Explore the vast range of titles available.

Aerial application of fire retardant is a critical tool for managing wildland fire spread. Retardant applications are carefully planned to maximize fire line effectiveness, improve firefighter safety, protect high-value resources and assets, and limit environmental impact. However, topography, wind, visibility, and aircraft orientation can lead to differences between planned drop locations and the actual placement of the retardant. Information on the precise placement and areal extent of the dropped retardant can provide wildland fire managers with key information to (1) adaptively manage event resources, (2) assess the effectiveness of retardant slowing or stopping fire spread, (3) document location in relation to ecologically sensitive areas; and perform or validate cost-accounting for drop services. This study uses Sentinel-2 satellite data and commonly used machine learning classifiers to test an automated approach for detecting and mapping retardant application. We show that a multiclass model (retardant, burned, unburned, and cloud artifact classes) outperforms a single-class retardant model and that image differencing (post-application minus pre-application) outperforms single-image models. Compared to the random forest and support vector machine, the gradient boosting model performed the best with an overall accuracy of 0.88 and an F1 Score of 0.76 for fire retardant, though results were comparable for all three models. Our approach maps the full areal extent of the dropped retardant within minutes of image availability, rather than linear representations currently mapped by aerial GPS surveys. The development of this capability allows for the rapid assessment of retardant effectiveness and documentation of placement in relation to sensitive environments.

Recent advances in image classification of fine spatial resolution imagery from unoccupied aircraft systems (UASs) have allowed for mapping vegetation based on both multispectral reflectance and fine textural details. Convolutional neural network (CNN)-based models can take advantage of the spatial detail present in UAS imagery by implicitly learning shapes and textures associated with classes to produce highly accurate maps. However, the spatial resolution of UAS data is infrequently examined in CNN classification, and there are important tradeoffs between spatial resolution and classification accuracy. To improve the understanding of the relationship between spatial resolution and classification accuracy for a CNN-based model, we captured 7.6 cm imagery with a UAS in a wetland environment containing graminoid (grass-like) plant species and simulated a range of spatial resolutions up to 76.0 cm. We evaluated two methods for the simulation of coarser spatial resolution imagery, averaging before and after orthomosaic stitching, and then trained and applied a U-Net CNN model for each resolution and method. We found untuned overall accuracies exceeding 70% at the finest spatial resolutions, but classification accuracy decreased as spatial resolution coarsened, particularly beyond a 22.8 cm resolution. Coarsening the spatial resolution from 7.6 cm to 22.8 cm could permit a ninefold increase in survey area, with only a moderate reduction in classification accuracy. This study provides insight into the impact of the spatial resolution on deep learning semantic segmentation performance and information that can potentially be useful for optimizing precise UAS-based mapping projects.

The introduction of Zika virus to the USA in 2015 engendered heightened interest in its known vectors. Aedes aegypti is the primary vector, with Ae. albopictus considered a potential secondary vector, together with several other possible marginal vectors. In Delaware, Ae. aegypti has been collected rarely, but no breeding populations were detected during past intensive statewide surveillance efforts. However, there is an abundance of Ae. albopictus statewide. Both species are container breeders and are peri-domestic—increasing the risk for virus transmission to humans. From July through September 2017, Delaware Mosquito Control conducted surveillance in 16 container-breeding hot spots to search for Ae. aegypti , and also ascertain the virus-positive pool rates of Ae. albopictus and Ae. triseriatus for West Nile virus (WNV) and Zika virus (ZIKV). The survey concluded that there were no known breeding populations of Ae. aegypti in Delaware, and no WNV- or ZIKV-positive pools were detected among pools of mosquitoes of the aforementioned species.

A relatively new machine learning method in remote sensing, deep learning, has gained interest in recent years. Deep learning models are often applied to the analysis of fine spatial resolution imagery acquired by relatively low-cost unoccupied aircraft systems (UAS). Deep learning models such as convolutional neural networks (CNNs) can classify imagery using complex, nonlinear relationships using spatial and spectral information. Particularly, CNNs have powerful capabilities to learn spatial patterns in an image. However, the spatial detail in an image typically decreases with spatial resolution. In August 2020 at the Howard Slough Waterfowl Management Area, near Ogden, Utah, the Utah Department of Natural Resources (DNR) flew a UAS that captured 7.6 cm resolution imagery to monitor an invasive plants species, Phragmites australis, along with native vegetation and land covers. They applied a CNN model that reliably made predictions, but their UAS had a slow capture rate at about 1.2 km2 /day. It is possible to increase the capture rate by using a different UAS and increasing the flight altitude, with a consequence that the spatial resolution will coarsen. For this reason, the Utah DNR is considering different options for UAS platforms, however, they need to evaluate the effect of a coarser spatial resolution on the CNN model’s classification accuracy. To test this relationship, this study used a subset of the 2020 image data and simulated coarser resolutions using spatial averaging to 15, 23, 30, 38, 53, and 76 cm resolution. A CNN model was produced at each resolution, and the accuracy metrics were compared to that of the original 7.6 cm resolution imagery. The testing indicated that model accuracy decreased as the spatial resolution coarsened. At the 23 cm resolution, the model overall accuracy decreased from 5-11% compared to the original resolution. One of the options for a new UAS system can capture about 65 km2 /day at a 20 cm resolution. If selected, the large increase in capture rate is likely worth the relatively small decrease in model accuracy.