are formed by the combination of holes and electrons trapped by the phosphorescent dopants. The optimized device utilizing the co-host system exhibits highly superior performance relative to the single-host device. The maximum external quantum efficiency and maximum luminance are 14.88% and for the co-host device, being 1.6 times and 3.6 times the maximum external efficiency and maximum luminance for the single-host device, respectively. Significantly, the critical current density, evaluating the device efficiency roll-off characteristic, is as high as , which is highly superior to for the single-host device, indicating the notable alleviation in efficiency roll-off for the co-host device. The significant improvement in device performance is attributed to eliminating the exciton quenching resulting from the captured holes and the efficient energy transfer from the exciplex-type co-host to the phosphorescent emitter incurred by the reverse intersystem crossing process.
An exciplex with significant thermally activated delayed fluorescence properties was realized, comprising diphenyl-[3′-(1-phenyl-1H-phenanthro[9,10-d]imidazol-2-yl)-biphenyl-4-yl]-amine as a donor and 2,4,6-tris[3-(diphenylphosphinyl)phenyl]-1,3,5-triazine as an acceptor. A very small energy difference between the singlet and triplet levels and a large rate constant of the reverse intersystem crossing were attained simultaneously, contributing to the efficient upconversion of triplet excitons from the triplet state to the singlet state and thermally activated delayed fluorescence emission. A high-efficiency organic light-emitting device based on the exciplex was fabricated, which exhibited a maximum current efficiency, power efficiency, external quantum efficiency, and exciton utilization efficiency of 23.1 cd/A, 24.2 lm/W, 7.32%, and 54%, respectively. The efficiency roll-off of the exciplex-based device was slight, as illustrated by a large critical current density of 34.1 mA/cm2. This efficiency roll-off was ascribed to triplet–triplet annihilation, as confirmed by the triplet–triplet annihilation model. We proved the high binding energy of the excitons and excellent charge confinement within the exciplex by performing transient electroluminescence measurements.
We demonstrate a directly visible approach to adjust the coupling conditions of photonic components based on fluorescent indication. Under an optical microscope, the fluorescence of the dye-doped polymer microfiber (PMF) is excited by the evanescent coupling method and is used to visualize the light path in the PMF. By simply moving the coupling position of input probe in a few microns according to the fluorescent indication, we adjust the coupling conditions of the components, such as $Y$ beam splitter and Mach–Zehnder interferometer, and the splitting ratio varies from 1.8 to 0.8. Besides, we directly visualize and adjust the energy coupling of two overlapping PMFs in the pig-tail microring via the fluorescent indication. The fluorescent indication in PMF demonstrated here simplifies the adjusting process of the component based on free standing PMFs and promises a potential for directly visible adjusting in microfibers-based photonic circuit integration for photonic neural networks.