Unexpected Shallow Earthquake of August 1st, 2020 in the North of Indramayu, West Java, Indonesia

During the last one hundred years, there are no shallow seismicity in the north of Java. This area is dominated by intermediate and deep focus earthquakes due to the subducted Indo-Australian slab. An earthquake with magnitude ML 4.5 struck Indramayu, north of West Java on August 1, 2020. According to the Agency for Meteorology, Climatology, and Geophysics (BMKG), the earthquake was felt III MMI scale in Indramayu and its vicinity. We used waveform data from BMKG seismic station in West Java, then we picked P-and S-waves arrival times from each station and hypocenter location was determined by Geiger method. We have detected Pn before Pg phase on four BMKG seismic stations, indicating a shallow crustal earthquake. Our inversion show that the earthquake occurred in 6.1805° S, 108.2612° E with 5 km focus depth at 16:24:38 GMT+7. Our focal mechanism solution was determined by using moment tensor inversion shows a strike-slip faulting, which corresponds to the active fault in the north of Indramayu.

Whole Slide Preview Image Segmentation and Setup for Digital Pathology Scanners

The problem of tissue finding is of special interest in automating WSI scanners where it decomposes the preview image of tissue glass slides into a simplified and abstract level of localization and identification to setup WSI scanner for high-resolution scan. Prior to such scanning, a preview image is captured to calibrate the scanner’s parameters. Scan parameters such as focus depth and scan region are determined using a tissue finding software package. This paper introduces a series of pipelines (e.g. binary mask segmentation, tissue/artifact classification, region-of-interest allocation) to automate tissue preview segmentation in both brightfield and darkfield microscopy.

Accuracy of a neural net classification of closely-related species of microfossils from a sparse dataset of unedited images

Identification of biologic objects in images is a major source of biodiversity data. Currently this is done by scarce taxonomic experts and data is thus limited in scope and reproducibility. Automated identification in fields such as plankton research or micropaleontology, where enormous numbers of objects are available, would significantly improve data quantity and quality, particularly in applied studies of environmental and climate change. We describe a machine learning workflow based on the MobileNet convolutional network. The software can identify closely related species of radiolarians, a morphologically challenging group of microfossils, and from complete species populations (not only ideal specimens) as they are normally identified in standard transmitted light microscope preparations. Multiple, partial focus, depth of field limited images were obtained for each fossil specimen from multiple radiolarian microslides. Images were normalized and in one test also cropped to remove most systematic slide-linked image biases (e. g. type of background particles) that could be used by a classifier as non-taxonomic clues to species assignment. An average of 60 specimens per species for 16 species in two distinct clusters of closely related forms (9 species in the Antarctissa group and 7 species in the genus Cycladophora) were used to train and test the system. An overall average classification accuracy of ca 73% was achieved, and for some species >85%. Using a cutoff for specimens with classifier-calculated low certainty values boosts overall accuracy close to 90%, but at the cost of ca 1/3 reduction in identifiable specimens. This latter accuracy is close to the reproducibility of human experts, albeit with more unidentifiable specimens. The most important constraint to broader use is the time and effort needed by taxonomic experts to collect and label images to be used in training, as many species in these diverse biotas are rare, and the numbers of taxonomic experts available are very limited.

Accuracy of a neural net classification of closely-related species of microfossils from a sparse dataset of unedited images

Identification of biologic objects in images is a major source of biodiversity data. Currently this is done by scarce taxonomic experts and data is thus limited in scope and reproducibility. Automated identification in fields such as plankton research or micropaleontology, where enormous numbers of objects are available, would significantly improve data quantity and quality, particularly in applied studies of environmental and climate change. We describe a machine learning workflow based on the MobileNet convolutional network. The software can identify closely related species of radiolarians, a morphologically challenging group of microfossils, and from complete species populations (not only ideal specimens) as they are normally identified in standard transmitted light microscope preparations. Multiple, partial focus, depth of field limited images were obtained for each fossil specimen from multiple radiolarian microslides. Images were normalized and in one test also cropped to remove most systematic slide-linked image biases (e. g. type of background particles) that could be used by a classifier as non-taxonomic clues to species assignment. An average of 60 specimens per species for 16 species in two distinct clusters of closely related forms (9 species in the Antarctissa group and 7 species in the genus Cycladophora) were used to train and test the system. An overall average classification accuracy of ca 73% was achieved, and for some species >85%. Using a cutoff for specimens with classifier-calculated low certainty values boosts overall accuracy close to 90%, but at the cost of ca 1/3 reduction in identifiable specimens. This latter accuracy is close to the reproducibility of human experts, albeit with more unidentifiable specimens. The most important constraint to broader use is the time and effort needed by taxonomic experts to collect and label images to be used in training, as many species in these diverse biotas are rare, and the numbers of taxonomic experts available are very limited.

PENGEMBANGAN PAKET PERKULIAHAN READING BERKELANJUTAN JURUSAN PENDIDIKAN BAHASA INGGRIS UNDIKSHA

This research aims at developing the syllabus and instructional materials for Reading course series in English Education Department, Ganesha University of Education. This Reading course comprises four series: Reading I (Literal Reading), Reading II (Intepretive Reading), Reading III (Critical Reading) and, Advanced Reading (Reading from the Media). The research was inspired by preliminary findings about unclear transition and focus of each serial of reading course. These include the degree of complexity and coverage of the texts, instructional strategies and teaching strategies. As a matter of fact, reading is a stepping stone for developing other language skills such as writing, speaking and listening (Guthrie& Kirsch, 2007). The first year of the research has resulted in the adapted syllabus and blueprint for the material development. In the second year the instructional materials for levelled reading courses were developed and validated and was checked its relevance and readability. Data analysis found that the product was highly relevant to the syllabus and provide with Engliah language teaching. This indicates that the quality of the product can be categorized as very good and as systematic gradation, focus, depth and scope.

Relocation and Focal Mechanism of Aftershocks Pidie Jaya Earthquake (Mw6.5) Dec 7th, 2016 using BMKG Network

We have successfully relocated 74 out of 89 aftershocks until December 19, 2016, by using hypocenter double-difference method. We also have conducted focal mechanism analysis to estimate the type of fault slip. The results indicate improvement in hypocenter location, where the initial earthquakes focus depth at a fixed depth of 10 km have been updated and have described the patterns of active fault in the area trending Northwest-Southeast. The validity through the histogram of travel-time residual shows fairly good data processing where the residual value is close to zero (t.obs - t.cal ~ 0). Based on focal mechanism solutions of mainshock and two selected aftershocks, the type of fault is right lateral strike-slip.