- freely available
ISPRS Int. J. Geo-Inf. 2018, 7(8), 314; https://doi.org/10.3390/ijgi7080314
2. Study Site
- BW satellite images, 1968: digital scanned copies of photogrammetric BW films acquired in 1968 from one of the first generation of U.S. photo intelligence satellite systems, namely CORONA. Images were originally used by U.S. intelligence agencies and eventually declassified by Executive Order 12,951 in 1995 because of their potential usefulness in global change research. Images used in this study were in part downloaded for free and in part purchased from USGS (United States Geological Survey). The images were acquired in stereo mode by means of a system of two twin cameras named KeyHole (KH) 4A. The spatial resolution was about 2.8 m.Mission data: 1049-2 12/22/1968; film IDs: DS1049-2155DA038, DS1049-2155DA042, DS1049-2155DF044, DS1049-2155DF045, DS1049-2155DA045.
- RGB orthophoto, 2007: color orthophotos acquired in 2007. Images were geocoded in WGS84 (World Geodetic System 1984), and coordinates were given in a UTM cartographic plane. Images used in this study were saved from the Albanian State Authority for Geospatial Information (ASIG) portal using WMTS (Web Map Tile Services).
- Multi-band LANDSAT imagery, 1972–2017: a set of multi-band LANDSAT imagery from 1972–2017. Considering the typical time scale of geomorphological phenomena involving natural river systems of interest, a scene every 5–7 years was considered. When possible, and in order to guarantee an acceptable presence of vegetation, scenes within the May–July period were selected. All LANDSAT data were downloaded for free from the Earth-Explorer website. Table 2 contains details concerning LANDSAT data, including mission codes, sensing instruments, acquisition dates and spectral bands used in this work.
- Multi-band Sentinel-2 imagery, 2017: Three tiles of the Level-1C product were considered. Each tile was acquired from the satellite S2A on 11 July 2017, at 9.30 UTC. All data of the COPERNICUS program were downloaded for free from the COPERNICUS HUB website. Bands used in the work had a spatial resolution of 10 m.
- DEM: European DEM provided by EEA (European Environment Agency). The model was a hybrid product based on SRTM and ASTER GDEM data fused by a weighted averaging approach. The spatial resolution was 25 m.
4. Analysis of Multi-Temporal Imagery Dataset
- Color space transformation: The RGB to HSL (Hue, Saturation, Lightness) color transformation was applied to facilitate the detection of the gravel class (see Figure 1). The lightness component of the HSL bands corresponds to the arithmetic mean of the largest and smallest RGB components.
- Pan sharpening: In scenes with LANDSAT 7 or LANDSAT 8 data, i.e., where the panchromatic band is present, pansharpening was performed.
- Thresholding of vegetation and water indices. Plants have a typical absorption/reflection pattern; in fact, chlorophyll strongly absorbs visible light (0.4–0.7 m) for use in photosynthesis, whereas the cell structure of the leaves strongly reflects near-infrared light (from 0.7–1.1 m). The Normalized Difference Vegetation Index (NDVI) is a specific radiometric index that exploits such a typical absorption/reflection pattern, and it is used to map green leaf vegetation. Typically, the NDVI is computed according to the following equation:Likewise NDVI, the Normalized Difference Water Index (NDWI)  is a radiometric index that exploits a typical absorption/reflection pattern for water mapping purposes. After , the NDWI is computed according to the following equation:Thresholds on the NDVI and NDWI maps were set in order to select only pixels corresponding to water. NDVI and NDWI histograms were studied, and pixels with value simultaneously greater than the 90th percentile of the NDWI map and lower than the 10th percentile of the NDVI map were labeled as water pixels. Once the initial water map was obtained, a buffer at a suitable distance was applied to ensure the inclusion of all parts of the wet channel in further processing. Afterwards, the enlarged water map was refined, excluding outside parts of the river reach, such as small pools or secondary channels outside the active channel by means of area thresholding. Likely, the vegetation map was produced selecting pixels with a value greater than the 80th percentile of the NDVI map; the gravel map was produced selecting pixels with a value simultaneously greater than the 95th percentile of the lightness map and lower than the fifth percentile of the saturation map. For every coverage class, the actual percentile values were slightly modified according to the specific imagery dataset considered. Figure 2 and Figure 3 show examples of NDVI and NDWI maps computed from Sentinel-2 imagery.
