The term "river health" refers to the assessment of river conditions, but it is still unclear what aspects of river health sets of ecosystem-level indicators actually identify or how physical, chemical, and biological characteristics may be integrated into measures rather than observations of cause and effect (Norris and Thoms 1999). Assessing environmental flow (eflow) is crucial for maintaining river health as it helps in understanding and managing the ecological requirements of rivers and their associated ecosystems.
The term environmental flow (eflow) has recently become increasingly popular as concerns about the destruction of freshwater ecosystems and the impacts of development activities (i.e., urban development and energy production) on river have intensified. Eflow is defined as "the quantity, timing, and quality of water flows required to sustain freshwater and estuarine ecosystems, and the human livelihoods and well-being that depend on these ecosystems" (Brisbane Declaration 2007). Alternatively, eflow is described as the foundation of water security for achieving sustainable development. Managing eflow is relevant to meet the most targets of SDG 6, but especially SDG 6.4 on water use efficiency (6.4.2 level of water stress) and SDG target 6.6 on the protection of water-dependent ecosystems.
Numerous academics (e.g., Pastor et al. 2014; Acreman et al. 2014; Hairan et al. 2021) underlined the significance of eflow by emphasizing over the necessity of flow for the support of aquatic organism life cycles (e.g., fish flow), the maintenance of water quality (e.g., maintenance flow), and the provision of habitat for riparian plants (e.g., ecological flow). There are over 300 methods of eflow assessment, from simple ones dating back to the 1970s to holistic methods of today (see figure 1).
Overall, eflow assessment methods can be grouped into four categories: hydrological, hydraulic rating, habitat simulation, and holistic methodologies (Tharme, 2003). Hydrological methods use summary statistics from hydrological data sets to set a “minimum flow” of the river. In hydraulic-rating methods, simple hydraulic variables (e.g., wetted perimeter) are used to predict how these change with variations in discharge. Habitat-simulation techniques are used to model how much of the experimented hydraulic habitat would be available over a range of flows. Lastly, holistic methods address the overall condition of river ecosystems connected to societal, resource and economic issues and cover the full spectrum of river conditions. The required data input and example applications are described in the table 1.
Traditionally, guaging stations have been used to measure variety of data including water height, chemistry, velocity and temperature which are necessary for assessing quantity, timing, and quality of water flows (Grimes and Diop 2003). According to Stisen et al., existing gauges nowadays are providing false recording data in some countries due to the poor management practices. To a certain extent, the problem can be solved by applying various models using meteorological variables, as well as the appropriate values of catchment and model parameters which can be extracted from the remote sensing images (Stisen et al. 2008).
Fundamentals of eflow assessment
Environmetal flow (eflow) assessment is based on understanding and analysing the natural flow regime of a river. The flow regime of a river is the statistical description of its long-term behaviour, including the timing, duration, and magnitude of its flows. It is characterized by a variety of components, including high flows, low flows, and their seasonality (UNESCO 2012). The river flow regime is influenced by natural processes and anthropogenic activities, and is a key determinant of the ecological, economic, and social functioning of river systems. Figure 2 shows how the river infrastructure projects such as dams, weirs, barrages, levees, and other development activities on riverine ecosystems increased the river fragmentation and flow alteration.
Protecting and restoring the river flow regimes ensure sustaining eflow and it is a major aspect of river basin management. The six components of a flow regime that have an impact on river ecology and ecosystems are frequency, magnitude, timing, duration, rate of change, and overall variability of flow (Acreman et al. 2014). The following questions on each component can help in planning the assessment:
- Frequency: How often do certain flows or levels occur?
- Magnitude: How much do flows or what levels occur?
- Timing: When do certain flows or levels occur?
- Duration: How long do certain flows or levels last?
- Rate of change: How fast do flows or levels change from one condition to another?
