Surface sediment classification using a deep learning model and unmanned aerial vehicle data of tidal flats.

This study proposes a deep learning model, U-Net, to improve surface sediment classification using high-resolution unmanned aerial vehicle (UAV) images. We constructed training datasets with UAV images and corresponding labeling data acquired from three field surveys on the Hwangdo tidal flat. The labeling data indicated the distribution of surface sediment types. We compared the performance of the U-Net model trained in various implementation environments, such as surface sediment criteria, input datasets, and classification models. The U-Net trained with five class criteria-derived from previous classification criteria-yielded valid results (overall accuracy:65.6 %). The most accurate results were acquired from trained U-Net with all input datasets; in particular, the tidal channel density caused a significant increase in accuracy. The accuracy of the U-Net was approximately 20 % higher than that of other classification models. These results demonstrate that surface sediment classification using UAV images and the U-Net model is effective.

Impact of activator type on the immobilisation of lead in fly ash-based geopolymer.

Immobilisation of heavy metals in geopolymers has attracted attention as a potential means of treating toxic wastes. Lead is known to be effectively immobilised in a geopolymer matrix, but detailed explanation for the mechanisms involved and the specific chemical form of lead are not fully understood. To reveal the effect of the activator types on the immobilisation of lead in geopolymers, 0.5 and 1.0wt% lead in the form of lead nitrate was mixed with fly ash and alkaline activators. Different alkaline activators (either combined sodium hydroxide and sodium silicate or sodium aluminate) were used to achieve the target Si:Al ratios 2.0 and 5.0 in geopolymers. Zeolite was formed in aluminate-activated geopolymers having a Si:Al ratio of 2.0, but the zeolite crystallization was suppressed as lead content increased. No specific crystalline phase of lead was detected by X-ray diffraction, electron diffraction or FT-IR spectrometry. In fact, double Cs corrected TEM analysis revealed that lead was evenly distributed with no evidence of formation of a specific lead compound. A sequential extraction procedure for fractionation of lead showed that lead did not exist as an exchangeable ion in geopolymers, regardless of activator type used. Aluminate activation is shown to be superior in the immobilisation of lead because about 99% of extracted lead existed in the oxidizing and residual fractions.

Lattice parameter determination of mullite by energy-filtered needle-texture electron diffraction pattern.

The thermal transformations of pyrophyllite to mullite by heating were re-examined using mainly energy-filtering transmission electron microscopy and, for the first time, the texture electron diffraction pattern of the mullite was completely interpreted. Through a temperature range in which pyrophyllite dehydroxylate maintained a long-range order with a fluctuation of approximately 1% in d-spacings of (100) and (010) planes at 1000 degrees C, without prominent exothermic feature, pyrophyllite dehydroxylate was gradually decomposed and transformed into mullite through topotaxy. Pyrophyllite dehydroxylate did not collapse completely until 1100 degrees C, which promoted the rapid growth of mullite in random orientation at 1200 degrees C and the crystallization of amorphous silica to cristobalite at 1300 degrees C. The mullite needles, having their c-axis (texture axis) parallel to the elongation direction, lined up along the b(*)-axis of the pyrophyllite dehydroxylate in the needle-texture electron diffraction patterns. The mullite needles had monoclinic symmetry with lattice parameters of 7.27 A (a), 7.75 A (b), 2.90 A (c), 90 degrees (alpha), 90 degrees (beta) and 88.41 degrees (gamma), which, because of the structural affiliation to the parent pyrophyllite dehydroxylate, differ to the orthorhombic 3/2-mullite.

Seabed classification from acoustic profiling data using the similarity index.

We introduce the similarity index (SI) for the classification of the sea floor from acoustic profiling data. The essential part of our approach is the singular value decomposition of the data to extract a signal coherent trace-to-trace using the Karhunen-Loeve transform. SI is defined as the percentage of the energy of the coherent part contained in the bottom return signals. Important aspects of SI are that it is easily computed and that it represents the textural roughness of the sea floor as a function of grain size, hardness, and a degree of sediment sorting. In a real data example, we classified a section of the sea floor off Cheju Island south of the Korean Peninsula and compared the result with the sedimentology defined from direct sediment sampling and side scan sonar records. The comparison shows that SI can efficiently discriminate the bottom properties by delineating sediment-type boundaries and transition zones in more detail. Therefore, we propose that SI is an effective parameter for geoacoustic modeling.