“Assessment of the geomorphic effectiveness of controlled floods in a braided river using a reduced-complexity numerical model” Ziliani et al

Most Alpine rivers have undergone strong alteration ofsignificant alterations in flow and sediment regimes. These alterations have notable effects on river morphology and ecology. One option to mitigate such effects is the flow regime management, specifically bythrough the re-introductionreintroduction of channel-forming discharges. The aim of this work is to assess the morphological changes induced in the Piave River (Italy) due toby two differentdistinct controlled flood strategies, the first characterized by a single artificial flood per year and the second by higher magnitude, but less frequent, floods. TheThis 190 work was carried outinvolved applying a 2D reduced-complexity morphodynamic model (CAESAR-LISFLOOD) to a 7 -kmlong reach, characterized by a braided pattern and highly regulated discharges. The numericalNumerical modelling allowed the assessment of morphological changes for four long-term scenarios (2009-2034). The scenarios were defined taking into accountconsidering the current flow regime and the natural regime, which was estimated by a stochastic physically-based hydrologic model. Changes in channel morphology were assessed by measuring active channel width and braiding intensity. 195 ComparingA comparison of controlled flood scenarios to a baseline scenario (i.e., no controlled floods) it turned outshowed that artificial floods had small effectslittle effect on channel morphology. The highestMore channel widening (13.5%) was produced byresulted from the release strategy with higherhigh magnitude floods, whileflood strategy than from the application of the other strategies produced lower wideningstrategy (8.6%). Negligible change was observed in terms of braiding intensity. Results pointed outThe results indicate that controlled floods maydo not represent an effective solution for morphological 200 recovery in braided rivers with strongly impacted in their flow and sediment regimes.

(4) The test for sediment transport rate is quite weak: the authors find that the computed transport rates are within typical range for such gravel-bed braided rivers. A more 100 sensitive test would be to evaluate how the bed grain size changes over time. You specify an initial grain size -does that grain size shift dramatically over the course of the model run?
Thanks for pointing out this. We analyzed grain size changes, specifically we 105 compare D50 of bed sediment at the beginning (i.e. 24.9 mm) and at the end of each

110
(5) River are a combination of sediment-feed and sediment-recirculating systems. I suspect that the model results are sensitive to this choice. The problem, of course, is specifying an upstream sediment boundary condition. I would be interested in learning how an increase in sediment supply changes the predicted channel morphology. Perhaps that is affect the morphology. Ultimately, it shows how controlled releases of larger flows have some but not significant impacts on channel widths and depths within the reach studied.
In doing so it also provides an excellent opportunity to evaluate/test a morphodynamic model (which is a complex and far from straightforward exercise) demonstrating how such methods can be used as management tools in such environment to answer questions about hydro-140 geomorphic interactions.
The paper is well written, produced and structured. As per R1, there are several minor grammatical/typo mistakes that can be picked up in proof reading if the paper progresses.
Some specific comments and suggestions for further literature that has not been cited (some has only just come out) are provided below.  230-236 -felt a bit clunky and repetetive -might be worth having a closer look at this 155 section. Also there have been a series of CL papers and studies since 2013 that might be useful for the paper to cite here as well. There are others but two more recent ones CL Sensitivity analysis paper: https://www.  Surian, 2012;Magilligan et al., 2013;Mueller et al., 2014;Lobera et al., 2016). Nowadays, damDam construction is now considered a viable strategy to support the increasingmeet energy and water demands due to climate change and population growth (World Bank, 2009;Lehner et al., 2011). As outlined by Overeem et al. (2013), large reservoirs with a volumevolumes 210 greater than 0.5 km 3 intercept globally more than 40% of river discharge and ~26% of the sediments transported by the rivers, reducing the global sediment delivery to oceans, and commonly leading to coastal erosion.
Environmental flow management plans aim to mitigate someare aimed at mitigating undesired channel adjustments due to 240 dam operations. Due toGiven the costcosts of these programs, decision-makers are increasingly requesting the scientific community to develop appropriate tools able tocapable of (i) identifyidentifying and controlcontrolling the factors that cause channel alterations, and (ii) to assessof assessing the effectiveness of management programs. Environmental agencies in several countries require dam operations to respect releasing protocol in an attemptrelease protocols to mitigate adverse impacts on downstream ecosystems (Schmidt and Wilcock, 2008;Olden and Naiman, 2010;Watts et al., 2011;Konrad et al., 245 2012). Beisdes someAlthough successful empirical experiences do exist Konrad et al., 2011), robust predictive tools and models are becoming much more impellenturgently needed to predict channel responseresponses to dam operations and interruption of the longitudinal river continuum (Bliesner et al., 2009;McDonald et al., 2010;Melis, 2011;Coulthard and Van De Wiel, 2013;Gaeuman, 2014).
The assessment of future evolutionary trajectory of channel morphology maycan be achievedassessed using conceptual models 250 (e.g., Channel Evolution Models -CEMs, as described in Schumm et al., 1984;Simon and Hupp, 1986;Simon, 1989), empirical (Lane, 1955;Schumm, 1977;Rhoads, 1992) orand numerical models, either Computational Fluid Dynamic (CFD) models or Reduced Complexity Models (RCM).) (Larsen et al., 2014). Previous applications of RCMs on braided rivers have focused mainly on theoretical scale-independent analysis (Murray and Paola, 1994), laboratory experiments (Doeschl-Wilson and Ashmore, 2005;Doeschl et al., 2006;Nicholas, 2010), or short gravel-bed river reaches (Coulthard et al., 2002;Thomas 255 and Nicholas, 2002;Coulthard et al., 2007;Thomas et al., 2007;Van De Wiel et al., 2007). In this study, such as in Ziliani et al. (2013) and Ziliani and Surian (2016), an attempt has been made towe apply aan RCM model at mesospatial (i.e., 5-50 km) and mesotemporal (i.e., 10-100 years) scales. In particular, the CAESAR -LISFLOOD model (Bates et al., 2010;Coulthard et al., 2013) has been hereinis applied to a 7 -km -long braided reach of the Piave River (Italy), one of the most heavily and historically regulated river systemsystems in Italy. 260 We applied the CAESAR -LISFLOOD model (hereafter C-L) to assess the morphological effects related toof two different kinds of flow regime management strategies: the. The first is characterized by yearly controlled floods with peaks able to transportcapable of transporting sediments;, while the second with more infrequent andconsists of less frequent higher magnitude floods (i.e., floods with 5-year recurrence interval equal to 5 yearsintervals) released only when notable channel narrowing is observed in the evolutionary trajectory. Both strategies have been developed according to two main criteria: (i) 265 to maximize the flow regime "re-naturalization",, meaning that the "Controlled Flood" (CF) duration has to be set in orderof the controlled flood (CF) is designed to increase its yearly likelihood to occur approaching theapproach natural scenario conditionconditions as much as possible; and (ii) the "feasibility" of the strategy, which is verified by the fact that the cumulative volume released per year has to be loweris less than the maximum volume stocked in the reservoirs existing upstream offrom the studystudied reach. 270 This paper aims to addressaddresses two main issues: (i) to what extent the effectiveness of controlled floods can be effective for the geomorphic recovery of a strongly regulated braided river?, and (ii) canthe suitability and reliability of the reducedcomplexity CAESAR-LISFLOOD morphodynamic model CAESAR-LISFLOOD be considered a suitable and reliable tool to reproduce the morphological evolution of a large gravel bed riverrivers at the given mesoscales?. In the first section of the paper, we provide a brief description of the studied river reach. The second section presents the available data, the two models 275 used (i.e.,, the morphodynamic model CAESAR-LISFLOOD, model (Bates et al., 2010;Coulthard et al., 2013), and the hydrological model, (Botter et al., 2007)), and the criteria adopted forto design the scenario strategy design. The third section presents the results concerningin terms of (i) the historical river reach morphological river reach adjustments, (ii) the flow regime alterationalterations, (iii) the CAESAR-LISFLOOD calibration, and (iv) the simulations of three different "Controlled Floods" (CFs) releasescontrolled flood release scenarios. Finally, we critically discuss the results and examine the strengths 280 and weaknesses of CAESAR-LISFLOOD and the effectiveness of the flow management strategies under investigation.
2 General setting of the study area