- Overall variability of flow: How do flows function differently in dry, rainy and flood season
Remote sensing applications in environmental flow assessment
To learn more about the information on remote sensing of meteo-hydrological components, read this Space4Water article by Kitambo (2022). The satellite missions of Sentinel-1, Sentinel-2 and Landsat 8 which are providing free global data is described in Table 2 and briefly discussed here since they are important in this study. While the data from these satellites is freely available and governed under open licenses, it requires registration and approval to access.
Spatial resolution of these satellites varies from 5-100 m and temporal resolution varies from 5-16 days. The lower the number of spatial resolutions, the more detail of the objects can be seen. Depending on the sensor type, spectral resolution is also different. Many sensors are multispectral having 3-10 bands, while some sensors are hyperspectral which have hundreds to even thousands of bands. The spectral resolution is finer when the wavelength range for a given band is narrower.
Satellite Mission |
Launch Date |
Operation life |
Type of sensor |
Resolution (Spatial /Temporal) |
Data portal |
Sentinel-1 | April 3, 2014 | 7 years with consumables for 12 years (still operational) | Optical: Synthetic Aperture Radar (SAR) | 5-20m (8days) | European Space Agency (Copernicus Open Access Hub) |
Sentinel-2 | June 23, 2015 | 7 years with consumables for 12 years (still operational) | Optical: multispectral imaging instrument (MSI) | 10-60m (5/10 days) | European Space Agency (Copernicus Open Access Hub) |
Landsat 8 | February 11, 2013 | Still operational and expected more years to be operated | Optical | 15-100 m (16 days) | United States Geological Survey (USGS Earth Explorer) |
Sentinel-1
Sentinel-1 is a synthetic aperture radar (SAR) based earth observation system consisting of two identical satellites: Sentinel-1A and Sentinel-1B. It is equipped with a C-band SAR having a frequency of 5.405 GHz and wavelength of 56 mm (Ahmad and Kim 2019). It provides imagery in four different acquisition modes (Figure 3): Stripmap (SM), Interferometric Wide Swath (IW), Extra Wide Swath (EW), and Wave mode (WV), and is independent of weather conditions. The selection of an acquisition mode is influenced by the aim of imaging, download capacity, and the total time available for acquisitions at each orbital pass.
Ahmad and Kim (2019) introduced streamflow estimation using Sentinel-1 images. In order to correct various types of image distortions and errors caused by the sensor, the images are first pre-processed for thermal noise, radiometric calibration, speckle filtering and terrain correction using the ESA Sentinels Application Platform software tool (SNAP). As a next step, the satellite look angle and local incident angle is addressed by adopting the image histogram matching technique to remove the inconsistency due to different atmospheric conditions. This is followed by using the selective area filtering to exclude the non-water part and to focus the analysis exclusively on the water area. In a final step, the Optimum Threshold Classification (OTC) method was used to determine the optimal threshold intensity which identifies the water area. As a result of this process, a series of the water area corresponding to the multi-temporal SAR images was estimated for each of the incremental threshold intensity (Figure 4).
Sentinel-2
Among the multispectral images freely distributed with systematic global coverage, the Sentinel-2 mission provides the highest spatial resolution and revisit frequency. It comprises a constellation of two identical satellites, Sentinel-2A and Sentinel-2B. Both satellites carry on board the Multispectral Instrument (MSI), which provides 13 spectral bands in the visible, near infrared (NIR) and short-wave infrared (SWIR) wavelengths, with four bands at 10 m (B2, B3, B4, B8), six bands at 20 m (B5, B6, B7, B8a, B11, B12), and three bands at 60 m (B1, B9, B10). In Table 3, the spectral characteristics of Sentinel-2 data are shown.
