| 1. |
- Plavetic, Marko, et al.
(author)
-
Identification of environmentally relevant benthic foraminifera from the Skagerrak fjords by deep learning image modelling
- 2023
-
In: International Congress FORAMS2023, Perugia, Italy 25-30th June, 2023.
-
Conference paper (other academic/artistic)abstract
- Over the several past decades there has been increasing interest in using foraminifera as environmental indicators for coastal marine environments. As compared to macrofauna, which are currently used in environmental studies, foraminifera offer several distinct advantages as bioindicators, including short generation times, high number of individuals per small sample volume, hard and durable tests with high preservation potential, and low cost of sample extraction. One of the main problems with foraminifera identification is reliance on manual identification and expert judgement, which is a tedious and slow process prone to errors and subjectivity. Deep learning, a subfield of machine learning, has emerged as a promising solution to this challenge, since a neural network can learn to recognize subtle differences in shell morphology that may be difficult for the human eye to distinguish. Benthic foraminifera mounted on microslides from several Skagerrak fjords including Gullmar Fjord, Hakefjord and Idefjord were imaged using a Nikon SMZ-10 stereomicroscope and DeltaPix DP450 microscope camera. Images were then processed in Roboflow API, where individual foraminifera were labelled and classified. This resulted in 3003 images and 22 138 labelled individuals. Using the labeled images, a dataset was created to be used for deep learning training. We used the YOLO (You Only Look Once) v7 model implemented in the PyTorch framework, which has demonstrated state-of-the-art speed and performance for object detection as of the time of writing. Models were trained using a Nvidia RTX A4000 GPU (graphical processing unit). Preliminary results show a 90,3% mAP (mean average precision) and 78,8% mAP on the best and the worst performing models, respectively. Even though the imaging and labelling was done in a short amount of time, the results look promising and show that even a relatively small dataset can be used for training a reliable deep learning species identification model.
|
|
| 2. |
- Shahabi-Ghahfarokhi, Sina, et al.
(author)
-
The response of metal mobilization and redistribution to reoxygenation in Baltic Sea anoxic sediments
- 2022
-
In: Science of the Total Environment. - : Elsevier. - 0048-9697 .- 1879-1026. ; 837
-
Journal article (peer-reviewed)abstract
- To bring life back to anoxic coastal and sea basins, reoxygenation of anoxic/hypoxic zones has been proposed. This research focuses on the metals released during the oxidization of sediments from two locations in the anoxic Eastern Gotland Basin under a laboratory-scale study. Triplicate experimental cores and reference cores were collected from the North and South Eastern Gotland Basins. The oxygenation of the water column took place over a 96-hour experiment in a dark and 5 °C environment. In 12 and 24 hour intervals, the surface waters were exchanged and, over time, analyzed for pH, electroconductivity (EC), total organic carbon (TOC), soluble metal concentrations, and the top samples (0–10 cm) were analyzed with 3-step (E1: water-soluble, E2: exchangeable, and E3: organic-bound) sequential chemical extraction (SCE). Results show stable pH and decreasing EC in the column waters. The EC indicates that metals are released in the initial phases (12 h) of reoxygenation for both sites. Arsenic, Ba, Co, Mn, Rb, U, K, Sr, and Mo are released into the water column during the 96 hour experiment, and based on the calculations for the entire East Gotland Basin, would mean 8, 50, 0.55, 734, 53, 27, 347,178, 3468, and 156 μg L−1 are released, respectively. Elements Mn, Mo, U, and As are released in higher concentrations during the experiment than previously measured in the Eastern Gotland Basin, which provides vital information for future proposed remediation and natural geochemical processes with their known environmental impacts. The SCE results show that redox-sensitive metals (Mn, U, and Mo) are released in the highest concentrations into the solution. The relationship between the highest released metals (beside redox-sensitive) into solution over the oxygenation and their initial abundant phase is noticed, where the smallest released concentrations belong to K < Rb < Sr in E2, and As
|
|
