top of page
  • Writer's pictureThiaggo Tayer

The Power of Remote Sensing in Decoding Intermittent Rivers

And why remote sensing is a game-changer

Introduction

Intermittent rivers and ephemeral streams (IRES) are nature's water chameleons, changing with the seasons and keeping scientists and water managers on their toes. Unlike perennial rivers that maintain a consistent flow throughout the year, IRES may exhibit periods of flow, usually after rain events or when snow melts, followed by phases of reduced or no water flow during drier periods. IRES comprise about half of the world's river network, hosting unique aquatic life and supporting crucial ecosystem services.

And as our climate and planet continue to change, we expect them to become even more common.

Fitzroy River in Western Australia.
Image: Author

Bridging the gaps with remote sensing

Traditional fieldwork methods like gauging stations have only given us quick peeks at the true nature of IRES.

They capture moments that miss the full picture, especially when the water stops flowing—a signature trait of these rivers. Gauges just can't tell us much about where the river pools are, their shapes, or how persistent they are, and they certainly add up in costs when it comes to upkeep.

This is where remote sensing really shines. Satellite monitoring serves as a global diagnostic instrument – not just observing but actively diving into the task of tracking these rivers with great detail. Through this lens, we get to see another side of IRES's story, capturing their expansion and contraction over time. Remote sensing helps fill most of the limitations inherent to traditional gauging, offering us a clearer picture of IRES and their intricate ways.

The magic of multispectral imagery

Multispectral imagery captures pictures of the Earth in different wavelengths of light, including some beyond what the human eye can see, revealing hidden details of the environment. Legacy-bearers like the Landsat series have been our window to the Earth's changes for decades. More recently, satellites like Sentinel-2 have brought even greater clarity to our images, and Planetscope has upped the ante by offering almost daily updates, ensuring no significant change goes unnoticed.

This technology shines when analyzing the hydrology of intermittent rivers, offering a way to bypass some of the limitations we face with traditional monitoring methods. Its regular captures give us continuity and a big-picture view at a more manageable cost. To make sense of these images, we first need to map where water is, and for that, the Water Detect algorithm [1] stands out as potentially one of the best options due to its classification accuracy, open-source code, and because it does not require ancillary data (Check here to learn how to use this automated method).

By applying multispectral imagery to IRES, we can measure various aspects of these rivers, such as the area, width, length, and perimeter of river pools. These metrics are crucial for understanding the rivers' health and behaviour. They help us develop detailed metrics that illuminate the complexity and ecological roles of these rivers. The data becomes a narrative, telling the story of each river's ecological function in detail, as further explored in scientific research [2,3].

While multispectral imagery is a powerful resource, it does have its constraints. We're limited by what the imagery can visibly capture—essentially, we can't detect what we can't see. Clouds can block the view from space, and thick vegetation along the banks can hide smaller streams from sight, complicating the process of accurately mapping IRES. Also, the level of detail in each image must be suited to the size and type of landscape we're examining. For example, the 30-meter pixel resolution from Landsat may not be fine enough for mapping a tiny creek.

Image: Author

Digital elevation models (DEMs): The contours of complexity

Digital Elevation Models (DEMs) are another good resource to understand IRES.  They offer a 3D view of river basins and are great for modelling potential flow and hydrological scenarios, but they, too, come with limitations. The crux lies in their precision and time resolution: the more detailed the elevation data, the better we can predict how water flows and pools in these dynamic environments. Yet, in the arid expanses where IRES are often located, high-resolution DEMs are as scarce as the rivers themselves, and time series for this type of imagery are rare.

Image derived from 2m LiDAR overlapped with 3m Planetscope.
Image: Author

SWOT's promise

The Surface Water and Ocean Topography (SWOT) mission (https://swot.jpl.nasa.gov/) is set to be a leap forward in our understanding of Earth's waters. This recent advanced mission, launched on Dec 16th 2023, circles our planet with a satellite capable of gauging the height of water across vast expanses of oceans, lakes, reservoirs, and rivers. It's like having a network of virtual gauging stations that offer fresh data every three weeks, giving us a consistent pulse on river flows across the globe. Combined with the detailed insights from multispectral imagery, this promises a more comprehensive and nuanced comprehension of water dynamics.

