Prediction of three-dimensional ocean temperature in the South China Sea based on time series gridded data and a dynamic spatiotemporal graph neural network

Ocean temperature is an important physical variable in marine ecosystems,and ocean temperature prediction is an important research objective in ocean-related fields.Currently,one of the commonly used methods for ocean temperature prediction is based on data-driven,but research on this method is mostly limited to the sea surface,with few studies on the prediction of internal ocean temperature.Existing graph neural network-based methods usually use predefined graphs or learned static graphs,which cannot capture the dynamic associations among data.In this study,we propose a novel dynamic spatiotemporal graph neural network(DSTGN)to predict threedimensional ocean temperature(3D-OT),which combines static graph learning and dynamic graph learning to automatically mine two unknown dependencies between sequences based on the original 3D-OT data without prior knowledge.Temporal and spatial dependencies in the time series were then captured using temporal and graph convolutions.We also integrated dynamic graph learning,static graph learning,graph convolution,and temporal convolution into an end-to-end framework for 3D-OT prediction using time-series grid data.In this study,we conducted prediction experiments using high-resolution 3D-OT from the Copernicus global ocean physical reanalysis,with data covering the vertical variation of temperature from the sea surface to 1000 m below the sea surface.We compared five mainstream models that are commonly used for ocean temperature prediction,and the results showed that the method achieved the best prediction results at all prediction scales.

Comparisons of Criteria for Analyzing the Dynamical Association of Solutes in Aqueous Solutions

How to determine accurately the association states of solutes in aqueous systems is of fundamental importance in a variety of chemical, physical, and biological processes, We apply four widely used criteria to analyze the dynamic association processes of solutes, e.g., amphiphilic molecules, and to find the inappropriate selections of representative sites on solutes in these criteria may bring about appreciable influence on the estimation of dynamic association behaviors such as unrealistic packing radii and even misleading packing structures. It would be better to select dynamically representative sites on solute molecules based on the characteristic of solute associations. Our detailed discussions give a guide on how to determine an appropriate criterion to accurately analyze the association behaviors of solute molecules in aqueous solutions.

Comments on:Perioperative von Willebrand factor dynamics are associated with liver regeneration and predict outcome after liver resection

Recentlythearticle"PerioperativevonWillebrandfactordynamics are associated with liver regeneration and predict outcome afterliver resection" was published in Hepatology[1].Prof.Starlinger et al. aimed to assess the association of von Willebrand factor (vWF) levels and clinical outcome in patients with liver cancers post-liverresection(LR).Basedonthemechanismthatplatelets accumulation in the liver may promote liver regeneration after partial LR in mice, they found the vWF-dependent pattern of platelets accumulationduringliverregenerationinpatientsaftersurgery.

Kinematics of the Open Cluster System in the Galaxy

Absolute proper motions and radial velocities of 202 open clusters in the solar neighborhood, which can be used as tracers of the Galactic disk, are used to investigate the kinematics of the Galaxy in the solar vicinity, including the mean heliocentric velocity components （u1, u2, u3） of the open cluster system, the characteristic velocity dispersions （σ1,σ2,σ3）, Oort constants （A, B） and the large-scale radial motion parameters （C, D） of the Galaxy. The results derived from the observational data of proper motions and radial velocities of a subgroup of 117 thin disk young open clusters by means of a maximum likelihood algorithm are： （u1,u2,u3） = （-16.1 ± 1.0,-7.9 ±1.4,-10.4±1.5） km·s^-1, （σ1,σ2,σ3） = （17.0±0.7, 12.2±0.9, 8.0±1.3） km·S^-1, （A, B） = （14.8±1.0, - 13.0±2.7） km·s^-1 kpc^-1, and （C, D） = （1.5 ± 0.7, -1.2 ±1.5） km·s^-1 kpc^-1. A discussion on the results and comparisons with what was obtained by other authors is given.

Galactic kinematics and structure defined by open clusters

On the basis of recently published astrophysical parameters of the open clusters, we have selected 301 clusters with measurements of their kinematical parameters to trace the local structure and kinematics of the Galactic disk. The present sample covers a range of over 3.0 kpc from the Sun and gives significant estimates of the disk structure and kinematical parameters of the Galaxy. We derive the disk scale height, vertical displacement of the Sun to the Galactic plane, solar motion with respect to the local standard of rest, circular speed of the Galactic rotation, Galactocentric distance from the Sun, etc. We found that the average scale height of the disk defined by the open clusters is Zh = 58 ± 4pc, with a vertical displacement of the Sun below the Galactic plane of z0 -= - 16±4 pc. Clusters with ages older than 50 Myr are less concentrated in the average plane （Zh=67 ±6pc） than the younger clusters （Zh = 51±5pc）. Using the approximation of axisymmetric circular rotation, we have derived the distance to the Galactic center from the Sun R0 = 8.03 ±0.70 kpc, which is in excellent agreement with the best estimate of the Galactocentric distance. From a kinematical analysis, we found an agedependent rotation of the Galaxy. The older clusters exhibit a lower velocity of vorticity, but have the same shear as the younger clusters. The mean rotation velocity of the Galaxy was obtained as 235 ± 10 km s-1.

The dynamics of ant mosaics in tropical rainforests characterized using the Self-Organizing Map algorithm

Ants, the most abundant taxa among canopy-dwelling animals in tropical rainforests, are mostly represented by territorially dominant arboreal ants （TDAs） whose territories are distributed in a mosaic pattern （arboreal ant mosaics）. Large TDA colonies regulate insect herbivores, with implications for forestry and agronomy. What generates these mosaics in vegetal formations, which are dynamic, still needs to be better understood. So, from empirical research based on 3 Cameroonian tree species （Lophira alata, Ochnaceae; Anthocleista vogelii, Gentianaceae; and Barteriafistulosa, Passifloraceae）, we used the Self-Organizing Map （SOM, neural network） to illustrate the succession of TDAs as their host trees grow and age. The SOM separated the trees by species and by size for L. alata, which can reach 60 m in height and live several centuries. An ontogenic succession of TDAs from sapling to mature trees is shown, and some ecological traits are highlighted for certain TDAs. Also, because the SOM permits the analysis of data with many zeroes with no effect of outliers on the overall scatterplot distributions, we obtained ecological information on rare species. Finally, the SOM permitted us to show that functional groups cannot be selected at the genus level as congeneric species can have very different ecological niches, something particularly true for Crematogaster spp., which include a species specifically associated with B. fistulosa, nondominant species and TDAs. Therefore, the SOM permitted the complex relationships between TDAs and their growing host trees to be analyzed, while also providing new information on the ecological traits of the ant species involved.