The Piave River basin
The Piave River is located in north-eastern ofnortheastern Italy, and it. It flows for about 220 km from the Alps to the Adriatic Sea ( Fig. 1). The basin area is about 3,900 km 2 , and its average elevation is about 1,300 m a.s.l. (maximum elevation is 3,364 285 m a.s.l.). The climate is temperate-humid, with an average annual precipitation of aboutapproximately 1,350 mm. Significant annual variations in theAnnual rainfall amount have been measuredvaried substantially over the 20 th century, but without any statistically relevantsignificant trends (Surian, 1999).
AsLike most of theItalian Alpine Italian rivers (Surian and Rinaldi, 2003;Comiti, 2012), the Piave River has suffered heavy human impact, which has altered the basin and the river channel dynamics (Surian, 1999;Botter et al., 290 2010;Comiti et al., 2011;Comiti, 2012). Especially during the 20 th century, theThe Piave basin has experienced a rapid increase ofin anthropogenic exploitation byin the 20 th century, with the construction of a series of dams and reservoirs (nowadays therebetween the 1930s and 60s. There are now 13 major reservoirs, (Botter et al., 2010) built along the main stem and some tributaries from the 1930s to 1960s. At present, a (Botter et al., 2010). A complex regulation scheme existsis in place (for details see Surian, 1999;Botter et al., 2010), designed) to maximize production of hydroelectric power production and the 295 provision ofprovide irrigation water (Fig. 1). Flow regulation alters both the flow duration characteristics and the volume of annual runoff in the river. The reservoirs and diversions along the river and its tributaries also affect sediment transport and supply.
The Piave basin had also experienced has historically experienced strong changes due to land use modifications. Especially afterSince the 1950s, the development of industry and tourism boosted the abandonment of traditional agricultural and 300 cropping activitiesproduction on the mountain slopes, causing have been abandoned largely because of the development of industry and tourism, resulting in natural reforestation in the upper parts of the basin (Del Favero and Lasen, 1993). In addition to the reductions in sediment supply due to trapping by dams and reforestation, intense in-channel gravel mining has also contributed to alterreducing sediment fluxes since the 1960s. Furthermore, humanHuman pressure on the river channel dynamics has also resulted from construction of bank protection structures and torrent control works. As, as a result of these 305 bank protection works, at presentwhich the river can still move laterally, although the available width for planform shifting is narrower than its natural braided belt.

Study reach
The studystudied reach is ~7 km -long ( Fig. 1) and is located between Ponte nelle Alpi and Belluno (the drainage area at Belluno is 1,826 km 2 ). In this reach the The morphology of the reach is mainly braided and wandering. The average slope of 310 the reach is 0.47%, and the median surface grain size ranges between 18 and 32 mm (Tomasi, 2009). The active channel width ranges between 43 and 452 m, beingand is 241 m on average, while the fluvial corridor width, defined by the presence of Holocene fluvial terraces, ranges between 106 and 1,110 m, beingand is 672 m on average. Previous studies (Surian, 1999;Comiti et al., 2011;Picco et al., 2016) have outlined that, over the last 200 years, the studystudied reach havehas undergone notable lateral adjustments (narrowing up to 66%), but notwith no significant changes ofin channel pattern. 315 In terms of bed-level changes, twopatterns. Two phases have been identified: a in terms of changes in bed levels, first a phase of moderate incision (1970-1990s) followed by a more recent phase (1990s-2003/2007) during which the river has exhibited equilibrium or slight aggradation Comiti et al., 2011).

Channel morphology and reconstruction of its evolutionary trajectory 320
Channel morphology was analysed in order to gather (i) input data for the CAESAR -LISFLOOD model, (ii) data for model calibration, and (iii) evidence of the evolutionary trajectory of the studystudied reach. RiverThe river channel, islands, flowing channelchannels, bank protection structures and groynes were digitized using the available aerial photos and terrain models covering the studystudied reach (i.e., 2003, 2009 - Table 1). The analysis was carried out usingemployed ArcGIS 10.2. The flowing channels and the unvegetated or sparsely vegetated bars with little or no vegetation were merged to obtain 325 measurement ofmeasure channel width. BraidingThe braiding index was calculated using the average number of anabranches across the river (Ashmore, 1991;Egozi and Ashmore, 2008). The historical analysis carried out byof Comiti et al. (2011), which covered the period 1805 --2006 has been, was extended up to 2009. A LiDAR Digital Elevation Model (DEM) was provided by the Autorità di Bacino delle Alpi Orientali. It was created using an airborne LiDAR survey that was acquired in July 2003 (orthometric elevations adopted, vertical error estimate ±20 cm) 330 almost contemporary to one of thean aerial photosphoto used in this study (Table 1). Even though the river reach is characterized at low flow by the presence of rather small inundated areas, it was not possible to obtain bed elevation in the flowing channel areas with the standard LiDAR data. Therefore, to complete theTo obtain bed elevation extraction, the water depth was estimated through the application ofby applying the method proposed by Bertoldi et al. (2011) using the 2003 aerial photos. This is an optical remote sensing technique (Marcus, 2012) for retrieving shallow water depth information using the 335 color of the pixel, as. Legleiter et al. (2009) demonstrated that the log transformation of the green over red band ratio correlates linearly with water depth across a wide range of substrate types. The linear regression is usually should be calibrated by direct measurementsmeasurement of water depths at the time of the aerial survey. Since such data were not available, we calibrated the regression coefficients by referring to the topography of boththe 2003 and 2009 cross -section surveys.