In a research study assessing the flowing status of non-perennial rivers by Cavallo et al. (2022), Sentinel-2 multispectral satellite images are used since these are fulfilling the required spatial and temporal resolution needed. All Sentinel-2 bands are used except the atmospheric bands B1, B9 and B10. The images are processed with the calculation of near infrared (NIR) and short-wave infrared (SWIR) to generate false colour images (FCIs) in which the water pixels stand out from the background. Three flowing statuses were distinguished: (1) continuous flow, (2) disconnected pools and ponds, (3) dry bed by visual interpretation of the FCIs. Comparison with ground truths: a field survey and Google Earth Pro image confirm the actuality of being able to observe the flowing state according to the FCI analysis. A comparison of a Google Earth Pro image and an FCI image is as shown in figure 5. This result allows the identification of objects with a minimum width between 6 and 15 m.
Landsat-8
Landsat-8 carries two sensors: The Operational Land Imager (OLI) and Thermal Infrared Sensor (TIRS). A 12-bit dynamic range is used by OLI to record data with higher radiometric precision, which improves the signal-to-noise ratio overall. OLI provides nine spectral bands at 30 m, and TIRS provides two spectral bands at 100m. The description of band wavelengths and uses is as shown in table 4.
Yue et al. (2021) checked whether or not environmental flow needs are met by comparing the surface widths of two rivers. The first river surface width (WI) measurement is done by interpreting cloud free Landsat-8 scenes through a manual visual method (Figure 6). The second river surface width is calculated by combining the river cross-section data with the ecological water level (WE) which can be calculated from a relation curve of flow rate and the water level based on the daily hydrological data; in our example from 2011 to 2017. Following the Tennant method, 10% of mean annual flow is set up as minimum flow. When WI > WE, the eflow requirements are met; contrarily, when WI < WE, the eflow requirements are not met. This result allows examining eflow in larger river basins or several basin types to account for the ecological diversity, connectivity, and competing water demands in these systems. It includes a range of river sizes, including small rivers, and the minimum width requirement depends on the specific objectives and characteristics of the waterbody being assessed.
Limitations and solutions of remote sensing for environmental flow assessment
It is unnecessary to reiterate how important remote sensing is for the monitoring process in areas with a lack of data. But given that a lot of satellite images and remotely sensed data would be collected based on the size of the area and the parameters needed, there are some drawbacks to remote sensing that you should be aware of. Careful consideration of the factors as described in table 5, would help ensure the accuracy and effectiveness of the assessment.
Limitations |
Description of Limitations |
Solutions |
Description of Solutions |
Atmospheric interference | Atmospheric conditions such as clouds, haze and aerosols could reduce the quality of data collected. | Atmospheric correction |
Atmospheric effects can be corrected by: |
Sensor limitations | The degree to which image resolution is offered determines the accuracy and reliability of the data. In some cases, specialized sensors might be needed. | Sensor fusion |
To improve quality and accurate representation of the scene being observed, data from multiple sensors must be combined and resolved by processing the image distortions and errors. |
Data processing time | Pre-processing and analysing data using complex algorithms and software can be time-consuming and specialized expertise is demanded. | Shortening the data processing time |
1. Parallel processing by breaking down the data into smaller subsets and processing them with multiple computers, |
Cost | High cost is expected to achieve high-resolution data or for large areas that require multiple acquisitions (i.e., hardware, software, experienced staff, and training) | Reducing cost |
1. Collaboration: sharing resources and expertise with other organizations or research groups, |
Data access limitation | Data restriction can occur due to licensing agreements, data sharing policies, or political and economic factors. | Data access acquisition |
Data restriction problem could be avoided by applying above mentioned resolving strategy with open-source multispectral data. |
Complexity of hydrological processes | Hydrological processes are complex and vary spatially and temporally, which can make it challenging to interpret remote sensing data accurately. | Resolving complexity of hydrological processes |
Integration of multiple remote sensing data sources; use of machine learning algorithms; calibration and validation; collecting field research data; collaboration and interdisciplinary approaches would support in interpreting of remote sensing data. |
Limited understanding of ecological needs and data integration | Knowledge and experience of real ground condition is needed even though remote sensing data can provide information on water quantity. | Enhancing understanding of ecological needs and data integration | The best result would be integrating field observations, remote sensing datasets and expert local knowledge. |
Conclusion
Rivers are the arteries of the Earth, and they are connecting the landscape and people through the flow of water. Environmental flow refers to a specific flow regime of river, capable of sustaining the association of aquatic habitats and ecosystems processes. The term eflow has a variety of names worldwide within different (scientific) disciplines (e.g., instream flow, ecological reserve, ecological demand of water, environmental water allocation, compensation flow or minimum flow). Being a concept that has evolved over time, eflow shifted from referring to the conventional assignment of minimum water amounts to a more holistic understanding of a river system and its dynamic.