- Contextual supervised image classification: To distinguish surface coverages of different types in multi-spectral imagery, a contextual classifier based on a Sequential Maximum A Posteriori (SMAP) estimation was used. The classifier models a spectral class by means of simple spectral mean and covariance as parameters of a Gaussian mixture distribution. The SMAP segmentation algorithm attempts to improve segmentation accuracy by segmenting the image into regions rather than segmenting each pixel separately. The SMAP algorithm exploits the fact that nearby pixels in an image are likely to have the same class. It works by segmenting the image at various scales or resolutions and using the coarse-scale segmentations to guide the finer scale segmentations. False-color composite maps facilitated the classification of water, vegetation and gravel classes. In Figure 4, an example of the classified coverage units is shown; the orthophoto is used as a background image, and image contrast is reduced in order to enhance readability.
- Voronoi diagrams: The channel centerlines were extracted starting from the Voronoi diagrams of the areas representing the active and wet channels. The centerline is represented by the polyline connecting the centers of all the highest circles inscribed in the Voronoi polygons. In Figure 5, an example of the detected active channel and centerline is shown; the orthophoto is used as a background image, and image contrast is reduced in order to enhance readability.From the active channel centerline, lastly, transects were created at a fixed distance of 200 m and with adequate lateral extension. From the channels’ centerlines and the transect sets, hydromorphological parameters were derived by means of an automatic procedure described in Section 5.2.
- BW satellite images: Because of the very limited radiometric content and the reference role of BW images, the procedure, up to the extraction of channel centerlines step, was implemented entirely by hand.
- RGB orthophoto and multi-band LANDSAT 1 and 2 imagery: DEM analysis for the extraction of river axis approximation; buffering approximate axis to limit the area of analysis; RGB to HSI color transformation for thresholding based initial gravel detection; buffering of detected gravel areas for the extraction of active channel; supervised contextual classification of active channel surface coverage; buffering and thresholding for coverage classes refinement; active channel identification after boundary vegetation removal.
- Multi-band LANDSAT 3–8 and Sentinel-2 imagery: NDVI and NDWI analysis for the extraction of river axis approximation; buffering approximate axis to limit the area of analysis; RGB to HSI color transformation for thresholding-based initial gravel detection; buffering of detected gravel areas for the extraction of the active channel; supervised contextual classification of active channel surface coverage; buffering and thresholding for coverage classes refinement; active channel identification after boundary vegetation removal.
5.1. Validation of Imagery Analysis Outcomes
- RGB orthophoto: Validation was performed against reference maps obtained by manual digitalization after expert visual inspection of the entire reaches set. Hydromorphological reference parameters were hence computed starting from the reference maps. The production of the reference maps took a couple of days. Table 4 shows the mean and the standard deviation of the differences between the active channel widths obtained from the applied procedure and from the reference maps. Values were, in general, lower than 10 m and had a percentage difference below 5%. Because of radiometric ambiguities between pixels of different coverage classes and the hard classification of mixed pixels, standard deviations presented greater values and variability.