Analysis of the hydrologic regime inof the Piave River
A variety ofThere are several approaches is available to analyse the impact of river regulation on the natural flow regime of rivers (Richter et al., 1996;Richter et al., 1997;Martínez Santa-María et al., 2008;Yin et al., 2015). In the case of the Piave 345 River, although several studies have investigated the degree of alteration to its hydrological regime alteration (Villi and Bacchi, 2001;Botter et al., 2010;Comiti et al., 2011), such analysis washas been hampered by (i) the unavailabilitylack of a longterm flow data series, and (ii) the difficulty in sortingdistinguishing between natural and artificial componentcomponents of the flow regime. In Da Canal et al. (2007) and Comiti et al. (2011), flow records derived from two gauging stations (Busche and Segusino; Fig. 1),) were modified using a specific corrective factor (Villi and Bacchi, 2001)  Furthermore, their analysis showed that the dischargedischarges with Recurrence Intervala recurrence interval (RI) of 2 years waswere not statistically different if calculated separately for pre-and post-regulation periods (1954 was used as the separation date between the periods). However, higher frequencythe peak discharges of the more frequent events (RI ≤ 1.5 year) show a reduction of peak dischargedecreased after 1954. Similar outcomes have also been reported in Picco et al. (2016) have applied an analytical stochastic model (Botter et al., 2007) to characterize the streamflow probability density function 360 (pdf) by means of climate, soil and vegetation parameters. After applying a preliminary model application to smaller, unregulated sub-catchments (that allowed tofor properly verifyverifying the capability of the model to reproduce locally the natural streamflow regime) locally), the authors haveauthor applied the model also into several regulated sections of the Piave River, including Soverzene (about 5 km upstream offrom Ponte nelle Alpi, Fig. 1), in order to evaluate the natural flow regime in regulated cross -sections and, bybased on the difference, the effect of regulation on the statistical hydrographic features of 365 the hydrograph. The approach conceptualizes the dynamics of daily streamflow as a sequence of peaks in response to rainfall and decays in between these jumps. These jump-decay dynamics are then linked to a catchment-scale soil-water balance wherein which the input is represented by stochastic daily rainfall. In this setting, flow-producing rainfall events (that lead to streamflow jumps) result from the censoring operated by catchment soils on daily rainfall, and theywhich are modeledmodelled as a marked Poisson process with mean depth α and mean frequency λ. The parameter α identifies the average intensity of daily rainfall events, while λ is the frequency of flow-producing events, which is smallerless than the underlying precipitation frequency because of the effect of soil moisture dynamics and evapotranspiration. As a consequence, several climate variables (such as rainfall attributes), as well as soil and vegetation properties were, are embedded in λ. Additionally, in thatIn this framework, streamflow recessions in between flow pulses are assumed asto be exponential, with a mean rate equal to k, which defines the inverse of the time scale of the hydrological response (i.e., the mean water retention time in the upstream 375 catchment). Under these assumptions, it can be shown that the steady-state pdf of the specific daily discharge (discharge per unit catchment area) is a Gamma distribution with shape parameter λ/k and scale parameter αk. The model is applied at the seasonal timescale, and then the annual pdf is calculated as the average of the four seasonal distributions. During winter, theThe presence of snow dynamics in winter in the uppermost regions of the catchment is accounted for by reducing the size of the active contributing catchment and increasing the recession rates as described by Schaefli et al. (2013), with an elevation 380 threshold of about 1,900 m a.s.l. In spring, a base flow value is added to the modeledmodelled streamflow distribution, which corresponds to a rigid rightward shift of the pdf. The probability distribution of the natural daily streamflows estimated by the model is then compared to the pdf of the observed daily flows to assess the extent of the impact of regulation in the lower reaches of the Piave River, and to get someobtain guidelines for devising meaningful strategies of the flow regime management strategies. In particular, the daily streamflow series used in this study has beenwas recorded from 1995 to 2009 at Belluno 385 gauging station located at the downstream section of the studystudied reach (Fig. 1). The highest flood event peaks observed in the reference periods (1996, 2000 and 2002) were checked and modified combining data atfrom Belluno and discharge measurements at the Soverzene weir (Braidot, 2003).

The CAESAR-LISFLOOD model
Over the last 20 years theThe application of hydro-morphodynamic physically-based numerical models (generally known as 390 Computational Fluid Dynamic models, CFD) over the last 20 years has mainly been focused on laboratory idealized channel configurations at the laboratory scale (Wu et al., 2000;Defina, 2003;Rüther and Olsen, 2005;Abad et al., 2008) or referred to the morphological dynamicdynamics of natural channels over short time periods (Darby et al., 2002;Chen and Duan, 2008;Li et al., 2008;Wang et al., 2008;Zhou et al., 2009). Although their recent development, the restriction of their fieldThe limited range of application of these models reflects unsolvedunresolved issues in terms of data availability and high computational 395 demands (Escauriaza et al., 2017). Only a few recent works (i.e., Nicholas, 2013a;Williams et al., 2016) have shown that CFD models can be applied at larger spatial and temporal contextsscales. This limitation has drivenled to develop the development of alternative two-dimensional alternative models that have beenare commonly referred to as cellular automata (Murray, 2007), cellular models (Murray and Paola, 1994;Coulthard et al., 2002;Thomas and Nicholas, 2002;Coulthard et al., 2007;Parsons and Fonstad, 2007;Van De Wiel et al., 2007), exploratory models, andor reduced-complexity models (RCM - Murray, 2007;400 Nicholas et al., 2006;. These models have a common solution to the problem that, which is the adoption of simplified hydrodynamic and sediment transport equations derived by the abstractions of the governing physics. A major advantage of RCMs is their computational efficiency that, which allows to simulatefor simulating river evolution over historic and Holocene timescales (e.g., Coulthard et al., 2002;Coulthard et al., 2005;Nicholas and Quine, 2007;Thomas et al., 2007;Van De Wiel et al., 2007). However, the physical realism of suchthese models has received relatively little attention 405 (Nicholas, 2009;Nicholas, 2013b;Ziliani et al., 2013), and there have been only a few studies have shown that deal with the high sensitivity of these models are highly sensitive to the grid resolution of the computational domain (Doeschl-Wilson and Ashmore, 2005;Doeschl et al., 2006;Nicholas and Quine, 2007). Despite the progress shownmade in several works (Nicholas, 2009;Nicholas, 2013a), there are still few applications into natural rivers characterized by complex channel morphology (Ziliani et al., 2013;Ziliani and Surian, 2016). 410 In this study, we applied the CAESAR-LISFLOOD model  has been applied (1.2 version, reach mode; see "the Supplementary material"Material file for details about the model)). This model), which is an integration of the LISFLOOD-FP (Bates et al., 2010) and CAESAR Van De Wiel et al., 2007) models. The C-L model links the hydraulics of the former with the erosion and deposition components of the latter. LISFLOOD-FP is a onedimensional inertial model derived from the full shallow water equations that is applied in the x and y directions to simulate 415 two-dimensional flow over a rasterized spatial domain (Bates et al., 2010). LISFLOOD-FP has been successfully tested to simulate hydraulics in shallow water environments affected by a strongly unidirectional flow (Bates et al., 2010;Neal et al., 2011;Coulthard and Van De Wiel, 2013;Lewis et al., 2013;Skinner et al., 2015;Wong et al., 2015) and for flood inundation simulations characterized by rapid wetting and drying condition (Bates et al., 2010). The CAESAR model Van De Wiel et al., 2007;Ziliani et al., 2013) represents the morphodynamic component of the C-L integrated C-L 420 model. Ziliani et al. (2013) submitted CAESAR to a rigorous and objective performance evaluation procedure, and showedfound that (i) CAESAR can be a very powerful tool for modelling spatial and temporal scales still hardlythat are not well supported by 2D -3D CFD morphodynamic CFD models, (ii) it can be very useful for setting "what-if scenario" strategies over meso -spatial and -temporal scales, and (iii) it provides reliable bedload sediment budget estimations. From a morphological point of view, Ziliani et al. (2013) have shownshowed that CAESAR is able tocan reproduce the average change 425 in channel width, but it performedperforms poorly in reproducing the braided in-channel pattern dynamicdynamics and the typical topographic complexity of a braided river at low water stages (e.g., braiding intensity). LISFLOOD-FP and CAESAR have been efficiently integrated and tested intoin the new CAESAR-LISFLOOD (see Coulthard et al., 2013 for details). Herein the hydraulic element embedded into the model has been verified to be consistent with the LISFLOOD-FP developed by Bates (2010), but the geomorphic component of the The C-L model has not been fully evaluated referring to real case study data. 430 The embedded erosionbeen applied at reach (Feeney et al., 2020) and deposition modules have beencatchment scales (Coulthard and Van De Wiel, 2017), whereas Skinner et al. (2018) assessed only through the intercomparison of CAESAR and C-L sediment yield resultsthe sensitivity of the model. Feeney et al. (2020) used the model to reconstruct geomorphic changes for ten alluvial reaches in northern England, and found that the model accurately reproduces channel and floodplain dynamics at meso-temporal scales. 435

