Eye in the Sky: Using UAV Imagery of Seasonal Riverine Canopy Growth to Model Water Temperature

Willis, Ann; Holmes, Eric

doi:10.3390/hydrology6010006

Cited by 6 publications

(3 citation statements)

References 51 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…When studied at the reach scale, however, macrophytes become a seasonal control of spring-fed thermal regimes at both the local and reach scales. Extensive emergent macrophyte growth forms a seasonal riverine canopy and creates a buffer against warming [46,67]. Within reaches where extensive, emergent macrophyte growth occurs, spring-fed streams reach their annual maximum temperatures over 30 days earlier than runoff streams and have less diurnal heating during what is typically the warmest period of the year [46].…”

Section: Macrophytes and The Stream Environmentmentioning

confidence: 99%

Not All Rivers Are Created Equal: The Importance of Spring-Fed Rivers under a Changing Climate

Lusardi

Nichols

Willis

et al. 2021

Water

Self Cite

View full text Add to dashboard Cite

In the Western United States, volcanic spring-fed rivers are anticipated to become increasingly more important for salmonids and other native fishes, as these rivers will retain coldwater habitats as the climate warms. Despite this, little is known about the hydro-biogeochemical interactions within these ecosystems. A review of existing literature on spring-fed rivers, coupled with a decade of research on volcanic spring-fed rivers of northern California, finds that these systems are exceptionally productive and exhibit stable environmental conditions. These unique conditions stem from hydrogeologic processes typical of young volcanic terrains. Aquatic macrophytes, common to some nutrient-rich spring-fed systems, play a disproportionate role in hydrologic and geomorphic processes by facilitating ecological interactions and velocity conditions that improve juvenile salmonid growth. We find that volcanic spring-fed rivers are also resilient to climate change, due not only to their ability to dampen water temperature changes through deep groundwater flow but also because of their nutrient-driven high ecosystem productivity, which may enable coldwater species to metabolically compensate for marginal increases in water temperature. Understanding the fundamental geomorphic and ecological differences between these rare ecosystems and their numerically dominant runoff rivers is essential for developing long-term conservation strategies for coldwater species under a rapidly changing climate.

show abstract

Section: Macrophytes and The Stream Environmentmentioning

confidence: 99%

Not All Rivers Are Created Equal: The Importance of Spring-Fed Rivers under a Changing Climate

Lusardi

Nichols

Willis

et al. 2021

Water

Self Cite

View full text Add to dashboard Cite

show abstract

“…There are many factors affecting the remote measurement of this surface temperature, such as air temperature and humidity, wind speed, distance from objects, recording time, and the sensor used for the measurement [10][11][12]. Using an Unmanned Aerial Vehicle (UAV) and thermal camera, it is possible to study water surface temperatures [13][14][15][16][17]. The most commonly used wavelength range is 8-14 µm [18] at resolutions of 640 × 512 pixels [19], and the final products are thermal orthophotos [20][21][22][23][24].…”

Section: Introductionmentioning

confidence: 99%

Attempt to Combine Physicochemical Data with Thermal Remote Sensing to Determine the Extent of Water Mixing between River and Lake

2022

View full text Add to dashboard Cite

The mixing of river and lake waters is important for the functioning of a reservoir, especially in the case of shallow polymictic reservoirs such as Lake Swarzędzkie. The extent of this mixing depends largely on the river flow rate. In lakes, which rivers with low flow values flow through, it should be expected that the flow currents only reach the narrow zone adjacent to the mouth of the river to the lake. The water of rivers generally has different chemical compositions and physicochemical parameters in relation to lake water. Therefore, to determine the range of the river in the lake and characterize the water mixing, measurements of temperature, electrolytic conductivity, and the concentrations of selected chemical elements were made in the estuary zone and at other points located on the lake and on the river near the tributary. In addition, the values and directions of horizontal currents were determined, and thermal photos were taken from a low-altitude ceiling.

show abstract

“…Bjerklie et al, 2003;Kuhn et al, 2019;Langat et al, 2020)(eg Bjerklie et al, 2003;Kuhn et al, 2019;Langat et al, 2020) , algorithms for quantifying in-stream metrics such as grain size, water depth, or sub-meter scale temperature topography (e.g. Black et al, 2014;Dietrich, 2016;Willis and Holmes, 2019) require the acquisition of very high resolution (usually <10 cm) RGB images which are not available from any satellite platform. Carbonneau and Piégay (2012a) define 'hyperspatial' resolution threshold as <10 cm and state that such images have increasing value in the analysis of river systems.…”

Section: Introductionmentioning

confidence: 99%

Adopting deep learning methods for airborne RGB fluvial scene classification.

Carbonneau¹,

Dugdale²,

Breckon³

et al. 2020

Preprint

View full text Add to dashboard Cite

River environments are among the world’s most threatened ecosystems. Enabled by the rapid development of drone technology, hyperspatial resolution (<10 cm) images of fluvial scenes are now a common data source used to better understand these sensitive habitats. However, the task of image classification remains challenging for this type of imagery and the application of traditional classification algorithms such as maximum likelihood, still in common use among the river remote sensing community, yields unsatisfactory results. We explore the possibility that a classifier of river imagery based on deep learning methods can provide a significant improvement in our ability to classify fluvial scenes. We assemble a dataset composed of existing RGB images from 11 rivers in Canada, Italy, Japan, the UK, and Costa Rica. The images were labelled into 5 classes. In total, >5 billion pixels were labelled and partitioned for the tasks of training (1 billion pixels) and validation (4 billion pixels). We develop a novel supervised learning workflow based on the NASNet convolutional neural network (CNN) called ‘CNN-Supervised Classification’ (CSC). First, we compare the classification success of maximum likelihood, a multilayer perceptron, a random forest, and CSC. Results show F1 scores (a best-practice quality meter in machine learning) of 71%, 78%, 72% and 95%, respectively. Second, we train our classifier using data for 5 of 11 rivers. We then predict the validation data for all 11 rivers. For the 5 rivers that were used in model training, F1 scores reach 98%. For the 6 rivers not used in model training, F1 scores are 90%. We reach two conclusions. First, in the traditional workflow where images are classified one at a time, CSC delivers an unprecedented mix of labour savings and classification F1 scores above 95%. Second, deep learning can determine land-cover classifications (F1 = 90%) for rivers not used in training. This demonstrates the potential to train a generalised open-source deep learning model for airborne river surveys suitable for most rivers ‘out of the box’. Research efforts should now focus on further development of a new generation of deep learning classification tools that will encode human image interpretation abilities and allow for fully automated, potentially real-time, interpretation of riverine landscape images.

show abstract

Eye in the Sky: Using UAV Imagery of Seasonal Riverine Canopy Growth to Model Water Temperature

Cited by 6 publications

References 51 publications

Not All Rivers Are Created Equal: The Importance of Spring-Fed Rivers under a Changing Climate

Not All Rivers Are Created Equal: The Importance of Spring-Fed Rivers under a Changing Climate

Attempt to Combine Physicochemical Data with Thermal Remote Sensing to Determine the Extent of Water Mixing between River and Lake

Adopting deep learning methods for airborne RGB fluvial scene classification.

Contact Info

Product

Resources

About