Eflow assessment is helpful in ensuring that the natural flow regimes are maintained, and the water needs of ecosystems are met. Assessing Eflow eflow is a first step to maintain natural flow regimes, which is essential to fulfil the water needs helps ensures that the water needs of ecosystems are met by ensuring that the natural flow regimes are maintained. It also helps to identify , and the critical habitats and vulnerable ecosystems and maintain critical habitatsare identified. In the sustainable management of water resourcesR, remote sensing plays a crucial role in the sustainable management of water resources, especially with regards to monitoring, planning and informedming decision- making. For example, remote sensing allows to monitor large areas continuously and detect changes over time. The accuracy of e-flow monitoring with remote sensing images is still not guaranteed owing to the limitations of remote sensing system and the complexity of ecological responses. Nonetheless, integrating data from multiple sources including in-situ measurements, hydrological models, and satellite imagery can promise improving data quality and reducing limitations of either mode of acquisition.
Future research and development could focus on developing and improving (1) satellite sensors that can capture images at higher spatial and enhanced temporal resolutions; (2) methods for integrating and analysing data from multiple sources; (3) advanced data analysis techniques to extract meaningful information from large volumes of satellite imagery data, (3) holistic methodologies considering the interacting components of aquatic systems including sediments and (4) standardized methods for validating and verifying eflow assessments. While satellite data in very high-resolution exist and are offered by the private sector, making them accessible in the open domain is crucial to ensure eflow assessment for the many watersheds globally that we need to sustainably maintain.
Acreman, Mike, Angela H Arthington, Matthew J Colloff, Carol Couch, Neville D Crossman, Fiona Dyer, Ian Overton, Carmel A Pollino, Michael J Stewardson, and William Young. 2014. “Environmental Flows for Natural, Hybrid, and Novel Riverine Ecosystems in a Changing World.” Frontiers in Ecology and the Environment 12 (8): 466–73. https://doi.org/10.1890/130134.
Ahmad, Waqas, and Dongkyun Kim. 2019. “Estimation of Flow in Various Sizes of Streams Using the Sentinel-1 Synthetic Aperture Radar (SAR) Data in Han River Basin, Korea.” International Journal of Applied Earth Observation and Geoinformation 83 (November): 101930. https://doi.org/10.1016/j.jag.2019.101930.
Baldan, Damiano, David Cunillera-Montcusí, Andrea Funk, and Thomas Hein. 2022. “Introducing ‘Riverconn’: An R Package to Assess River Connectivity Indices.” Environmental Modelling & Software 156 (October): 105470. https://doi.org/10.1016/j.envsoft.2022.105470.
Bonzi, Christopher, Günther Grill, Phyoe Kyaw Htet, Natalie Shahbol, and Frank Van der Valk. 2020. “Mapping Myanmar’s Free-Flowing Rivers.” The World Wide Fund for Nature (WWF).
Cavallo, Carmela, Maria Nicolina Papa, Giovanni Negro, Massimiliano Gargiulo, Giuseppe Ruello, and Paolo Vezza. 2022. “Exploiting Sentinel-2 Dataset to Assess Flow Intermittency in Non-Perennial Rivers.” Scientific Reports 12 (1): 21756. https://doi.org/10.1038/s41598-022-26034-z.