Morphodynamic model performance assessment
Several works (Darby and Van De Wiel, 2003;Hoey et al., 2003;Wilcock and Iverson, 2003) have emphasized the challenge of a proper calibration of process-based models in fluvial geomorphology due to the increase ofincreasing uncertainty proportionally withproportional to the increased complexity of the modeled processesmodelled process and the number of parameters to be estimated (Formann et al., 2007;Papanicolaou et al., 2008). Despite this intrinsic complexity, it is crucial to 440 understand the limitations and performance of RCMs (Aronica et al., 2002;Hall et al., 2005;Lane, 2006), adopting methods that (i) are able tocan include all the limitations inherent in calibration ofcalibrating this type of model, and (ii) are mainly based on field and remote sensing data (Nicholas, 2010). There are currently no standard international standard methods for the calibration and validation of fluvial morphodynamic models (Mosselman, 2012), and those previously proposed are typically designed for hydrodynamicshydrodynamic CFD models (ASME, 1993;Lane et al., 2005). Furthermore, the 445 calibration of RCMs has toIt must also be performed keepingkept in mind that a calibrated RCM model can just be "simply empirically adequate" (Van Fraassen, 1980) and its validation is justsimply a "confirmation" (Oreskes et al., 1994) that cannot be considered conclusive (Haff, 1996;Lane et al., 2005;Murray, 2007). Calibration of the C-L model was specifically addressed by Feeney et al. (2020), who pointed out the need for reach-specific calibration to increase model performance.
In light of all the issues above, the C-L was calibrated referring to the July 5 th 2003 -August 5 th 2009 period (Fig. 2) by 450 comparing the model output (i.e., morphological features such as channel boundaries, islands, wet channel positions) to the channel morphology digitized using the 2009 aerial photos (see "Supplementary material" for a detailed description of model calibration). The hourly discharge series was used as upstream flow boundary condition. At the downstream end of the reach, a constant energy slope was fixed at 0.0047 m m -1 , equal to the local bed slope. The initial bed sediment grain size was set according to Tomasi (2009) results. The grain size distribution was defined using nine classes and was considered 455 homogeneous in the whole reach. Due to the lack of field estimates of bed load at the upstream end of the reach, we assumed the sediment recirculation option available in C-L (i.e., sediment input equals the output at the downstream end of the reach).
The model factor called "Sediment Proportion Recirculated" (SPR) was assumed to be 1, which assumes that upstream sediment load being the same as at the downstream end of the reach (i.e., sediment transport equilibrium condition). Vegetated areas (i.e., islands, recent and old terraces covered by arboreal vegetation) and channelization structures (i.e., bank protection 460 structures, groynes, and levees) were digitized combining 2003 aerial photos and LiDAR Digital Surface Model (DSM, 2 m grid dimension, Table 1). The vegetation cover has been used as the model initial condition for vegetation (maturity fixed to 1). The initial bed elevation was established using a 10 m cell DEM achieved by resampling the 2003 LiDAR DEM (bilinear interpolation, original cell dimension 2×2 m). The 10 m cell dimension was chosen to ensure a reasonable computational time for long term scenario runs and also a spatial resolution higher than previous works (e.g. 25 m in Ziliani et al., 2013). . The hourly discharge series was used as an upstream flow boundary condition. At the downstream end of the reach, a constant energy slope was fixed at 0.0047 m m -1 , equal to the local bed slope. The initial bed sediment grain size was set according to Tomasi (2009). Nine grain size classes were used to define size distribution, which was considered homogeneous throughout the reach. Due to the lack of field estimates of bed load at the upstream end of the reach, we assumed 475 the sediment recirculation option available in C-L (i.e., sediment input equals the output at the downstream end of the reach).
The model factor termed "recirculated sediment proportion" (SPR) was assumed to be 1, which assumes that sediment load upstream is the same as it is downstream (i.e., sediment transport equilibrium condition). Vegetated areas (i.e., islands and recent and old terraces covered by arboreal vegetation) and channelization structures (i.e., bank protection structures, groynes, and levees) were digitized combining 2003 aerial photos and the LiDAR Digital Surface Model (DSM, 2 m grid dimension, 480 Table 1). The vegetation cover was used as the initial modelled vegetation condition (maturity fixed at 1). The initial bed elevation was established using a 10 m DEM cell obtained by resampling the 2003 LiDAR DEM (bilinear interpolation, original cell dimension 2×2 m). The 10 m cell dimension was chosen to ensure a reasonable computational time for long-term scenario runs and for a spatial resolution higher than previous works (e.g. 25 m in Ziliani et al., 2013). The DEM was corrected in the wetted areas (about 8% of the total spatial domain) through the application of the method proposed in Bertoldi et al. 485 (2011) and forced to be not erodible in the areas occupied by both channelization structures still effective in 2003 and undamaged structures built since the 19 th century.