- Multi-band LANDSAT and Sentinel-2 imagery: Regarding the processing of LANDSAT data, it was noted that parts of the scene that did not belong to the river channel were still inserted into the final map, despite carefully-chosen threshold values. This occurred in the areas adjacent to the river where gravel roads or quarries were present, or simply vegetation-free areas that were wrongly classified as gravel. Other times, vegetation-rich river portions were not included in the final map because of the removal of boundary vegetation. To fix these problems, when available data made it possible, a refinement step was set up. The refinement step was based on the application of the standard procedure on two scenes belonging to the same year, but in different seasons. In the summer standard scene, the vegetation recognition threshold was lowered, so that the classification of the other two classes was reduced to the minimum. The water class recognition threshold was modified accordingly to avoid distortion. A map of the active channel was obtained where the vegetation was predominant and the gravel parts were partially filtered. This first new map helped in solving the problem of incorrect classification of quarries and of other gravel disturbances adjacent to the river. Then, for the same river, a late autumn or winter scene of the same year was selected. The standard procedure was applied to produce a map in which, for seasonal issues, vegetation was poor and the gravel class was increased. This map typically improved the contour outlines of the river channel areas where summer vegetation led classification disturbances. This second map helped in solving the problem of incorrect classification of parts of the river that in summer were covered by vegetation and thus excluded from the first map. The maps were therefore intersected, and the result were used to cut the initial map of the procedure. In this way, a final active channel classification included all significant parts of the scene. Validation was performed against hydromorphological reference parameters derived from manual digitalization of RGB color maps. Table 5 shows the results of the validation process for the Vjosë River. Overall, percentage differences between length and width values obtained from the applied procedure and from the reference maps were below 6%. Again, differences were mainly due to radiometric ambiguities between pixels of different coverage classes and the hard classification of mixed pixels. In cases where it was not possible to apply the refinement procedure, maps were refined manually. Comparable results were obtained for the rivers Seman and Shkumbin.
5.2. Characterization of River Morphology
- RGB orthophoto: Table 6 shows coverage classes’ distribution, active channel widths and standard deviations, active channel areas, number of transects, channel centerline lengths, sinuosity indices and fluvial reaches’ classification obtained from the processing of the 2007 aerial orthophotos. From the coverage classes- distribution, two states can be distinguished: (1) the wet area was close to or greater than the gravel area, and the vegetation was very scarce; (2) the wet area was less than the gravel area, and vegetation area was not negligible. Differences in widths (between mean, minimum and maximum values) and standard deviations, when jointly considered, permitted distinguishing reaches with respect to different fluvial typologies: single-thread reaches presented low mean widths and variations with respect to transitional or multi-thread reaches. The sinuosity indices were all greater than 1.05, as expected for natural reaches .
- Multi-band LANDSAT and Sentinel-2 imagery. In Table 7, Table 8 and Table 9, the lengths of each river reach per year of study are reported. Subsequently, in order to visualize the time evolution of the watercourses, the planimetric patterns of some selected centerlines of active and wet river channel for the three rivers are plotted in Figure 6a–f. Alongside the centerline plots in Figure 7a–c and Figure 8a–f, the plots of the temporal variations of the sinuosity index and of the wet and active channel widths are also presented. Observing the charts of the centerlines, it can be seen how some meanders evolved over time, changing their curvature and moving both laterally and longitudinally. In the case of the Seman River, a large meander that reached its peak in 2008 and then underwent a cut-off, is visible; see Figure 9. The progressive growth and consequential cut-off of the meander in the Seman River scene is also visible in the chart of active channel lengths. The highest values of the sinuosity indices appeared in the case of the Vjosë River, which showed peak values of about 2.3 in 2008.In this work, Sentinel-2 data were basically considered for comparison with respect to LANDSAT 8 to support a further application of the presented procedure in a real-time river monitoring system based on a rolling processing mechanism. Satellite-2A was acquired on 11 July 2017, and LANDSAT 8 data were acquired almost at the same time on 1 June 2017. Outcomes of the application of the procedure to the Sentinel-2 data of the Vjosë River are presented in the Table 10. The results of the Sentinel-2 data processing differed very little from those obtained from the LANDSAT 8 data. The average wet channel mean width () was an exception. The value obtained from the Sentinel-2 data was 33% lower than the value derived from LANDSAT 8 data. In fact, the extent of the wet channel can be easily affected by the rainfall. Investigation of such a relevant difference was performed on the basis of rainfall data acquired at a weather station located near Vorë, a coastal city located within the Vjosë River basin. According to the records, significant rainfall events occurred in May 2017, just a few days before the acquisition of the LANDSAT 8 imagery.
