Éditeur(s) : HAL CCSD BioMed Central
Résumé : International audience
Background: HCV treatment uptake has drastically increased in HIV-HCV coinfected patients in France since direct-acting antiviral (DAA) treatment approval, resulting in HCV cure in 63% of all HIV-HCV patients by the end of 2015. We investigated the impact of scaling-up DAA on HCV prevalence in the whole HIV population and in various risk groups over the next 10 years in France using a transmission dynamic compartmental model.Methods: The model was based on epidemiological data from the French Dat’AIDS cohort. Eight risk groups wereconsidered, including high-risk (HR) and low-risk (LR) men who have sex with men (MSM) and male/female heterosexuals,intra-venous drug users, or patients from other risk groups. The model was calibrated on prevalence and incidence dataobserved in the cohort between 2012 and 2015.Results: On January 1, 2016, 156,811 patients were registered as infected with HIV in France (24,900 undiagnosedpatients) of whom 7938 (5.1%) had detectable HCV-RNA (722 undiagnosed patients). Assuming a treatmentcoverage (TC) rate of 30%/year (i.e., the observed rate in 2015), model projections showed that HCV prevalenceamong HIV patients is expected to drop to 0.81% in 2026. Sub-analyses showed a similar decrease of HIV-HCVprevalence in most risk groups, including LR MSM. Due to higher infection and reinfection rates, predicted prevalencein HR MSM remained stable from 6.96% in 2016 to 6.34% in 2026. Increasing annual TC rate in HR MSM to 50/70%would decrease HCV prevalence in this group to 2.35/1.25% in 2026. With a 30% TC rate, undiagnosed patients wouldaccount for 34% of HCV infections in 2026.Conclusions: Our model suggests that DAA could nearly eliminate coinfection in France within 10 years for most riskgroups, including LR MSM. Elimination in HR MSM will require increased TC.
ISSN: 1741-7015

Droits : info:eu-repo/semantics/OpenAccess
inserm-01685739
http://www.hal.inserm.fr/inserm-01685739
http://www.hal.inserm.fr/inserm-01685739/document
http://www.hal.inserm.fr/inserm-01685739/file/2017%2C%20Virlogeux%20-%20Modeling%20HIV-HCV%20coinfection.pdf
DOI : 10.1186/s12916-017-0979-1

Éditeur(s) : HAL CCSD Elsevier
Résumé : International audience
In this paper, we propose a benchmarking of supervised machine learning techniques (neural networks, Gaussian processes and support vector machines) in order to forecast the Global Horizontal solar Irradiance (GHI). We also include in this benchmark a simple linear autoregressive (AR) model as well as two naive models based on persistence of the GHI and persistence of the clear sky index (denoted herein scaled persistence model). The models are calibrated and validated with data from three French islands: Corsica (41.91°N; 8.73°E), Guadeloupe (16.26°N; 61.51°W) and Reunion (21.34°S ; 55.49°E). The main findings of this work are, that for hour ahead solar forecasting, the machine learning techniques slightly improve the performances exhibited by the linear AR and the scaled persistence model. However, the improvement appears to be more pronounced in case of unstable sky conditions. These nonlinear techniques start to outperform their simple counterparts for forecasting horizons greater than one hour.
ISSN: 0038-092X

Droits : info:eu-repo/semantics/OpenAccess
hal-01101564
https://hal.archives-ouvertes.fr/hal-01101564
https://hal.archives-ouvertes.fr/hal-01101564/document
https://hal.archives-ouvertes.fr/hal-01101564/file/Manuscript_Lauretetal_Rev_unmarked.pdf
DOI : 10.1016/j.solener.2014.12.014

Éditeur(s) : HAL CCSD Springer Verlag
Résumé : International audience
PurposeMarine mangrove sediments in the Manche-à-Eau lagoon (Guadeloupe, Caribbean Sea) harbor locally extensive, white microbial mats. These mats cover the surface of reduced sediments near the roots of red mangrove trees, Rhizophora mangle, and are mainly composed of sulfur-oxidizing bacteria belonging to the Beggiatoaceae family, with some filamentous cyanobacteria. The goal of this study was to investigate the possible influence of sediment characteristics on the presence of these microbial mats.Materials and methodsFour push cores were collected in April 2013, two from zones with microbial mats and two from zones without mats. Sediment characteristics (grain-size distribution, mineralogy, total organic carbon (TOC) and total nitrogen (TN) contents, atomic TOC/TN ratios, and organic matter (OM) δ13C values) were compared for all four cores.Results and discussionSignificant differences were observed between sediments below microbial mats and those without mats. Sediments with microbial mats contained greater amounts of clay, and higher TOC, TN, and TOC/TN ratios, with lower total carbonate content and δ13C values. The higher clay content most likely results from lower fluid flow velocity near to mangrove roots, while higher TOC/TN ratios and lower δ13C values indicate higher inputs of OM from mangrove trees. These results are consistent with the fact that microbial mats were observed near the roots of mangrove trees, which trap OM from terrestrial vegetation and fine sediments.ConclusionsThe grain-size distribution of sediment particles, the total carbonate content, and the δ13C values are the main parameters discriminating between zones with microbial mats and those without mats. Variations in total carbonate content, which is mainly of biogenic origin, result from conditions that are more favorable for benthic organisms in zones without microbial mats. Variations of the TOC/TN ratios are controlled by the presence of a non-negligible amount of inorganic nitrogen bound to surface clay mineral particles and/or by microbial processes.
ISSN: 1439-0108

hal-01545903
https://hal.archives-ouvertes.fr/hal-01545903
DOI : 10.1007/s11368-016-1555-6