Flow-regime management strategies
In rivers Water management of historically regulated, water management rivers may be oriented to restorerestoring the flow regime to close to the priorthat before impact conditions, typically aiming to reactivateaimed at reactivating physical processes 490 linked to specific components of the flow regime (Wohl, 2011). Nevertheless, existing priorities in uses of the water resourcesresource use often limit the feasibility and the effectiveness of any flow regime re-naturalization strategies, and in most cases the strategy is merely reduced to the definition of a minimum volumesvolume of water released for partial restoration goals. Olden et al. (2014) provided a systematic review of systematically reviewed flood experiments to evaluate globally the success of this practice in flow regime management. They considered 113 flood experiments acrossin 20 countries 495 were reviewed revealingand found that only 11% of the case studies were aimed directly toat morphological effects and about, while around 80% of the experiments involved only low magnitude flow events. Major part ofMost experimental flow releases focussesare focussed on a biological variablevariables (primarily fishesfish), aquatic organisms (Konrad et al., 2011) and reestablishing vegetation reestablishment (Shafroth et al., 2010), rather than on abiotic factorfactors (e.g., channel morphology) (Wohl et al., 2015b). The so-calledWhat are termed "environmental flows" and "Green Hydro" are concepts widely accepted 500 even though theyconcepts that refer mainly to quantity, timing, duration, frequency and quality of water flowsflow releases, as required to sustain freshwater, estuarine and near-shore ecosystems, according to social interests (Acreman and Ferguson, 2010;Rivaes et al., 2017).
Thus, there arehave been few examples of flow regime recovery strategies which have been designed on a geomorphological basis or rather planned to with primarily achieve morphological targetsgoals. The Colorado River below the Glen Canyon 505 Dams represents the most relevantimportant exception: several. Several controlled floods (i.e., five High Flood Experimentshigh flood experiments) were released in thebetween 1996-and 2013 period to maintain and rehabilitate sandbars that occur in lateral flow separation eddies (Schmidt and Wilcock, 2008;Melis, 2011;Mueller et al., 2014). Also theThe Trinity River (California, USA) represents ananother excellent case study, as a in which morphological goals were among the multiple objective flow designed included morphological issueobjectives (Trinity Management Council, 2014). Still, Nevertheless, with 510 the exception of the lower Waitaki River in New Zealand, there are no experiences are available onexamples of morphological recovery or conservation of braided rivers, with the exception of the Lower Waitaki River (New Zealand), even thoughand no controlled floods have been used in that riveron the Waitaki (Hicks et al., 2003;Hicks et al., 2006;Environment Canterbury Regional Council, 2015).

ControlledProvoking controlled floods andor removing vegetation removal actions are expensivecostly and potentially 515
foregocan preclude other positive feedback effects. It is therefore worth to preliminary evaluate and testactions, so it worthwhile to assess these kinds of actions taking into accountstrategies and consider the morphological effects of different channel-forming discharges . As emphasized by Rathburn et al. (2009),) stressed that flow releases from dams must exceed a series of thresholds to be morphologically effective. Four discharge thresholds of increasing threshold dischargesintensity should be identified in a morphological recovery/conservation plan, each able to activatecapable of 520 activating a specific morphological process: (i) the mobilization of interstitial sediment essential for the hyporheic exchange maintenance; (ii) the mobilization of the streambed to maintain natural bedforms; (iii) the inundation of overbank units (i.e., berms, floodplains, terraces) to confine encroachment by xeric plants; and (iv) the lateral channel mobility that may promote the removal ofcan remove senescent woody vegetation and create opportunities for seedlings to germinate and mature. Once each threshold discharge value or range is quantified, the flow duration can be tuned within the limits imposed by the 525 actualavailable flow availability, assuming, using the natural flow regime as a reference.
In this work, the scenarios of The flow regime management scenarios used in this work were defined referringin reference to three flood threshold levels that partially matchingmatch those proposed inby Rathburn et al. (2009): (i) thefull in-channel full transport discharge assumed to be able to mobilize thecapable of mobilizing interstitial sediment, (ii) the bankfull discharge able to maintain thecapable of maintaining natural bedforms; and (iii) the overbank discharge able to affectcapable of affecting 530 the main lateral units by inundation. A field study conducted in the reach using painted sediments similar to (those used in Mao and Surian., (2010;) and Mao et al., . (2017) allowed toresulted in an estimate thatof the first reference threshold is aboutat 80 m 3 s -1 (RI < 1 year, full transport discharge able to mobilizecapable of mobilizing sediments, irrespective of their size). The bankfull discharge was estimated using the calibrated C-L model and was approximated to the discharge filling the active channels and bars without overflowing onto the oldest island assumed to be morphologically equivalent to the recent 535 fluvial terraces (Williams, 1978;Pickup and Rieger, 1979;. The bankfull discharge is about 500 m 3 s -1 , and corresponds to a 2.5 years RI, coherentlywhich is consistent with other literature references in the literature (Leopold, 1994).
Only the > 1,000 m 3 s -1 floods were identified to be able toas capable of completely inundateinundating the oldest island and overflowing locally overflow into the recent fluvial terraces.
Four simulation scenarios, each were considered, all covering the same 25 years long-year period (2009-2034), were explored) 540 (Table 2). The first scenario (i.e., "baseline scenario",, SC1) corresponds tois the current condition, characterized by a strongly altered flow regime. In this case, the discharge series measured infor 1995-2009 (hourly data at Belluno gauge station) has beenwas repeated twice. ScenariosBoth scenarios 2 (SC2) and 3 (SC3) were both set up usingused a flow regime strategy characterized by one CF per year. In SC2 theThe yearly CF in SC2 had a constant value of 135 m 3 s -1 (RI ~ 1.08 years). This value has been), which was calculated as the average of the maximum annual floods observed in 1995-2009,. This was higher 545 than the reference threshold discharge, i.e. the full in-channel full transport discharge (80 m 3 s -1 ). In SC3, theThe yearly CF values for SC3 were randomly selected above the threshold discharge using the natural streamflow pdf estimated by the model (minminimum value = 80 m 3 s -1 ; maxmaximum value = 276 m 3 s -1 , RI ~ 1.4 years). In this case, the average value of all the CF values was found to beis equal to the SC2 yearly CF discharge for SC2, assuming values included in the [80 -276] m 3 s -1 range.
All the CFs had a fixed duration of 5 days, according to the re-naturalization maximization criteria. In SC2 and SC3, the 550 cumulative likelihood to occur associated towith the released peaks raisedincreased from 0.025 (actual observed altered regime) to 0.04, being the natural reference value being 0.14. Scenario 4 (SC4) was planned to represent a different management strategy, consisting inof larger CFCFs released by dams (constant value for one-day) only following the observation of notable channel narrowing. Specifically, we assumed 200 m as a threshold for average channel width, considering the evolutionary trajectory over the last 200 years and, in particular, the most intense narrowing that took place in 555 the early 1990s (Fig. 3). Taking into account theBased on channel width measurements conductedtaken by a one-year step duringin the SC4 simulation, only two CFs were released, one in 2020 and the other in 2032 respectively.. The released discharge was fixed equal toat 600 m 3 s -1 (RI 5 years)), ranging between the second and the third reference threshold discharges discussed above, so that it was surely able to maintain thewould undoubtedly be capable of maintaining in-channel bedformsbedform dynamics, while completely avoiding any hydraulic risk issuesrisks and damages in the overbank units (i.e., 560 recent terraces), locally occupied by secondary roads and cultivated fields.
All the CFs have beenwere released in November avoiding to overlap existing, while overlapping floods. were avoided. All the scenarios made use of the same model setting achieved, which was obtained in the calibration phase, using the 2009 calibration run output data in raster format as the initial boundary conditions (i.e., bed elevation, not erodibleunerodible hydraulic structures, vegetation cover, grain size bed sediment distribution). As reported in Table 2, all the scenarios required 565 the release of a cumulative volume to be released per year (i.e., per flood)), which was considerably lowerless than the maximum stocked volume in the reservoirs existing upstream offrom the studystudied reach (90.8 Mm 3 ). SC2 and SC3 required similar volumes of about 58 Mm 3 of water, corresponding to aboutor around 1.45x10 3 Mm 3 onover the wholeentire scenarios period. SC4 represents the cheapestlast expensive scenario indeed because it needed 51.8 Mm 3 per year and 104 Mm 3 globally, one order of magnitude lowerless than in the other scenarios. 570

