Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

Tropical forests in Central Africa are found under dry and seasonal climates. The vegetation adopts a seasonal pattern as depicted by localized field studies, but its importance and extent are barely known. The recent Sentinel-2 mission provides new opportunities to monitor vegetation phenology from space. 
The first objective of this study was to assess the potential of Sentinel-2 data to monitor vegetation seasonality in Central Africa. The use of Sentinel-2 data further aimed to address the questions of how seasonal are Central African forests, and how is this seasonal functioning related to rainfall seasonality in Central African Republic (Mbaïki) and Democratic Republic of Congo (Luki) subjected to reverse rainfall regimes.
This work relied on three data types: (i) Sentinel-2 images, (ii) ground data consisting of regular observations of phenophases and (iii) rainfall data from the Global Precipitation Measurement mission.
The resulting rainfall and Enhanced Vegetation Index times series allowed respectively the retrieval of the start of rain and the start of the season further compared to the ground observations. An additional wavelet analysis was performed on the Mbaïki site to determine the frequency and timing of the periodic vegetation events. 
This work has demonstrated the suitability of the recently available Sentinel-2 data for monitoring vegetation dynamics when cloud contamination remains reasonable. In addition, annual vegetation cycles dominated at the study sites, in line with the seasonality of rainfall. Comparison between sites confirmed a shift in vegetation seasonality from Mbaïki to Luki in response to the inversion of rainfall patterns across the Equator. In Mbaïki, the forest EVI signal was on average 16.9 ± 4.4 days before the rainfall signal, supporting the ultimate control of rainfall. In contrast, for both study sites, the earliest onset of rainfall resulted in the earliest onset of the season in 2019, supporting the hypothesis that rainfall exert a proximal control.
To conclude, Sentinel-2 data are suited to monitor vegetation seasonality but would require the combination with additional images to cope with continuously clouded areas. Les forêts tropicales d'Afrique centrale se trouvent sous des climats secs et saisonniers. La végétation adopte un schéma saisonnier comme le montrent des études de terrain localisées, mais son importance et son étendue sont à peine connues. La récente mission Sentinel-2 offre de nouvelles opportunités pour suivre la phénologie de la végétation depuis l'espace. 
Le premier objectif de cette étude était d'évaluer le potentiel des données Sentinel-2 pour suivre la saisonnalité de la végétation en Afrique centrale. L'utilisation des données Sentinel-2 visait à répondre aux questions suivantes : quelle est la saisonnalité des forêts d'Afrique centrale, et comment ce fonctionnement saisonnier est-il lié à la saisonnalité des précipitations en République centrafricaine (Mbaïki) et en République démocratique du Congo (Luki) soumises à des régimes pluviométriques inversés.
Ce travail s'est appuyé sur trois types de données : (i) les images Sentinel-2, (ii) les données au sol consistant en observations régulières des phénophases et (iii) les données pluviométriques de la mission "Global Precipitation Measurement".
Les séries temporelles de précipitations et d'"Enhanced Vegetation Index" qui en résultent ont permis respectivement de détecter le démarrage des pluies et le début de la saison. Le début de la saison a ensuite été comparé aux observations de terrain. Une analyse supplémentaire des ondelettes a été effectuée sur le site de Mbaïki pour déterminer la fréquence et le moment des événements périodiques de végétation. 
Ce travail a démontré l'adéquation des données Sentinel-2 récemment disponibles pour le suivi de la dynamique de la végétation lorsque l'ennuagement reste raisonnable. En outre, les cycles annuels de végétation ont dominé sur les sites étudiés, en lien avec la saisonnalité des précipitations. La comparaison entre les sites a confirmé une inversion de la saisonnalité de la végétation de Mbaïki à Luki en réponse à des régimes pluviométriques inversés de part et d'autre de l'Equateur. À Mbaïki, en moyenne, le signal de la forêt était 16,9 ± 4,4 jours avant le signal des précipitations, ce qui supporte le contrôle ultime des précipitations. En revanche, pour les deux sites d'étude, le démarrage le plus précoce des précipitations a entraîné le début le plus précoce de la saison en 2019, soutenant l'hypothèse des précipitations comme facteur proximal.
En conclusion, les données Sentinel-2 sont adaptées au suivi de la saisonnalité de la végétation mais nécessiteraient la combinaison avec des images supplémentaires pour faire face aux zones continuellement ennuagées.