Čotar, Klemen, Krištof Oštir, and Žiga Kokalj. 2016. “Radar Satellite Imagery and Automatic Detection of Water Bodies,” 5–15.
Grimes, D. I. F., and M. Diop. 2003. “Satellite-Based Rainfall Estimation for River Flow Forecasting in Africa. I: Rainfall Estimates and Hydrological Forecasts.” Hydrological Sciences Journal 48 (4): 567–84. https://doi.org/10.1623/hysj.48.4.567.51410.
Jumani, Suman, Matthew J Deitch, David Kaplan, Elizabeth P Anderson, Jagdish Krishnaswamy, Vincent Lecours, and Matt R Whiles. 2020. “River Fragmentation and Flow Alteration Metrics: A Review of Methods and Directions for Future Research.” Environmental Research Letters 15 (12): 123009. https://doi.org/10.1088/1748-9326/abcb37.
Li, Jiaxin, Ronghua Ma, Zhigang Cao, Kun Xue, Junfeng Xiong, Minqi Hu, and Xuejiao Feng. 2022. “Satellite Detection of Surface Water Extent: A Review of Methodology.” Water 14 (7): 1148. https://doi.org/10.3390/w14071148.
Masud, Muhammad Jehanzeb, and Wim GM Bastiaanssen. 2017. “Remote Sensing and GIS Applications in Water Resources Management,” 351–73.
Mokhtari, Ali, Morteza Sadeghi, Yasamin Afrasiabian, and Kang Yu. 2023. “OPTRAM-ET: A Novel Approach to Remote Sensing of Actual Evapotranspiration Applied to Sentinel-2 and Landsat-8 Observations.” Remote Sensing of Environment 286 (March): 113443. https://doi.org/10.1016/j.rse.2022.113443.
Norris, Richard H., and Martin C. Thoms. 1999. “What Is River Health? What Is River Health?” Freshwater Biology 41 (2): 197–209. https://doi.org/10.1046/j.1365-2427.1999.00425.x.
Sahoo, Debi Prasad, Bhabagrahi Sahoo, Manoj Kumar Tiwari, and Goutam Kumar Behera. 2022. “Integrated Remote Sensing and Machine Learning Tools for Estimating Ecological Flow Regimes in Tropical River Reaches.” Journal of Environmental Management 322 (November): 116121. https://doi.org/10.1016/j.jenvman.2022.116121.
Sood, Aditya, Vladimir Smakhtin, Nishadi Eriyagama, Karen G. Villholth, Nirosha Liyanage, Yoshihide Wada, Girma Ebrahim, and Chris Dickens. 2017. Global Environmental Flow Information for the Sustainable Development Goals. IWMI Research Report 168. Colombo, Sri Lanka: International Water Management Institute (IWMI).
Stisen, Simon, Karsten H. Jensen, Inge Sandholt, and David I.F. Grimes. 2008. “A Remote Sensing Driven Distributed Hydrological Model of the Senegal River Basin.” Journal of Hydrology 354 (1–4): 131–48. https://doi.org/10.1016/j.jhydrol.2008.03.006.
Tharme, R. E. 2003. “A Global Perspective on Environmental Flow Assessment: Emerging Trends in the Development and Application of Environmental Flow Methodologies for Rivers.” River Research and Applications 19 (5–6): 397–441. https://doi.org/10.1002/rra.736.
Unesco, ed. 2012. Managing Water under Uncertainty and Risk. Vol. 3: Facing the Challenges. Vol. 3. United Nations World Water Development Report, 4.2012,3. Paris: UNESCO [u.a.].
World Meteorological Organization (WMO). 2019. Guidance on Environmental Flows: Integrating E-Flow Science with Fluvial Geomorphology to Maintain Ecosystem Services. 2019th ed. WMO- No.1235. WMO.