3D Modeling of Composite Membranes from TEM

Overview

A team of researchers led by KAUST graduate Lena Li with KAUST faculty Maggie Johnson recently published a paper entitled: "Rhodoliths can act as daily resolution paleotemperature archives in the Red Sea".

Paper summary by M365 Copilot: This study shows that rhodoliths — free‑living, calcified red algae found widely on the seafloor—can serve as highly detailed natural “thermometers” of past ocean conditions. The researchers developed a new method combining chemical markers, micro‑CT imaging, and advanced time‑alignment techniques to reconstruct daily seawater temperatures from the complex growth patterns of a single rhodolith collected in the Red Sea, producing a continuous 133‑day record. By integrating data from multiple growth directions and chemical indicators, their approach closely matched real temperature measurements, demonstrating much greater accuracy than previous methods. Overall, the work provides proof of concept that rhodoliths can deliver high‑resolution climate records, which could help scientists better understand short‑term temperature variability and improve predictions of how marine ecosystems respond to ongoing ocean warming. 

The paper's graphical abstract shown above illustrates the approach used in this work. The paper entitled “Impact of active layer morphology on salt permeability in RO composite membranes: 3D modelling from TEM geometry and effective membrane thickness” can be accessed here. Please note that all figures used in this article are directly from the original published paper.

KVL's visualization scientist Dr. Ronell Sicat supported Dr. Martinez in pre-processing the TEM scans, testing segmentation methods, and eventually post-processing the segmentation and extracted surfaces in order to get 3D geometries that were suitable for the target simulation software. KVL also provided free access to Avizo software for data visualization, segmentation, and analysis and a powerful workstation with a pen-and-tablet interface that enabled faster manual segmentation of the TEM data. KVL's contribution is acknowledged in the paper.