The rapid development of self-driving vehicles requires integrating a sophisticated sensing system to address the various obstacles posed by road traffic efficiently. While several datasets are available to support object detection in autonomous vehicles, it is crucial to carefully evaluate the suitability of these datasets for different weather conditions across the globe. In response to this requirement, we present a novel dataset named the Canadian Vehicle Datasets (CVD). Subsequently, we present deep learning models that use this dataset. The CVD comprises street-level videos which were recorded by Thales, Canada. These videos were collected with high-quality cameras mounted on a vehicle in the Canadian province of Quebec. The recordings were made during daytime and nighttime, capturing weather conditions such as hazy, snowy, rainy, gloomy, nighttime and sunny days. A total of 10000 images of vehicles and other road assets are extracted from the collected videos. A total of 8388 images were annotated with corresponding generated labels 27766 with their respective 11 different classes. We analyzed the performance of the YOLOv8 model trained using the existing RoboFlow dataset. Then, we compared it with the model trained on the expanded version of RoboFlow using the proposed weather-specific dataset, CVD. Final values of improved accuracy of 73.26 %, 72.84 %, and 73.47 % (Precision/Recall/mAP) were reported upon adding the proposed dataset. Finally, the model trained on this diverse dataset exhibits heightened robustness and proves highly beneficial for both autonomous and conventional vehicle operations, making it applicable not only in Canada but also in other countries with comparable weather conditions.

We used a full remote sensing-based approach to assess plant species diversity in large and inaccessible areas affected by Lantana camara L., a common invasive species within the deciduous forests of Western Himalayan region of India, using spectral heterogeneity information extracted from optical data. The spread of L. camara was precisely mapped by Pléiades 1A data, followed by comparing Pléiades 1A, RapidEye and Landsat-8 OLI – assessed plant species diversities in invaded areas. The single plant species analysis was improved by Pléiades 1A-based diversity analysis, and higher species diversity values were observed for mixed vegetation cover. Furthermore, lower Coefficient of Variation and Renyi diversity values were observed where L. camara was the only species, while higher variations were observed in areas with a mixed spectral reflectance. This study was concluded to add a crucial baseline to the previous studies on remote sensing-based solutions for rapid estimation of biodiversity attributes.

Late frost can cause damage to trees, especially to the developing bud of broadleaf species in spring. Through long-term adaptation, plants adjust leaf phenology to achieve an optimal trade-off between growing season length and frost avoidance. In this study, we aim to assess ecotypic differentiation in leaf development of sugar maple populations planted in a common garden. A total of 272 sugar maple seedlings from 29 Canadian provenances were planted at the northern boundary of the natural range, and the phenological phases of bud and leaf development were monitored during spring 2019. The wide geographical area under evaluation showed a complex seasonal pattern of temperature, with spring warming occurring later in the north and close to the sea. Overall, leaf development lasted between 20 and 36 days, from the end of May to end of June. We observed different timings and rates of leaf development among provenances, demonstrating the occurrence of ecotypes in this species. Minimum April temperatures of the original sites were able to explain such differences, while maximum April temperatures were not significant. Seedlings from sites with colder minimum April temperatures completed leaf development earlier and faster. On average, leaf development diverged by up to 6 days among provenances, with minimum April temperatures ranging from −3 to 3 °C. Our results demonstrated that the avoidance of late spring frost is a driving force of leaf development in sugar maple populations. In the colder sites, the growing season is a limiting factor for tree growth. Thus, when thermal conditions become favorable in spring, an earlier growth reactivation and high metabolic activity ensure a fast leaf emission, which maximizes the period available for photosynthesis and growth. These patterns demonstrate the long-term phenological adaptation of sugar maple populations to local climatic conditions and suggest the importance of frost events for leaf development.

Satellite remote sensing is a widely accessible tool to investigate the spatiotemporal variations in the bud phenology of evergreen species, which show limited seasonal changes in canopy greenness. However, there is a need for precise and compatible data to compare remote sensing time series with field observations. In this study, fortnightly MODIS-NDVI was fitted using double-logistic functions and calibrated using ordinal logit models with the sequential phases of bud phenology collected during 2015, 2017 and 2018 in a black spruce stand. Bud break and bud set were spatialized for the period 2009–2018 across 5000 stands in Quebec, Canada. The first phase of bud break and the last phase of bud set were observed in the field in mid-May and at the beginning of September, when NDVI was 80.5% and 92.2% of its maximum amplitude, respectively. The NDVI rate of change was estimated at 0.07 in spring and 0.04 in autumn. When spatialized on the black spruce stands, bud break was detected earlier in the southwestern regions (April–May), and later in the northeastern regions (mid to end of June). No clear trend was observed for bud set, with different patterns being detected among the years. Overall, the process bud break and bud set lasted 51 and 87 days, respectively. Our results demonstrate the potential of satellite remote sensing for providing reliable timings of bud phenological events using calibrated NDVI time series on wide regions that are remote or with limited access.