Reliable sampling technology for deep-sea sediments is a prerequisite for studying the sedimentary environment in the geological history, the mechanical characteristics of sediments and sediment biology. In this study, we develop a new sampling device in order to achieve a convenient and economic pressure sampling for hadal sediments. The overall structure, as well as the key component such as the pressure-holding part, and the sealing structure are designed. The feasibility of the device was verified through field tests. Furthermore, the sampling process is analyzed by the computer simulation. It is found that the sampling length is determined by the open diameter and the installation position of the core keepers. According to the simulation, the structure is optimized, which is beneficial to the further improvement of the sampling length.

The accurate prediction of vertical deformation (VD) in tidal flat terrains is crucial for their rational development and disaster warning. However, the high-precision and strong robustness prediction method of VD data in tidal flat terrain is still a great challenge due to the complex environmental conditions. In this study, VD data was collected over 7 consecutive days using Micro-Electro-Mechanical system (MEMS) sensor arrays deployed in a tidal flat. Particle swarm optimization (PSO)-optimized variational mode decomposition (VMD) is applied to decompose the acquired VD data for noise removal. The Fast Fourier Transform (FFT) and Spearman correlation coefficient analysis were employed to identify the inducing factors by comparing them with the decomposed VD. Four machine learning algorithms were evaluated to predict the trend and periodic components of VD, and the results showed that long short-term memory neural network (LSTM) outperformed the other three, achieving a root mean square error (RMSE) less than 0.15 and an R-squared ( R ² ) greater than 0.90. The proposed combined PSO-VMD-LSTM model effectively addresses the challenges of feature recognition, noise removal and deformation prediction of vertical deformation data in tidal flat terrains.

The ecosystem of the abyss is one of the fields that humans hardly know. The ultra-high hydrostatic pressure makes it very difficult to obtain abyssal organisms. Samples are often severely broken during recovery due to changes in environmental pressure, temperature, and other factors. Currently, there are no macro-organism samplers suitable for the abyss. The development of a pressure-maintaining sampler for the abyss is a prerequisite for abyssal ecosystem research. This paper mainly proposed a pressure-maintaining trapping instrument (PMTI) designed to work at a depth above 10,000 m. Unlike typical deep-sea equipment, this instrument is lightweight (about 65 kg in water). The instrument adopts a new structure, using a hollow piston as the sampling space and sealing the mechanism with O-rings at both ends of the piston, thus avoiding sealing methods such as ball valves and greatly reducing the weight of the equipment. The structure and working process of the instrument are described in detail in this paper. Meanwhile, in this paper, the movement resistance of the piston (mainly the resistance of the O-ring) is analyzed using a dynamic explicit method in Abaqus. The factors affecting the friction of the O-rings are analyzed via the method of orthogonal tests and ANOVA. In addition, high-pressure tests were conducted on key parts of the instrument, and the results showed that the instrument works well at 100 MPa.

Terrain monitoring of intertidal beach is important to the coastline evolution research. Due to the complexity of the intertidal environment and the dynamics of terrain changes, in-situ, real-time and long-term monitoring of intertidal beach topography are big difficulties for existing observation methods. This paper describes the design and field experimental research of a terrain in-situ monitoring system based on sensor array. The monitoring system has two 20-metres long flexible sensor arrays containing 42 sensor nodes can monitor the terrain changes of an area with 10m×10m simultaneously. The optimal arrangement conditions of sensor arrays are deduced to completely reconstruct the terrain changes. A calculation model for terrain change is also proposed to convert the three-axis acceleration into vertical displacement of sensor array. To evaluate the monitoring system and validate the arrangement condition and reconstruction model, a field experiment is conducted on an intertidal beach in Zhoushan Island. The cumulative displacement and vertical velocity of silt zone and tidal height and waves are observed and obtained. The results demonstrate the feasibility and stability of the proposed system under real-time field monitoring conditions.

The accurate recording of terrain and strata deformation time series is vital for geodesy and geophysical studies. However, missing values often occur due to various reasons, necessitating robust imputation methods to utilize all available data effectively. In this study, a novel missing data imputation method is proposed based on the in-situ long-term monitoring data obtained from Micro-Electro-Mechanical System (MEMS) sensor arrays deployed in the Beilun Port of Ningbo, China. We investigate the spatiotemporal correlation between different sensor nodes and employ separate fitting of deformation data from highly correlated sensors during high-tide and low-tide times for imputation. The analysis indicates that simple regression imputation achieves relative error of approximately 2% when the correlation coefficient exceeds 0.90 for single node. However, for multiple regression, the relative error does not significantly decrease with an increasing number of independent variables beyond three. To validate the accuracy of the proposed imputation method, the simulations using four baseline methods and four regression-iterative (CI) models are conducted, which includes encompassed four different missing patterns and missing rates ranging from 5% to 40%. The results demonstrate that the correlation-iterative random forest regression (CI-RFR) model outperforms other methods across various missing patterns and rates, particularly in the in-row pattern.