Evolutionary trajectory of channel morphology
Using the available dataset on morphological changes of the Piave River (Comiti et al., 2011), we reconstructed and updated the channel adjustments up to 2009. The analysis focused on two sub-reaches, respectively upstream and downstream offrom San Pietro in Campo, where the river is naturally more confined (Fig. 1). The division in reach was divided into two sub-575 reaches was used to better describe the morphological adjustments over the 1800-2009 period. The trends in average width trends are similar for bothin the sub-reaches are similar, and are characterized by four main adjustment phases (Fig. 3) was the lowest value observed in the studystudied reach over the last 200 years. While widening during phase III widening was likely due to the termination of in-channel gravel mining (Comiti et al., 2011), the most recent phase of narrowing (i.e., phase (phase IV) was likely due to the absence of major floods (see also Fig. 2).

Flow regime alterations 590
TheA comparison betweenof the frequency distribution of observed daily streamflows at the Belluno cross -section and the model-based estimate of the streamflow pdf under unregulated conditions (Fig. 4) shows the extent of thesignificant impact of regulation in the lower reaches of the Piave River. The mean and the mode of the streamflow distribution are significantly reduced by the anthropogenic exploitation of water resources (i.e., by-pass flows and diversions). Accordingly, the exceedance probability of moderate to high flows is significantly reduced under current regulated conditions. In particular, the probability 595 to observe discharges larger than 80 m 3 s -1 is reduced by about one order of magnitude (i.e., from 0.14 to 0.025). Such results are crucial for setting the flow-regime management scenario since (i) they show that a strategy aiming to improveaimed at improving the current flow-regime should be implementedis needed, (ii) this strategy should compensate the expected low morphological dynamism of the river caused by the decreased occurrence of discharges able to mobilizecapable of mobilizing sediments and produceproducing significant morphological changes in the studystudied reach. 600 It is worth to notenoting that the hydrological model underestimates the frequency of the highest flows (i.e., discharges largerof more than 300 m 3 s -1 ) because all the non-linearities of the hydrologichydrological response (e.g., the presence of different flow components such as surface runoff) are neglected in this version of the model (Basso et al., 2015). As a consequence, the probability associated towith the highest flows in regulated conditions is largergreater than the corresponding value estimated by the stochastic model for the natural setting. This model However, this limitation, however, of the model does not bearhave 605 any significant consequenceconsequences for the analysis carried out in this paper, provided that the frequency of such high flows is relatively low.

Calibration of the morphodynamic model
The results presented inof Ziliani et al. (2013) and Coulthard et al. (2013) have been takenwere used as a reference to achievecalibrate the C-L calibrationmodel. According to the results of the sensitivity analysis in Ziliani et al. (2013), the lateral 610 erosion rate and maximum erosion limit have been assumed aswere the most sensitive factors that required accurate tuning.
The other factors (see Table 3), including the main new parameters introduced in the C-L version, were tuned manually through a "by trial-and-error" calibration strategy (n. (75 runs in total). Following the performance evaluation techniques used by Ziliani et al. (2013), the calibration was based on performance indices developed specifically for data available in a raster format (Bates and De Roo, 2000;Horritt and Bates, 2001). The performance indices reported in Table 4 were calculated for 615 all the calibration runs at the end of the simulation (2009), that isthese being (i) the vegetation performance index (Fveg), (ii) the wet area performance index (Fwet) and (iii) the active channel performance index (Fc). In addition, several planimetric features have beenwere calculated, including (i) average active channel width, (ii) equivalent wet area width (Lw), and (iii) the mean braiding index (Egozi and Ashmore, 2008). The results (see Table 4, Fig. 5, Fig. S3 and S4 in the "Supplementary material" fileMaterial File) show a "very good performance" (performance class as defined in Henriksen et al., 2003;Allen et 620 al., 2007) for both the vegetation cover (Fveg 69.7%) and the active channel area (Fc 54.2%). Output values of the active channel width and braiding index values confirmed these results. The difference between the real and modeledmodelled 2009 active channel width (6 m) is lowerless than the input DEM cell size (10 m)), and the modeledmodelled braiding index value (1.71) is very close to the real value (1.69). The model performance is poor only performed poorly in reproducing the flowing channel position (Fw 15.7 %)%), which partially confirmingconfirms the results presented inof Ziliani et al. (2013). 625 In order to integrate the morphological performance evaluations, we carried out an estimation of theWe estimated mean annual bed load sediment yield at the downstream end of the reach and along the wholeentire reach. In the 2003-2009 period, the modeled average to integrate morphological performance evaluations. Average modelled bed load sediment yield resulted of aboutin the 2003-2009 period was around 21.5 x 10 3 m 3 yr -1 . ModeledModelled yield varies significantly along the reach (up to 30%) taking%), with higher yearly values in the sub-reach upstream San Pietro in Campo. Significant differences exist average annual sediment yield ofis about 260 m 3 yr -1 versus the 2008 maximum of about 53.3 x 10 3 m 3 yr -1 . SuchThese sediment transport values agree with estimates for gravel-bed rivers with similar characteristics to those of the Piave River reach (Martin and Church, 1995;Ham and Church, 2000;Nicholas, 2000;Liebault et al., 2008;Ziliani et al., 2013;Mao et al., 2017).