Conclusion: A fluid future

As the planet changes, so too do its waterways. Remote sensing stands as a powerful ally in adapting to these changes, allowing us to better understand IRES resources. We're now moving beyond mere observation to a deeper interpretation of IRES. As we integrate the insights from Multispectral imagery, DEMs, and missions like SWOT, we edge closer to a new frontier in hydrological science—a time when every river's rhythm, from the mightiest to the most intermittent, can be documented with precision.

Yet, our journey does not end with observation and interpretation; it begins with action. We need researchers, policymakers, managers, and community leaders to engage with this data, understand it, and use it to shape a sustainable future for our water resources. We must ensure that the insights we gain from above translate into action on the ground, nurturing every waterway from source to sea.

Getting started

Now that we've explored the vast potential of remote sensing for understanding intermittent rivers, you might wonder how to put this knowledge into practice. We are here to guide you through that process with a series of forthcoming articles, each designed to equip you with the necessary tools and knowledge:

  1. Detecting Water: Water Detection in High-Resolution Satellite Images using the WaterDetect Python package.

  2. Improving Automated Water Detection Accuracy: Precision is key when it comes to mapping water bodies. This upcoming article will take you through the intricacies of enhancing the accuracy of automated water detection algorithms. We will delve into the latest techniques and adjustments that can be made to ensure that your remote sensing data is as precise and reliable as possible. For now, check this original paper as a source: "Improving the accuracy of the Water Detect algorithm using Sentinel-2, Planetscope and sharpened imagery: a case study in an intermittent river"

  3. Calculating Pixel Persistence for Mapping Rivers and Wetlands: The persistence of water in a landscape is a critical measure for environmental monitoring, especially for wetland ecosystems. Our next piece will focus on calculating pixel persistence using time-series remote sensing data. This technique is invaluable for scientists and conservationists looking to understand and document the temporal changes in wetland areas, providing insights necessary for effective management and conservation efforts. For now, check this original paper as a source: "Ecohydrological metrics derived from multispectral images to characterize surface water in an intermittent river"

Stay tuned for these articles, designed to deepen your understanding and enhance your skills in using remote sensing technology for water management. Each piece will build upon the last, creating a comprehensive suite of resources for anyone looking to harness the power of this technology in the field of hydrology.

References

[1] Cordeiro, M. C. R.; Martinez, J.-M.; Peña-Luque, S. Automatic Water Detection from Multidimensional Hierarchical Clustering for Sentinel-2 Images and a Comparison with Level 2A Processors. Remote Sensing of Environment 2021, 253, 112209. https://doi.org/10.1016/j.rse.2020.112209.

[2] Tayer, T.C., Beesley, L.S., Douglas M.M., Bourke S.A., Callow J.N., Meredith K. Ecohydrological metrics derived from multispectral images to characterize surface water in an intermittent river. Journal of Hydrology 2023, 617, 129087. https://doi.org/10.1016/j.jhydrol.2023.129087

[3] Tayer, T.C., Beesley, L.S., Douglas M.M., Bourke S.A., Meredith K. Identifying intermittent river sections with similar hydrology using remotely sensed metrics. Journal of Hydrology 2023, 626, 130266. https://doi.org/10.1016/j.jhydrol.2023.130266

[4] Tayer, T. C., Douglas, M. M., Cordeiro, M. C., Tayer, A. D., Callow, J. N., Beesley, L., & McFarlane, D. Improving the accuracy of the Water Detect algorithm using Sentinel-2, Planetscope and sharpened imagery: a case study in an intermittent river. GIScience & Remote Sensing 2023 60(1), 2168676. https://doi.org/10.1080/15481603.2023.2168676



131 views0 comments

Recent Posts

See All

Comments


bottom of page