Éditeur(s) : HAL CCSD American Association for the Advancement of Science
Résumé : International audience
Most eukaryotic organisms are arthropods. Yet, their diversity in rich terrestrial ecosystems is still unknown. Here we produce tangible estimates of the total species richness of arthropods in a tropical rainforest. Using a comprehensive range of structured protocols, we sampled the phylogenetic breadth of arthropod taxa from the soil to the forest canopy in the San Lorenzo forest, Panama. We collected 6144 arthropod species from 0.48 hectare and extrapolated total species richness to larger areas on the basis of competing models. The whole 6000-hectare forest reserve most likely sustains 25,000 arthropod species. Notably, just 1 hectare of rainforest yields >60% of the arthropod biodiversity held in the wider landscape. Models based on plant diversity fitted the accumulated species richness of both herbivore and nonherbivore taxa exceptionally well. This lends credence to global estimates of arthropod biodiversity developed from plant models.
ISSN: 0036-8075

hal-00817285
https://hal.archives-ouvertes.fr/hal-00817285
https://hal.archives-ouvertes.fr/hal-00817285/document
https://hal.archives-ouvertes.fr/hal-00817285/file/1226727Revisedtext.pdf
DOI : 10.1126/science.1226727

Éditeur(s) : HAL CCSD
Résumé : International audience
Atomic Energy Review

hal-01319666
http://hal.univ-reunion.fr/hal-01319666
DOI : 10.1016/j.energy.2015.07.089

Éditeur(s) : HAL CCSD Springer Verlag
Résumé : International audience
Purpose: Marine mangrove sediments in the Manche-à-Eau lagoon (Guadeloupe, Caribbean Sea) harbor locally extensive, white microbial mats. These mats cover the surface of reduced sediments near the roots of red mangrove trees, Rhizophora mangle, and are mainly composed of sulfur-oxidizing bacteria belonging to the Beggiatoaceae family, with some filamentous cyanobacteria. The goal of this study was to investigatethe possible influence of sediment characteristics on the presence of these microbial mats. Materials and methods: Four push cores were collected in April 2013, two from zones with microbial mats and two from zones without mats. Sediment characteristics (grain-size distribution, mineralogy, total organic carbon (TOC) and total nitrogen (TN) contents, atomic TOC/TN ratios, and organic matter (OM) δ13C values) were compared for all four cores.Results and discussion: Significant differences were observed between sediments below microbial mats and those without mats. Sediments with microbial mats contained greater amounts of clay, and higher TOC, TN, and TOC/TN ratios, with lower total carbonate content andδ13C values. The higher clay content most likely results from lower fluid flow velocity near to mangrove roots, while higher TOC/TN ratiosand lower δ13C values indicate higher inputs of OM from mangrove trees. These results are consistent with the fact that microbial mats were observed near the roots of mangrove trees, which trap OM from terrestrial vegetation and fine sediments.Conclusions: The grain-size distribution of sediment particles, the total carbonate content, and the δ13C values are the main parameters discriminating between zones with microbial mats and those without mats. Variations in total carbonate content, which is mainly of biogenic origin, result from conditions that are more favorable for benthic organisms in zones without microbial mats. Variations of the TOC/TN ratios are controlled by the presence of a non-negligible amount of inorganic nitrogen bound to surface clay mineral particles and/or by microbial processes.
ISSN: 1439-0108

hal-01682110
https://hal.univ-antilles.fr/hal-01682110
DOI : 10.1007/s11368-016-1555-6

Éditeur(s) : HAL CCSD International Association for Energy Economics
Résumé : International audience
In this paper we propose to determinate and to test a set ofstatistical parameters (20)to estimate the predictability of the global horizontal irradiation time series and thereby propose a new prospective tool indicating the expected error regardlessthe forecasting methodsa modeller can possibly implement. The mean absolute log return, which is a tool usually used in econometry, proves to be a very good estimator. Some examples of the use of this tool are exposed, showing the interest of this statistical parameter in concrete cases of predictions or optimizations.
ISSN: 0195-6574

hal-01180041
https://hal.archives-ouvertes.fr/hal-01180041
https://hal.archives-ouvertes.fr/hal-01180041/document
https://hal.archives-ouvertes.fr/hal-01180041/file/discriminent_param.pdf

Éditeur(s) : HAL CCSD Elsevier
Résumé : International audience
In this paper, we propose a benchmarking of supervised machine learning techniques (neural networks, Gaussian processes and support vector machines) in order to forecast the Global Horizontal solar Irradiance (GHI). We also include in this benchmark a simple linear autoregressive (AR) model as well as two naive models based on persistence of the GHI and persistence of the clear sky index (denoted herein scaled persistence model). The models are calibrated and validatedwith data from three French islands: Corsica (41.91°N; 8.73°E), Guadeloupe (16.26°N; 61.51°W) and Reunion (21.34°S ; 55.49°E). The main findings of this work are, that for hour ahead solar forecasting,the machine learning techniques slightly improve the performances exhibited by the linear AR and the scaled persistence model. However, the improvement appears to be more pronounced in case of unstable sky conditions.These nonlinear techniques start to outperform their simple counterparts for forecasting horizons greater than one hour.
ISSN: 0038-092X

Droits : info:eu-repo/semantics/OpenAccess
hal-01099211
https://hal.archives-ouvertes.fr/hal-01099211
https://hal.archives-ouvertes.fr/hal-01099211/document
https://hal.archives-ouvertes.fr/hal-01099211/file/Manuscript_Lauretetal_Rev_unmarked.pdf