Channel response to flow regime management strategies: scenario results 635
ChannelThe channel adjustments induced by allthe different scenarios were assessed by comparing every year (in February) the active channel width and the braiding intensity (BI) in each year (in February) using the same techniques adopted in the calibration phase (Fig. 6). Channel width in Scenarios 2-4 was almost always highergreater than in SC1 (the "baseline scenario"). On average, during the whole scenario period, SC2 and SC4 produced comparable). Average channel widening of aboutin SC2 and SC4 were similar throughout the scenarios, being approximately 6.4% (~ 14 m), while at the end of the 640 scenarios (2034), widening was about 9% (~ 25 m) and 13.5% (~ 38 m) in SC2 and SC4, respectively. The SC3 induced a slightly lowerless widening, about 5.4%, duringthroughout the whole period, and 8.6% (~ 24 m) at the end of the simulated period. The maximum annual widening was observed in SC4 (~ 120 m in 2033), followed by SC3 (~ 77 m in 2020) and SC2 (~ 43 m in 2032 - Fig. 7).
Results suggest that the CFs scenarios (SC2-4) and the baseline scenario (SC1) provide similar long-term morphological 645 trajectories characterized by alternate phases of widening and narrowing and notable changes in active channel width (width varies between 150 and 360 m). Figure 7 shows that each channel width oscillation takesoscillations take place in about over a period of 6-to 7 years and it has 160 man amplitude of 160 m in response to the alternation of periods characterized by different magnitude floods series: in the 2011-2015 and 2022-2028 periods, during which seven floods > 400 m 3 s -1 (RI ~ 1.9 years) occur, channel width follows a quasi-steady trend and is largergreater than 300 m. Instead, channel width decreased 650 during the following periods (2017-2021 and 2029-2031) affected by lower magnitude floods (200 m 3 s -1 maximum peak value), channel width shows decreasing trajectories. Over the whole 25 years,). SC1 provideshad a slightly decreasing trend ( Fig. 7) that is not reversedover the entire 25 years (Figure 7). Channel width has quasi-zero slopes in the otherCF scenarios.
In all the CFs scenarios the channel width trend assumes quasi-zero slope, even if the channel width measured at the end of the simulationssimulation is about [8.6-13.5 %] highermore than width in the baseline scenario. 655 The braiding index inindices of Scenarios 2-4 waswere similar, to or lower, than inthat of SC1. SC4 resultedwas the most similar scenario closest to thethat of SC1, with aan average BI in-time averaged value equal to of 2.78, onlywhich was slightly lower than thethat of SC1 (-1.5%). During SC2 we measured relativeDifferences in BI were higher differencesin SC2 than in the BI value compared to SC1 (-7.3%, 0.21 BI unit - Fig. 7), although these differences are quitewere small. In termsThe behaviour of trajectory, braiding intensity shows awas different behaviour in comparison to from channel width, as in that 660 there are no clear oscillation phases butwas one period (from 2009 to 2023) with ano clear oscillations, and a clearly increasing trend, followed by a decreasing or quasi-steady (SC1) period until the end of the simulation. There is a non-linear correlation between BI and flooding series magnitude or the CFs. In particular, SC1 is the only scenario that doesdid not show aan inverse trend inversion after 2023, andwhile the SC2 scenario has a very anomalous trend showing a anomalously had a consistently lower BI value steadily lower that the other CFsCF scenarios, while SC3 and SC4 show a good agreement in theirhad very 665 similar BI trendsvalues.

Geomorphic effectiveness of controlled floods
Comparing theSeveral insights can be obtained from comparing future scenarios to the historical evolutionary trajectory ( less than in 1999. This may suggestsuggests that the study reach, after a period (phases I and II in Fig.Figure 3) of morphological instability characterized by a prevalentnarrowing tendency to narrowing, has reached, the studied reach had acquired a new morphological equilibrium configuration characterized by a periodic oscillations ofin channel width. Similar new equilibrium conditions, mainly controlled by the flow regime (i.e. frequency and magnitude of formative discharges) and 685 vegetation establishment, have been observed in the Tagliamento River (Ziliani and Surian, 2016).
The intercomparisoncomparison of our four simulations shows that a few high magnitude floods provide slightly better morphological recovery/conservation than small yearly floods, alsoand at a significantly lower operational cost. Therefore, SC4 should be preferredis preferable to SC2 and SC3 from a purepurely morphodynamic point of view. Nevertheless, results suggestthe comparison suggests that (i) none of the CFsCF scenarios are able to change can significantly thechange long-term 690 channel width and braiding intensity trends, (ii) CFs releaseCF releases have no significant morphological benefits and do not represent a solution for athe morphological recovery inof braided rivers that suffered such strong and historical impacts in terms of flow and sediment supply regimes. It is worth noting that the selected CFs are feasible, that is taking into account the water infrastructure in the Piave River basin, and it is unlikely that higher or more frequent floods could be released. These results partially confirm the outcomes of Hicks et al. (2003) referring toon the Waitaki River, a gravel-bed river with similar consisting in the frequent release of "channel maintenance floods" from dams should be pursued. if wider and more active channels are desired. Hicks et al. (2013) showshowed that this kind of strategy may behas been unsuccessful and only multiyearsyear high magnitude CFs can produce temporary stable effective channel widening channel condition.
The cost of CFs is probably smallerless than thosethat of alternative strategies focused on increasing sediment supply, such as 700 sediment augmentation, because flood releases commonly can be performed without redesign ofoften do not require redesigning reservoir structures. Nevertheless, reintroduction of flood flows implies "loss" of resourceresources stocked for other purposes (e.g., hydroelectric production, and supply of drinking or irrigation water supply). Another feasible way for sediment augmentation is the removal (at least in part) of non-strategic bank protections along the reach. However, as suggested by Picco et al. (2016),) suggested, this kind of strategy should be preventively assesseda last resort since these structures are 705 still viewed by local populations as necessary to protect riparian woodlands that are highly appreciated for recreation and timber production.
Overall, this work gives useful insights for the Piave River management and, in general, for management of braided rivers with heavily impacted in flow and sediment regimes: (i) none of the tested controlled flood strategies that was tested is able tocan significantly change the on-going morphological evolution; (ii) the baseline scenario, without controlled flood releases 710 (i.e., the no action strategy), provides a similar morphological evolutionevolutionary trajectory similar to that induced by the controlled floodsflood release scenarios. Therefore, a main outcome is that controlled floods (including high-magnitude floods, e.g. 5-year RI) may not have any significant effects on regulated rivers, specifically if formative discharges have been strongly altered.