Remote Sensing --- Phenology --- Sentinel-2 --- Tropical forest --- Central Africa --- Enhanced Vegetation Index --- Télédétection --- Phénologie --- Sentinel-2 --- Forêt tropicale --- Afrique centrale --- Enhanced Vegetation Index --- Sciences du vivant > Sciences de l'environnement & écologie

Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

Water stress damages the plant. It is associated with yield losses and pathogen proliferation. It can be caused by a lack or an excess of water. On a physiological point of view, stress is mainly manifested by stomata closure. It happens to control plant water balance. Plant water status can be estimated on the basis of physiological variables (i.e. transpiration, CO2 assimilation, stomatal conductance and leaf temperature). These ones are associated with the opening degree of the stomata. The literature review indicates that vegetation indices (i.e. NDVI) acquired by remote sensing can be correlated to physiological variables (i.e. stomatal conductance). The study deals with determining the physiological variables which allow control irrigation through remote sensing data acquisition.
To achieve this goal, an experimentation is designed on the basis of a trial field irrigated with different modalities. It was done in order to create different water stress intensities. Variables linked to water stress are quantified through a gas analyser Ciras-3. It is assumed that this machine is out to represent the plant's functioning. Thermal and multispectral data are acquired by remote sensing throughout the growing season. At the end of the season, yields and rotten roots are assessed in each treatment.
After being segmented, thermal images show a good correlation with leaves temperatures acquired through Ciras measurements. The normalized vegetation index (NDVI) index calculated from the multispectral image showed a good correlation with stomatal conductance. Temperature combined with stomatal conductance are two parameters that explain most of the variability in physiological data. On the other hand, different irrigation levels applied on the trial field were not sufficiently contrasted to obtain different stress levels between treatments. However, the data from cameras accurately represents this lack of contrast between water regimes of micro plots.
To conclude, multispectral and thermal data acquired by remote sensing would effectively control irrigation. Indeed, physiological variables associated with water stress can be known through remote sensing. However, in order to apply it on a commercial scale, the sensitivity of this data acquisition technique still needs to be tested on more or less pronounced stress levels. Le stress hydrique endommage la plante. Il est associé à des pertes de rendement et à la prolifération des pathogènes. Il peut être engendré par un manque ou un excès d’eau. D’un point de vue physiologique, le stress se manifeste notamment par la fermeture des stomates en vue d’équilibrer le bilan hydrique. Le statut hydrique de la plante peut être estimé sur base de paramètres physiologiques associés au degré d’ouverture des stomates. La revue littéraire indique que des indices de végétation (i.e. NDVI) acquis en télédétection peuvent être corrélés aux variables physiologiques (i.e. transpiration, assimilation de CO2, conductance stomatique, température de la feuille). La présente étude vise à déterminer les variables physiologiques qui permettent de piloter l’irrigation sur base d’informations acquises par télédétection.
A cette fin, une expérience est conçue sur base d’un essai en champ irrigué selon différentes modalités. Celle-ci a pour but de créer différentes intensités de stress hydrique. Les variables liées au fonctionnement de la plante sont quantifiées à l’aide d’une machine mesurant entre autres les échanges gazeux (Ciras-3). Des données thermiques et multispectrales sont acquises en télédétection au long de la saison culturale. En fin de saison, les rendements et la pourriture sont évalués dans chaque traitement.
Après avoir été segmentées, les images thermiques ont montré une représentation correcte des températures de feuilles. L’indice de végétation normalisé (NDVI) calculé à partir de l’image multispectrale a montré une corrélation satisfaisante avec la conductance stomatique. La température combinée à la conductance stomatique sont deux paramètres qui expliquent la majeure partie de la variabilité des données. En revanche, les niveaux d’irrigation appliqués n’ont pas été suffisamment contrastés pour obtenir des niveaux de stress différents entre les traitements. Toutefois, les données issues des caméras ont représenté fidèlement ce manque de contraste entre les régimes hydriques des micro parcelles.
Suite à cette étude, il est conclu que l’utilisation de données multispectrales et thermiques acquises en télédétection permettrait effectivement de piloter l’irrigation. En effet, les variables physiologiques associées au stress hydrique peuvent être connues au travers de la télédétection. Cependant, en vue d’appliquer cette technique à l’échelle commerciale, la fiabilité de cette méthode vis-à-vis du niveau de stress hydrique doit subir une étude plus approfondie.

remote sensing --- vegetation index --- water stress --- precision agriculture --- Chicory --- NDVI --- stomatal conductance --- Sciences du vivant > Agriculture & agronomie

Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

The recent revolution in deep learning has enabled considerable development in the fields of object and activity detection. Visual object detection tries to find objects of target classes with precise localisation in an image and assign each object instance a corresponding class label. At the same time, activity recognition aims to determine the actions or activities of an agent or group of agents based on sensor or video observation data. It is a very important and challenging problem to detect, identify, track, and understand the behaviour of objects through images and videos taken by various cameras. Together, objects and their activity recognition in imaging data captured by remote sensing platforms is a highly dynamic and challenging research topic. During the last decade, there has been significant growth in the number of publications in the field of object and activity recognition. In particular, many researchers have proposed application domains to identify objects and their specific behaviours from air and spaceborne imagery. This Special Issue includes papers that explore novel and challenging topics for object and activity detection in remote sensing images and videos acquired by diverse platforms.