- RGB orthophotos and BW CORONA: Another comparison was carried out between the results obtained from the processing of the 2007 RGB orthophotos and those of the 1968 BW images from the CORONA satellite mission. The two datasets have the highest spatial resolutions of the entire image dataset considered in this work, respectively 0.2 and 2.8 m. The comparison was conducted on two reaches of the rivers Seman and Vjosë. Figure 10 and Figure 11 show the compared regions in the 1968 and 2007 images, respectively. Table 11 shows the values of hydromorphological parameters derived from the historical images and those derived from the orthophotos, as already reported in Table 6.The sinuosity indices and the channel widths changed significantly. Mean and maximum widths showed greater differences, whereas smaller differences were found for the minimum width. Mean widths decreased by about 36% and 46% for the Vjosë and Seman reaches, respectively.
6. Discussion and Future Prospects
Conflicts of Interest
- Righini, M.; Surian, N. Remote Sensing as a Tool for Analysing Channel Dynamics and Geomorphic Effects of Floods. In Flood Monitoring through Remote Sensing; Springer: Berlin, Germany, 2018. [Google Scholar]
- Carbonneau, P.; Piégay, H. (Eds.) Fluvial Remote Sensing for Science and Management; Wiley-Blackwell: Hoboken, NJ, USA, 2012. [Google Scholar]
- Demarchi, L.; Bizzi, S.; Piégay, H. Regional hydromorphological characterization with continuous and automated remote sensing analysis based on VHR imagery and low-resolution LiDAR data. Earth Surf. Process. Landf. 2017, 42, 531–551. [Google Scholar] [CrossRef][Green Version]
- Sundermann, A.; Stoll, S.; Haase, P. River restoration success depends on the species pool of the immediate surroundings. Ecol. Appl. 2011, 21, 1962–1971. [Google Scholar] [CrossRef] [PubMed]
- Poppe, M.; Kail, J.; Aroviita, J.; Stelmaszczyk, M.; Gielczewski, M.; Muhar, S. Assessing restoration effects on hydromorphology in European mid-sized rivers by key hydromorphological parameters. Hydrobiologia 2016, 769, 21–40. [Google Scholar] [CrossRef]
- Piégay, H.; Mathias, K.G.; Sear, D.A. Integrating geomorphological tools to address practical problems in river management and restoration. In Tools in Fluvial Geomorphology; Wiley-Blackwell: Hoboken, NJ, USA, 2016. [Google Scholar]
- Marteau, B.; Vericat, D.; Gibbins, C.; Batalla, R.J.; Green, D.R. Application of Structure from Motion photogrammetry to river restoration. Earth Surf. Process. Landf. 2017, 42, 503–515. [Google Scholar] [CrossRef]
- Parasiewicz, P. Using MesoHABSIM to develop reference habitat template and ecological management scenarios. River Res. Appl. 2007, 23, 924–932. [Google Scholar] [CrossRef]
- Conallin, J.; Boegh, E.; Jensen, K.J. Instream physical habitat modelling types: An analysis as stream hydromorphological modelling tools for EU water resource managers. Int. J. River Basin Manag. 2010, 8, 93–107. [Google Scholar] [CrossRef]
- Bergeron, N.; Carbonneau, P.E. Geosalar: Innovative Remote Sensing Methods for Spatially Continuous Mapping of Fluvial Habitat at Riverscape Scale. In Fluvial Remote Sensing for Science and Management; John Wiley & Sons: Hoboken, NJ, USA, 2012; Chapter 9; pp. 193–213. [Google Scholar]