Assessment of CAESAR-LISFLOOD performance
In Ziliani et al. (2013) the authors concluded that the main factors causingreasons for the poor morphological poor response of CAESAR are (i) the DEM cell size, aswhich has been pointed out also in others works (Doeschl-Wilson and Ashmore, 2005;Doeschl et al., 2006;Nicholas and Quine, 2007), (ii) the quality of data (i.e., lack of wet channel topography)), and (iii) the low flow periodsperiod removal, and therefore the cut offelimination of the consequent morphological "gardening" 720 phenomena (Ziliani et al., 2013). The combination of these factors produced a smoother and simpler braided morphology. The Piave case study represents an effort to achieve a better performance by (i) the flow routinesroutine refinement included in the LISFLOOD-FP module (one of the most recent and advanced Reduced Complexity Hydraulic Model schemereduced complexity hydraulic model schemes), (ii) the adoption of input data of higher quality input data (higher resolution DEM, bathymetry and hourly boundary conditions) and (iii) the code conversion in parallel programming methods. The results lead 725 to an overall improvement of the model performance considering (i) the good channel width performance in the calibration phase, (ii) the excellent reproduction of braiding complexity reproduction, including the pioneerpioneering and complex islandsisland dynamics, both in the calibration and inthe long-term simulations, (iii) the reasonable estimation of bedload transport, and the small changes in bed grain size in the long-term simulations (e.g. D50 changed from 24.9 mm to 22.7, 24.3, Formattato: Tipo di carattere: +Corpo (Times New Roman), Inglese (Stati Uniti) Formattato: Colore carattere: Rosso 23.6 and 23.1 mm respectively in SC 1, 2, 3, and 4), and (iv) the adequate computation speed, close to the expectations (i.e., 730 what was expected (about 10 days of computation for 25 years of hourly series).
The suitability of the RCMs application for the investigation ofinvestigating river dynamics has been discussed in several previous studies (Doeschl-Wilson and Ashmore, 2005;Brasington and Richards, 2007;Nicholas and Quine, 2007;Murray, 2007;Nicholas, 2012Nicholas, , 2013bZiliani et al 2013;Ziliani and Surian, 2016). A general conclusion of these works is that RCMs may provide morphological responses both unrealistic and highly sensitive morphological responses to model grid resolution. 735 These problems are commonly interpreted as a direct consequence of both the adoption of flow routing schemes that neglect the momentum conservation and the use of local bed slopes for thein calculating bedload transport calculation (e.g., through the application of the uniform flow approximation).
The C-L model maycan be considered ana useful tool in the search of an effective combination of simplicity and physical realism in the context of reduced-complexity modelling, overcoming some of the previousproblems associated with earlier 740 simplified hydrodynamic simplification issues.models. The encouraging results achieved inof this case study seem to justifysuggest the effort faced in such further development ofto develop this RCM. Although is justified. While the physical realism of flow and morphodynamic rules can remain unsolved at smallersmall scales (i.e., scales lower thanbelow the DEM cell dimension), the improvementsuperiority of the C-L model response at reach scales compared to over the older CAESARearlier CESAR model is evident. Although the reduced at reach scale. Reduced-complexity modelling approach 745 model will probably will not provide insights into some of the answers to key reductionist key questions currently faced byof hydraulic engineers and fluvial geomorphologists, nevertheless the model can provide useful insights for management.
Specifically, insights, specifically about their macro morphologicalmacromorphological river features (e.g. average channel width, braiding intensity) and adjustments (e.g. prediction of future evolutionary trajectory) of braided rivers.
The inherent limitation in limitations of reduced-complexity modelling approach doesmodels do not preclude the adoption of 750 using RCMs where the aim is to represent meso-scale system behaviour at the mesoscale, rather than to makemaking reductionist predictions that are theoretically more accurate in quantitative terms. Reproducing the morphodynamic processes at each scale requiredrequires necessarily some forms of simplification, regardless of the level of complexity of the model adopted, and in CFD models as well.. Nevertheless, this cannot be considered a convincingis not sufficient reason to necessarily callingcall into question whether explanationsthe usefulness of river behaviour based onapplying this kind of model have any 755 application in the models in real world situation. In the light of the results presented inof this work and of the limitations faced anyway by theof reductionist alternative approaches (Williams et al., 2016), we believe that RCMs, and C-L model specifically, remain an the C-L model are attractive optionoptions for simulating river evolution over historical time periodshistorically and into the future scenarios, that is at the scale of interest for river management.
This work presents another case study in which an RCM has given realistic outputs in a large gravel-bed river, especially in 760 terms of evolutionary trajectories. The suitability in reproducing macro morphological features and meso-scale processes should not be questioned any longer (Nicholas, 2013b). The capability to model small-scale phenomena remains open for RCMs as for all CFDs that try to reproduce phenomena deeply influenced by initial and boundary conditions, for which a data gap persists for future scenario application in natural contexts where the addition of modelling details does not guarantee a significant reduction of the overall uncertainty associated to the model results. 765 This work presents another case study in which an RCM provided realistic outputs regarding a large gravel-bed river, especially in terms of evolutionary trajectories. The suitability in reproducing macromorphological features and mesoscale processes should no longer be in doubt (Nicholas, 2013b). The capability of modelling small-scale phenomena remains open for RCMs as for all CFDs that are applied to reproduce phenomena deeply influenced by initial and boundary conditions, for which a data gap persists for future scenario applications in natural contexts where the addition of modelling details does not guarantee 770 a significant reduction of the overall uncertainty associated with modelling results.

Conclusions
Hydrological and morphodynamic models have been applied to assess the long-term geomorphic effectiveness of controlled floodsflood strategies. The simulated future scenarios (with a duration of 25 years) show that: (i) none of the CFsCF strategies can provide significant long-term morphological benefits and is able toor reverse the ongoing channel width trend; (it is worth 775 noting that the selected CFs are feasible and it is unlikely that higher or more frequent floods could be released); (ii) fewa small number of high magnitude floods (i.e. SC4) provide slightly better morphological recovery than smalldo yearly lowmagnitude floods (i.e. SC2 and SC3), also at aas well as having significantly lower operational cost (the cumulative volume released in SC4 is an order of magnitude lower than in SC2 and SC3). These results suggest that this kind of strategy does not represent a solution for morphological recovery in braided rivers strongly and historically impacted braided rivers. 780 TheThis study confirms the suitability of the RCMs for modelling future long-term future scenarios at spatial and temporal scales still hardlythat are not well supported by 2D-3D CFD morphodynamic models. From athe morphological point of viewperspective, the C-L model has proven to be able to reproduce the capable of reproducing variations in channel width variation, to preserve, preserving the complexity of morphological braiding complexity, including the vegetation dynamics, and to estimate reasonably theestimating average bed load sediment yield. The model performance assessment shows 785 significant improvements of that the C-L model in comparison toperforms significantly better than previous CAESAR model versionversions (Ziliani et al. 2013). The application of thisWhile the RCM does not provide insights into the spatial and temporal scales of interest for a traditional reductionist approach (e.g., single branch and bar dynamics, local bank erosion) however), it providesdoes provide useful indications for management of braided rivers at meso-scalesmesoscales.

Data availability
LiDAR data and aerial photos (2003) are available upon request at the Autorità di Bacino delle Alpi Orientali.
Hydrological data are available upon request at the Environmental Regional Agency (ARPA Veneto).
2009 aerial photos and cross sections are freely available by contacting the authors. 795

Supplement
The supplementary material related to this article is available online at: http://researchdata.cab.unipd.it/id/eprint/157

Author contribution
LZ and NS designed the research. LZ performed most of the analyses. GB aided as expert in hydrology and hydrological modelling. LM provided guidance on sediment transport. All authors jointly contributed to the discussion and interpretation 800 of the data. The paper was prepared by LZ, with contributions from NS, GB and LM. NS managed and coordinated research activities.

Competing interests
The authors declare that they have no conflict of interest.