Remote sensing of riverine landscapes

Abstract

1. The fashion for examining riverine landscapes is changing as our technical instruments, from microprobes to satellites, expands to be able to examine the spatial and temporal relationships among biota, hydrology and geomorphology across scales from microhabitats to channel units to valleys to catchments. 2. The range of successful applications from remote sensing analyses of riverine landscapes has especially increased with the launch of many new instruments that record data across the electromagnetic spectrum. Engineering of the instruments has also improved such that knowledge of the radiometric properties of the digital data is more complete as a result of better instrumentation and installation of on‐board calibrations systems for many instruments. 3. With the development of faster processing on cheaper computers, it is now common for comprehensive data sets to be processed through algorithms that previously could only be applied to relatively small (<1 Mbyte) rasters of data. This technical advance is especially important for the statistical algorithms such as principal components and spectral mixture analysis that can decompose gradients in the spectral data. The combined effect is the production of regional views of riverine landscapes separated into components of water, vegetation and soil. 4. The landscape properties of riverine landscapes that have been most successfully measured with remote sensing data include community and habitat level classification and connectivity of waterbodies with optical and radar data. Laser and radar altimetric data measured from aircraft provide land elevations at resolutions as fine as decimetres. A remaining challenge is to achieve an exact match between the categories of landscape classification from the remote sensing analysis and data from field surveys or model outputs. 5. In contrast to many landscape properties, several water properties are now routinely measured as absolute values (water surface elevation, temperature, surface sediment concentration and algal concentration) with remote sensing. New analyses of both passive and active radar data in addition have led to measurements of inundation and wetness that are providing valuable insight into the dynamics of flooding and its effect on riverine landscapes. 6. Finally, an effective examination of the variability in landscape cover includes additional analyses of remote sensing products using pattern metrics that measure the scale of patchiness and distribution of the landscape properties. These types of variability measures at the regional scale contribute to an increased understanding of the way in which spatial heterogeneity of riverine landscapes varies across scales and how landscape filters (sensu Poff, 1997) influence the evolution of these diversity patterns.

Effects of channel morphology and sensor spatial resolution on image-derived depth estimates

“…Remote sensing has emerged as a potentially powerful 32 tool for detailed, quantitative characterization of fluvial 33 systems across broad geographic areas with improved 34 temporal coverage (Mertes, 2002). Since the early 1990's, 35 numerous studies have demonstrated the utility of remotely 36 sensed data for retrieving suspended sediment concentra-37 tions (Mertes et al, 1993), classifying in-stream habitat 38 (Legleiter & Goodchild, In press;Whited et al, 2002;39 Wright et al, 2000), and estimating water depth (Lyon et al, 40 1992;Marcus et al, 2003;Winterbottom & Gilvear, 1997).…”

Land Use

“…There are an increasing number of sensing technologies used for mapping both landscape and water properties including optical sensors, light detection and ranging (LiDAR), forward-looking infrared (FLIR), and radio detection and ranging (RADAR) (Mertes 2002). These high-tech devices are deployed from multiple platforms that range from low-altitude tethered balloons to drones, to helicopters and fixed-wing aircraft up to outer space satellites, thereby producing data with sub-centimeter to multimeter grain resolution across scales from microhabitats to channel units to valleys to catchments.…”

Performance measures for a Mississippi River reintroduction into the forested wetlands of Maurepas Swamp

“…Water temperature can also be monitored by remote sensing to allow distributed measurement of connectivity between channels and backswamps that are typically inaccessible to field crews. There are several satellites available for this monitoring (Mertes, 2002;Ritchie and others, 2003), and testing of sensitivity for each one at the onset of the river reintroduction would be beneficial to determine which is most appropriate for this application. Swarzenski and others (2008) showed that the ratio of calcium to magnesium was an effective tracer of river connection in two Louisiana marshes, and novel tracers such as this could be tested and developed in the Maurepas Swamp project area.…”