multi-camera system --- space alignment --- UAV-assisted calibration --- cross-view matching --- spatiotemporal feature map --- view-invariant description --- air-to-ground synchronization --- tidal flat water --- YOLOv3 --- similarity algorithm for water extraction --- arbitrary-oriented object detection in satellite optical imagery --- adaptive dynamic refined single-stage transformer detector --- feature pyramid transformer --- dynamic feature refinement --- synthetic aperture radar (SAR) --- ship detection --- convolutional neural network (CNN) --- deep learning (DL) --- feature pyramid network (FPN) --- quad feature pyramid network (Quad-FPN) --- crowd estimation --- 3D simulation --- unmanned aerial vehicle --- synthetic crowd data --- invasive species --- thermal imaging --- habitat identification --- deep learning --- drone --- multiview semantic vegetation index --- urban forestry --- green view index (GVI) --- semantic segmentation --- urban vegetation --- RGB vegetation index --- n/a

Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

Except for latitudinal and elevational extremes, lizards range across a vast variety of biotopes worldwide, including environments as disparate as deserts, prairies, temperate woodlands, rainforests, or anthropic habitats. Although most species thrive on the ground, numerous lizards are fossorial, arboreal, and even aquatic, found in either fresh- or seawater. With lizards being ectotherms, accurate thermoregulation and other physiological adaptations are in most cases fundamental for their survival in such a variety of habitats. Moreover, lizard coloration may mediate thermoregulation, reproduction, and social status, among others. Lizards have also evolved some unusual antipredator adaptations, such as tail autotomy. Consequently, the astonishing morphological, ecological, and functional diversity of lizards results from extremely intense selective pressures, oftentimes opposing, many of whose interrelationships have yet to be disentangled. This Special Issue provides the international scientific community with an integrative meeting point to discuss and synthesize the current knowledge on the evolutionary pathways and mechanisms that led to today’s lizards.

Research & information: general --- Biology, life sciences --- Animals & society --- enhanced vegetation index --- Lacerta --- Mediterranean --- niche partitioning --- Sauria --- Timon --- colouration --- social signals --- Psammodromus algirus --- lizards --- altitudinal gradient --- Indochina --- Southeast Asia --- phylogeny --- Indo-Australian Archipelago --- Bent-toed geckos --- karst --- conservation --- high elevation --- hyperoxia --- sprint performance --- thermal performance curve --- thermal preference --- lizard --- autotomy --- tail --- locomotion --- performance --- temperature --- predation --- n/a

Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

The recent revolution in deep learning has enabled considerable development in the fields of object and activity detection. Visual object detection tries to find objects of target classes with precise localisation in an image and assign each object instance a corresponding class label. At the same time, activity recognition aims to determine the actions or activities of an agent or group of agents based on sensor or video observation data. It is a very important and challenging problem to detect, identify, track, and understand the behaviour of objects through images and videos taken by various cameras. Together, objects and their activity recognition in imaging data captured by remote sensing platforms is a highly dynamic and challenging research topic. During the last decade, there has been significant growth in the number of publications in the field of object and activity recognition. In particular, many researchers have proposed application domains to identify objects and their specific behaviours from air and spaceborne imagery. This Special Issue includes papers that explore novel and challenging topics for object and activity detection in remote sensing images and videos acquired by diverse platforms.

Technology: general issues --- History of engineering & technology --- multi-camera system --- space alignment --- UAV-assisted calibration --- cross-view matching --- spatiotemporal feature map --- view-invariant description --- air-to-ground synchronization --- tidal flat water --- YOLOv3 --- similarity algorithm for water extraction --- arbitrary-oriented object detection in satellite optical imagery --- adaptive dynamic refined single-stage transformer detector --- feature pyramid transformer --- dynamic feature refinement --- synthetic aperture radar (SAR) --- ship detection --- convolutional neural network (CNN) --- deep learning (DL) --- feature pyramid network (FPN) --- quad feature pyramid network (Quad-FPN) --- crowd estimation --- 3D simulation --- unmanned aerial vehicle --- synthetic crowd data --- invasive species --- thermal imaging --- habitat identification --- deep learning --- drone --- multiview semantic vegetation index --- urban forestry --- green view index (GVI) --- semantic segmentation --- urban vegetation --- RGB vegetation index --- multi-camera system --- space alignment --- UAV-assisted calibration --- cross-view matching --- spatiotemporal feature map --- view-invariant description --- air-to-ground synchronization --- tidal flat water --- YOLOv3 --- similarity algorithm for water extraction --- arbitrary-oriented object detection in satellite optical imagery --- adaptive dynamic refined single-stage transformer detector --- feature pyramid transformer --- dynamic feature refinement --- synthetic aperture radar (SAR) --- ship detection --- convolutional neural network (CNN) --- deep learning (DL) --- feature pyramid network (FPN) --- quad feature pyramid network (Quad-FPN) --- crowd estimation --- 3D simulation --- unmanned aerial vehicle --- synthetic crowd data --- invasive species --- thermal imaging --- habitat identification --- deep learning --- drone --- multiview semantic vegetation index --- urban forestry --- green view index (GVI) --- semantic segmentation --- urban vegetation --- RGB vegetation index