Evaluation of distributed hydrologic impacts of temperature-index and energy-based snow models

Two commonly used strategies in modeling snowmelt are the energy balance and temperature-index methods. Here we evaluate the distributed hydrologic impacts of these two different snowmelt modeling strategies, each in conjunction with a physics-based hydrologic model (PIHM). Results illustrate that both the Isnobal energy-balance and calibrated temperature-index methods adequately reproduce snow depletion at the observation site. However, the models exhibit marked differences in the distribution of snowmelt. When combined with PIHM, both models capture streamflow reasonably during calibration year (WY06), but Isnobal model gives better streamflow results in the validation year (WY07). The uncalibrated temperature-index model predicts streamflow poorly in both years. Differences between distributed snowmelt, as predicted by Isnobal and calibrated temperature-index method, and its consequent effect on predicted hydrologic states suggest the need to carefully calibrate temperature-index models in both time and space. Combined physics-based snow and hydrologic models provide the best accuracy, while a temperature-index model using coefficients from the literature the poorest.     

Process-based snowmelt modeling: does it require more input data than temperature-index modeling?

Modeling snow hydrology for cold regions remains a problematic aspect of many hydro-environmental models. Temperature-index methods are commonly used and are routinely justified under the auspices that process-based models require too many input data. To test this claim, we used a physical, process-based model to simulate snowmelt at four locations across the conterminous US using energy components estimated from measured daily maximum and minimum temperature, i.e. using only the same data required for temperature-index models. The results showed good agreement between observed and predicted snow water equivalents, average R²>0.9. We duplicated the simulations using a simple temperature-index model best fitted to the data and results were poorer, R²<0.8. At one site we applied the process-based model without substantial parameter estimation, and there were no significant (=0.05) differences between these results and those obtained using temperature-estimated parameters, despite relatively poorly predicted specific energy budget components (R²<0.8). These results encourage the use of mechanistic snowmelt modeling approaches in hydrological models, especially in distributed hydrological models for which landscape snow distribution may be controlled by spatially distributed components of the environmental energy budget.     

CURRENT RESEARCH IN BRITISH SNOWMELT RIVER FLOODING

ABSTRACT

An adaptation of the degree-day method has been used to analyse a number of snowmelt events on two catchments as a first step in a programme of research on snowmelt river flooding in Britain. The analysis indicates that the degree-day factor varies during events and between events on the same catchment. A snowmelt event is seen as consisting of three phases, an initial lag phase, a phase of nearly constant degree-day factor and a recession phase. The degree-day factor in the constant part of each event has significant correlation with the total flood volume on both catchments and with liquid precipitation during the snowmelt on one catchment only. Separate procedures are considered necessary for forecasting the initial lag phase and runoff during the recession.     

Integrated Surface and Subsurface Hydrological Modeling with Snowmelt and Pore Water Freeze–Thaw

For the simulation of winter hydrological processes a gap in the availability of flow models existed: one either had the choice between (1) physically-based and fully-integrated, but computationally very intensive, or (2) simplified and compartamentalized, but computationally less expensive, simulators. To bridge this gap, we here present the integration of a computationally efficient representation of winter hydrological processes (snowfall, snow accumulation, snowmelt, pore water freeze–thaw) in a fully-integrated surface water-groundwater flow model. This allows the efficient simulation of catchment-scale hydrological processes in locations significantly influenced by winter processes. Snow accumulation and snowmelt are based on the degree-day method and pore water freeze–thaw is calculated with a vertical heat conduction approach. This representation of winter hydrological processes is integrated into the fully-coupled surface water-groundwater flow model HydroGeoSphere. A benchmark for pore water freeze–thaw as well as two illustrative examples are provided.     

Modelling the hydrological response of debris‐free and debris‐covered glaciers to present climatic conditions in the semiarid Andes of central Chile

We apply the process‐based, distributed TOPKAPI‐ETH glacio‐hydrological model to a glacierized catchment (19% glacierized) in the semiarid Andes of central Chile. The semiarid Andes provides vital freshwater resources to valleys in Chile and Argentina, but only few glacio‐hydrological modelling studies have been conducted, and its dominant hydrological processes remain poorly understood. The catchment contains two debris‐free glaciers reaching down to 3900 m asl (Bello and Yeso glaciers) and one debris‐covered avalanche‐fed glacier reaching to 3200 m asl (Piramide Glacier). Our main objective is to compare the mass balance and runoff contributions of both glacier types under current climatic conditions. We use a unique dataset of field measurements collected over two ablation seasons combined with the distributed TOPKAPI‐ETH model that includes physically oriented parameterizations of snow and ice ablation, gravitational distribution of snow, snow albedo evolution and the ablation of debris‐covered ice. Model outputs indicate that while the mass balance of Bello and Yeso glaciers is mostly explained by temperature gradients, the Piramide Glacier mass balance is governed by debris thickness and avalanches and has a clear non‐linear profile with elevation as a result. Despite the thermal insulation effect of the debris cover, the mass balance and contribution to runoff from debris‐free and debris‐covered glaciers are similar in magnitude, mainly because of elevation differences. However, runoff contributions are distinct in time and seasonality with ice melt starting approximately four weeks earlier from the debris‐covered glacier, what is of relevance for water resources management. At the catchment scale, snowmelt is the dominant contributor to runoff during both years. However, during the driest year of our simulations, ice melt contributes 42 ± 8% and 67 ± 6% of the annual and summer runoff, respectively. Sensitivity analyses show that runoff is most sensitive to temperature and precipitation gradients, melt factors and debris cover thickness. Copyright © 2016 John Wiley & Sons, Ltd.     

