| 1. |
- Di Baldassarre, Giuliano, et al.
(författare)
-
An integrative research framework to unravel the interplay of natural hazards and vulnerabilities
- 2018
-
Ingår i: Earth's Future. - : John Wiley & Sons. - 2328-4277. ; 6:3, s. 305-310
-
Tidskriftsartikel (refereegranskat)abstract
- Climate change, globalization, urbanization, social isolation, and increased interconnectednessbetween physical, human, and technological systems pose major challenges to disaster risk reduction(DRR). Subsequently, economic losses caused by natural hazards are increasing in many regions of theworld, despite scientific progress, persistent policy action, and international cooperation. We argue thatthese dramatic figures call for novel scientific approaches and new types of data collection to integratethe two main approaches that still dominate the science underpinning DRR: the hazard paradigm and thevulnerability paradigm. Building from these two approaches, here we propose a research framework thatspecifies the scope of enquiry, concepts, and general relations among phenomena. We then discuss theessential steps to advance systematic empirical research and evidence-based DRR policy action.
|
|
| 2. |
- Rhodes, Emma, 1990-, et al.
(författare)
-
Textural Insights Into the Evolving Lava Dome Cycles at Santiaguito Lava Dome, Guatemala
- 2018
-
Ingår i: Frontiers in Earth Science. - : Frontiers Media SA. - 2296-6463. ; 6
-
Tidskriftsartikel (refereegranskat)abstract
- The structures and textures preserved in lava domes reflect underlying magmatic and eruptive processes, and may provide evidence of how eruptions initiate and evolve. This study explores the remarkable cycles in lava extrusion style produced between 1922 and 2012 at the Santiaguito lava dome complex, Guatemala. By combining an examination of eruptive lava morphologies and textures with a review of historical records, we aim to constrain the processes responsible for the range of erupted lava type and morphologies. The Santiaguito lava dome complex is divided into four domes (El Caliente, La Mitad, El Monje, El Brujo), containing a range of proximal structures (e.g., spines) from which a series of structurally contrasting lava flows originate. Vesicular lava flows (with a'a like, yet non-brecciated flow top) have the highest porosity with interconnected spheroidal pores and may transition into blocky lava flows. Blocky lava flows are high volume and texturally variable with dense zones of small tubular aligned pore networks and more porous zones of spheroidal shaped pores. Spines are dense and low volume and contain small skeletal shaped pores, and subvertical zones of sigmoidal pores. We attribute the observed differences in pore shapes to reflect shallow inflation, deflation, flattening, or shearing of the pore fraction. Effusion rate and duration of the eruption define the amount of time available for heating or cooling, degassing and outgassing prior to and during extrusion, driving changes in pore textures and lava type. Our new textural data when reviewed with all the other published data allow a cyclic model to be developed. The cyclic eruption models are influenced by viscosity changes resulting from (1) initial magmatic composition and temperature, and (2) effusion rate which in turn affects degassing, outgassing and cooling time in the conduit. Each lava type presents a unique set of hazards and understanding the morphologies and dome progression is useful in hazard forecasting.
|
|
| 3. |
- Sullivan, Devin P., et al.
(författare)
-
Deep learning is combined with massive-scale citizen science to improve large-scale image classification
- 2018
-
Ingår i: Nature Biotechnology. - : NATURE PUBLISHING GROUP. - 1087-0156 .- 1546-1696. ; 36:9, s. 820-
-
Tidskriftsartikel (refereegranskat)abstract
- Pattern recognition and classification of images are key challenges throughout the life sciences. We combined two approaches for large-scale classification of fluorescence microscopy images. First, using the publicly available data set from the Cell Atlas of the Human Protein Atlas (HPA), we integrated an image-classification task into a mainstream video game (EVE Online) as a mini-game, named Project Discovery. Participation by 322,006 gamers over 1 year provided nearly 33 million classifications of subcellular localization patterns, including patterns that were not previously annotated by the HPA. Second, we used deep learning to build an automated Localization Cellular Annotation Tool (Loc-CAT). This tool classifies proteins into 29 subcellular localization patterns and can deal efficiently with multi-localization proteins, performing robustly across different cell types. Combining the annotations of gamers and deep learning, we applied transfer learning to create a boosted learner that can characterize subcellular protein distribution with F1 score of 0.72. We found that engaging players of commercial computer games provided data that augmented deep learning and enabled scalable and readily improved image classification.
|
|