- Demarchi, L.; Bizzi, S.; Piégay, H. Hierarchical object-based mapping of riverscape units and in-stream mesohabitats using LiDAR and VHR Imagery. Remote Sens. 2016, 8, 97. [Google Scholar] [CrossRef][Green Version]
- Belletti, B.; Rinaldi, M.; Bussettini, M.; Comiti, F.; Gurnell, A.M.; Mao, L.; Nardi, L.; Vezza, P. Characterising physical habitats and fluvial hydromorphology: A new system for the survey and classification of river geomorphic units. Geomorphology 2017, 283, 143–157. [Google Scholar] [CrossRef]
- Poole, G.C. Stream hydrogeomorphology as a physical science basis for advances in stream ecology. J. N. Am. Benthol. Soc. 2010, 29, 12–25. [Google Scholar] [CrossRef]
- Elosegi, A.; Díez, J.; Mutz, M. Effects of hydromorphological integrity on biodiversity and functioning of river ecosystems. Hydrobiologia 2010, 657, 199–215. [Google Scholar] [CrossRef]
- Legleiter, C.J.; Marcus, W.A.; Lawrence, R.L. Effects of sensor resolution on mapping instream habitats. Photogramm. Eng. Remote Sens. 2002, 68, 801–807. [Google Scholar]
- Neal, J.C.; Bates, P.D.; Fewtrell, T.J.; Hunter, N.M.; Wilson, M.D.; Horritt, M.S. Distributed whole city water level measurements from the Carlisle 2005 urban flood event and comparison with hydraulic model simulations. J. Hydrol. 2009, 368, 42–55. [Google Scholar] [CrossRef]
- Bizzi, S.; Demarchi, L.; Grabowski, R.C.; Weissteiner, C.J.; Van de Bund, W. The use of remote sensing to characterise hydromorphological properties of European rivers. Aquat. Sci. 2016, 78, 57–70. [Google Scholar] [CrossRef][Green Version]
- Rinaldi, M.; Belletti, B.; Bussettini, M.; Comiti, F.; Golfieri, B.; Lastoria, B.; Marchese, E.; Nardi, L.; Surian, N. New tools for the hydromorphological assessment and monitoring of European streams. J. Environ. Manag. 2017, 202, 363–378. [Google Scholar] [CrossRef] [PubMed]
- Mitidieri, F.; Papa, M.N.; Amitrano, D.; Ruello, G. River morphology monitoring using multitemporal SAR data: Preliminary results. Eur. J. Remote Sens. 2016, 49, 889–898. [Google Scholar] [CrossRef]
- Vijith, H.; Dodge-Wan, D. Morphology and channel characteristics of an equatorial tropical river in Malaysian Borneo: A detailed evaluation through spatially explicit geomorphometric modelling. Model. Earth Syst. Environ. 2018, 4, 325–337. [Google Scholar] [CrossRef]
- Henshaw, A.J.; Gurnell, A.M.; Bertoldi, W.; Drake, N.A. An assessment of the degree to which landsat TM data can support the assessment of fluvial dynamics, as revealed by changes in vegetation extent and channel position, along a large river. Geomorphology 2013, 202, 74–85. [Google Scholar] [CrossRef]
- Belletti, B.; Dufour, S.; Piégay, H. Regional variability of aquatic pattern in braided reaches (example of the French Rhône basin). Hydrobiologia 2013, 712, 25–41. [Google Scholar] [CrossRef]
- Clerici, A.; Perego, S. A set of GRASS GIS-based Shell scripts for the calculation and graphical display of the main morphometric parameters of a river channel. Int. J. Geosci. 2016, 7, 135–143. [Google Scholar] [CrossRef]
