No abstract

Region-based classification of PolSAR data can be effectively performed by seeking for the assignment that minimizes a distance between prototypes and segments. Silva et al. (2013) used stochastic distances between complex multivariate Wishart models which, differently from other measures, are computationally tractable. In this work we assess the robustness of such approach with respect to errors in the training stage, and propose an extension that alleviates such problems. We introduce robustness in the process by incorporating a combination of radial basis kernel functions and stochastic distances with Support Vector Machines (SVM). We consider several stochastic distances between Wishart: Bhatacharyya, Kullback-Leibler, Chi-Square, Rényi, and Hellinger. We perform two case studies with PolSAR images, both simulated and from actual sensors, and different classification scenarios to compare the performance of Minimum Distance and SVM classification frameworks. With this, we model the situation of imperfect training samples. We show that SVM with the proposed kernel functions achieves better performance with respect to Minimum Distance, at the expense of more computational resources and the need of parameter tuning. Code and data are provided for reproducibility.

Blends possessing the elastomeric properties of natural rubber (NR) and the conducting properties of conducting polymer (polyaniline, PANI) were obtained, which are promising for further application in deformation sensors. Blends containing 20% (v/v) of PANI in 80% of NR latex were fabricated by casting in the form of free-standing films and treated either with HCl or with corona discharge, which lead PANI to its conducting state (doping process). Characterization was carried out by Raman spectroscopy, d.c. conductivity and thermogravimetric analysis. Evidence for chemical interaction between PANI and NR was observed, which allowed the conclusion that the NR latex itself is able partially to induce both the primary doping of PANI (by protonation) and the secondary doping of PANI (by changing the chain conformation). Further improvement in the primary doping could be obtained for the blends either by corona discharge or by exposing them to HCl. The electrical conductivity reached in the blends was dependent on the doping conditions used, as observed by Raman scattering.

No abstract

No abstract

Abstract. On 15 February 2022, the city of Petrópolis in the highlands of the state of Rio de Janeiro, Brazil, received an unusually high volume of rain within 3 h (258 mm), generated by a strongly invigorated mesoscale convective system. It resulted in flash floods and subsequent landslides that caused the deadliest landslide disaster recorded in Petrópolis, with 231 fatalities. In this paper, we analyzed the root causes and the key triggering factors of this landslide disaster by assessing the spatial relationship of landslide occurrence with various environmental factors. Rainfall data were retrieved from 1977 to 2022 (a combination of ground weather stations and the Climate Hazards Group InfraRed Precipitation – CHIRPS). Remotely sensed data were used to map the landslide scars, soil moisture, terrain attributes, line-of-sight displacement (land surface deformation), and urban sprawling (1985–2020). The results showed that the average monthly rainfall for February 2022 was 200 mm, the heaviest recorded in Petrópolis since 1932. Heavy rainfall was also recorded mostly in regions where the landslide occurred, according to analyses of the rainfall spatial distribution. As for terrain, 23 % of slopes between 45–60∘ had landslide occurrences and east-facing slopes appeared to be the most conducive for landslides as they recorded landslide occurrences of about 9 % to 11 %. Regarding the soil moisture, higher variability was found in the lower altitude (842 m) where the residential area is concentrated. Based on our land deformation assessment, the area is geologically stable, and the landslide occurred only in the thin layer at the surface. Out of the 1700 buildings found in the region of interest, 1021 are on the slope between 20 to 45∘ and about 60 houses were directly affected by the landslides. As such, we conclude that the heavy rainfall was not the only cause responsible for the catastrophic event of 15 February 2022; a combination of unplanned urban growth on slopes between 45–60∘, removal of vegetation, and the absence of inspection were also expressive driving forces of this disaster.

No abstract

No abstract