- Cencetti, C.; De Rosa, P.; Fredduzzi, A. Geoinformatics in morphological study of River Paglia, Tiber River basin, Central Italy. Environ. Earth Sci. 2017, 76, 128. [Google Scholar] [CrossRef]
- Gilvear, D.; Bryant, R. Analysis of remotely sensed data for fluvial geomorphology and river science. In Tools in Fluvial Geomorphology; Wiley-Blackwell: Hoboken, NJ, USA, 2016; pp. 103–132. [Google Scholar]
- Geerling, G.W.; Vreeken–Buijs, M.J.; Jesse, P.; Ragas, A.M.J.; Smits, A.J.M. Mapping river floodplain ecotopes by segmentation of spectral (CASI) and structural (LiDAR) remote sensing data. River Res. Appl. 2009, 25, 795–813. [Google Scholar] [CrossRef]
- Wiederkehr, E.; Dufour, S.; Piégay, H. Localisation et caractérisation des géomorphosites fluviaux à l’échelle des réseaux hydrographiques, exemples d’applications géomatiques dans le bassin de la Drôme. Géomorphol. Relief Processus Environ. 2010, 2, 175–188. [Google Scholar] [CrossRef][Green Version]
- De Leo, F.; Besio, G.; Zolezzi, G.; Bezzi, M.; Floqi, T.; Lami, I. Coastal erosion triggered by political and socio-economical abrupt changes: The case of Lalzit Bay, Albania. Coast. Eng. Proc. 2017, 1, 13. [Google Scholar] [CrossRef]
- Cavazza, S. Sulla scabrezza di alcuni corsi d’acqua albanesi in rapporto al carico di sedimenti trasportati in sospensione. In Proceedings of the XI Convegno di Idraulca e Costruzioni Idrauliche, Genova, Italy, 27–28 October 1968. [Google Scholar]
- Ciavola, P. Relation between river dynamics and coastal changes in Albania: An assessment integrating satellite imagery with historical data. Int. J. Remote Sens. 1999, 20, 561–584. [Google Scholar] [CrossRef]
- Pano, N. Dinamica del littorale albanese (sintesi delle conoscenze). In Proceedings of the 19th Atti del XIX Convegno AIGI Meeting, Genova, Italy, 4–6 Novembre 1992; pp. 3–18. [Google Scholar]
- Simeoni, U.; Pano, N.; Ciavola, P. The coastline of Albania: Morphology, evolution and coastal management issues. Bull. Inst. Océanogr. 1997, CIESM 3, 151–168. [Google Scholar]
- Milliman, J.D.; Syvitski, J.P.M. Geomorphic/Tectonic Control of Sediment Discharge to the Ocean: The Importance of Small Mountainous Rivers. J. Geol. 1992, 100, 525–544. [Google Scholar] [CrossRef]
- Constantine, J.A.; Dunne, T.; Ahmed, J.; Legleiter, C.; Lazarus, E.D. Sediment supply as a driver of river meandering and floodplain evolution in the Amazon Basin. Nat. Geosci. 2014, 7, 899–903. [Google Scholar] [CrossRef][Green Version]
- Goossens, R.; De Wulf, A.; Bourgeois, J.; Gheyle, W.; Willems, T. Satellite imagery and archaeology: The example of CORONA in the Altai Mountains. J. Archaeol. Sci. 2006, 33, 745–755. [Google Scholar] [CrossRef]
- Schenk, T.; Csatho, B.; Shin, S.W. Rigorous panoramic camera model for disp imagery. In Proceedings of the ISPRS Workshop: High Resolution Mapping from Space, Hannover, Germany, 6–8 October 2003. [Google Scholar]
- Sohn, H.G.; Kim, G.; Yom, J. Mathematical modelling of historical reconnaissance CORONA KH-4B Imagery. Photogramm. Rec. 2008, 19, 51–66. [Google Scholar] [CrossRef]
- Hamandawana, H.; Eckardt, F.; Ringrose, S. Proposed methodology for georeferencing and mosaicking CORONA photographs. Int. J. Remote Sens. 2007, 28, 5–22. [Google Scholar] [CrossRef]
- Anyamba, A.; Tucker, C. NDVI anomaly patterns over Africa during the 1997/98 ENSO warm event. Int. J. Remote Sens. 2011, 22, 1847–1859. [Google Scholar]
- Gao, B. NDWI-A normalized difference water index for remote sensing of vegetation liquid water from space. Remote Sens. Environ. 1996, 58, 257–266. [Google Scholar] [CrossRef]
- McFeeters, S.K. The use of the Normalized Difference Water Index (NDWI) in the delineation of open water features. Int. J. Remote Sens. 1996, 17, 1425–1432. [Google Scholar] [CrossRef]
- Surian, N.; Rinaldi, M.; Pellegrini, L. Linee Guida per L’analisi Geomorfologica Degli Alvei Fluviali e Delle Loro Tendenze Evolutive; Cleup: Padova, Italy, 2009. [Google Scholar]
- Surian, N.; Rinaldi, M. Dinamica recente ed attuale degli alvei fluviali in italia: Stato dell’arte e prospettive. Il Quaternario 2008, 21, 233–240. [Google Scholar]
- Gurnell, A.M.; Bertoldi, W.; Tockner, K.; Wharton, G.; Zolezzi, G. How large is a river? Conceptualizing river landscape signatures and envelopes in four dimensions. WIREs Water 2016, 3, 313–325. [Google Scholar] [CrossRef]
- Fasola, M.; Bogliani, G. Habitat selection and distribution of nesting Common and Little Terns on the Po River (Italy). Colon. Waterbirds 1984, 7, 127–133. [Google Scholar] [CrossRef]
- Parasiewicz, P.; Rogers, J.N.; Vezza, P.; Gortázar, J.; Seager, T.; Pegg, M.; Wisniewolski, W.; Comoglio, C. Ecohydraulics: An Integrated Approach; Maddock, I., Harby, A., Kemp, P., Wood, P., Eds.; John Wiley & Sons: Hoboken, NJ, USA, 2013; pp. 109–124. [Google Scholar]
- Elosegi, A.; Sabater, S. Effects of hydromorphological impacts on river ecosystem functioning: A review and suggestions for assessing ecological impacts. Hydrobiologia 2012, 712, 129–143. [Google Scholar] [CrossRef]
- Surian, N.; Rinaldi, M. Morphological response to river engineering and management in alluvial channels in Italy. Geomorphology 2003, 50, 307–326. [Google Scholar] [CrossRef][Green Version]
- Comiti, F. How natural are Alpine mountain rivers? Evidence from the Italian Alps. Earth Surf. Process. Landf. 2012, 37, 693–707. [Google Scholar] [CrossRef]
- Niroumand-Jadidi, M.; Vitti, A. Reconstruction Of River Boundaries At Sub-Pixel Resolution: Estimation And Spatial Allocation Of Water Fractions. ISPRS Int. J. GeoInf. 2017, 6, 383. [Google Scholar] [CrossRef]
- Zarfl, C.; Lumsdon, A.E.; Berlekamp, J.; Tydecks, L.; Tockner, K. A global boom in hydropower dam construction. Aquat. Sci. 2014, 77, 161–170. [Google Scholar] [CrossRef]
|L-1||MSS||B4, B5, B6, B7|
|L-2||MSS||B4, B5, B6, B7|
|L-2||MSS||B4, B5, B6, B7|
|L-5||TM+||B1, B2, B3, B4|
|L-4||TM||B1, B2, B3, B4|
|L-7||ETM+||B1, B2, B3, B4, B8|
|L-7||ETM+||B1, B2, B3, B4, B8|
|L-7||ETM+||B1, B2, B3, B4, B8|
|L-8||OLI||B2, B3, B4, B5, B8|
|L-8||OLI||B2, B3, B4, B5, B8|
|River||# of GCPs||Max RMSE (m)||Mean RMSE (m)|
|River||Reach||Mean (m)||Std. Dev. (m)|
|Parameters||Applied Procedure (AP)||Manual Digitalization (HD)||(HD−AP)||%|
|River||Reach||Classes (%)||Area (km2)||Widths (m)||# Transects||Lengths (m)||Sinuosity Index||Type|
|River||Year||Widths (m)||Lengths (m)||Sinuosity Index|
© 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).