Modelling the spatial–temporal variability of spring snowmelt in an arctic catchment

Arctic spring landscapes are usually characterized by a mosaic of coexisting snow‐covered and bare ground patches. This phenomenon has major implications for hydrological processes, including meltwater production and runoff. Furthermore, as indicated by aircraft observations, it affects land‐surface–atmosphere exchanges, leading to a high degree of variability in surface energy terms during melt. The heterogeneity and related differences when certain parts of the landscape become snow free also affects the length of the growing season and the carbon cycle. Small‐scale variability in arctic snowmelt is addressed here by combining a spatially distributed end‐of‐winter snow cover with simulations of variable snowmelt energy balance factors for the small arctic catchment of Trail Valley Creek (63 km²). Throughout the winter, snow in arctic tundra basins is redistributed by frequent blowing snow events. Areas of above‐ or below‐average end‐of‐winter snow water equivalents were determined from land‐cover classifications, topography, land‐cover‐based snow surveys, and distributed surface wind‐field simulations. Topographic influences on major snowmelt energy balance factors (solar radiation and turbulent fluxes of sensible and latent heat) were modelled on a small‐scale (40 m) basis. A spatially variable complete snowmelt energy balance was subsequently computed and applied to the distributed snow cover, allowing the simulation of the progress of melt throughout the basin. The emerging patterns compared very well visually to snow cover observations from satellite images and aerial photographs. Results show the relative importance of variable end‐of‐winter snow cover, spatially distributed melt energy fluxes, and local advection processes for the development of a patchy snow cover. This illustrates that the consideration of these processes is crucial for an accurate determination of snow‐covered areas, as well as the location, timing, and amount of meltwater release from arctic catchments, and should, therefore, be included in hydrological models. Furthermore, the study shows the need for a subgrid parameterization of these factors in the land surface schemes of larger scale climate models. Copyright © 2005 John Wiley & Sons, Ltd.     

INCORPORATING RADIATION INPUTS INTO THE SNOWMELT RUNOFF MODEL

Process-based, distributed-area snowmelt runoff models operating at small scales are essential to understand subtle effects of climate change, but require data not commonly available. Temperature index models operating over large areas provide realistic simulations of basin runoff with operationally available data, but lack rigorous physically based algorithms. A compromise between the two types of models is required to provide realistic evaluations of basin response to environmental changes in cold regions. One adaptation that is uniformly required for snowmelt models is the use of remotely sensed data, either as input or in model validation. At a minimum, snowmelt forecasting models need to incorporate snowcover extent information, which is currently obtained operationally. As more remote sensing capabilities come on line, models should accept upgraded information on snow water equivalent; additional remotely sensed information on landcover, frozen soil, soil moisture, cloudiness and albedo would also be useful. Adaptations to the semi-distributed snowmelt runoff model (SRM) are underway to make it more physically based for use in large area studies. A net radiation index has been added to the model, which formerly used only a temperature (degree–day) index to melt snow from a basin's elevation zones. The addition of radiation to the SRM allows the basin to be subdivided into hydrological response units by general aspect (orientation) as well as elevation. Testing of the new radiation-based SRM with measured radiation from a small research basin is the first step towards large scale simulations. Results from the W-3 research basin in Vermont, USA are promising. In the radiation version, the factor that multiplies the degree–day index is estimated independently of model output and is held constant throughout the season, in contrast with the degree–day version, where the corresponding factor is allowed to increase throughout the season. Without calibrating or optimizing on this important parameter, the goodness-of-fit measure R² is improved in two out of six test years when the radiation version of the SRM is used in place of the degree–day version in melt season simulations. When the accumulation of error is eliminated with periodic updating of streamflow, more significant improvement is noted with radiation included.     

Future runoff from a partly glacierized watershed in Central Switzerland: A two-model approach

We present a comprehensive hydrological modeling study in the drainage area of a hydropower reservoir in central Switzerland. To investigate the response of this 95 km² alpine watershed to a changing climate, we used both a conceptual and a physically based hydrological model approach. The multi-model approach enabled detailed insights into the uncertainties associated with model projections of future runoff based on climate scenarios. Both hydrological models consistently predicted changes of the seasonal runoff dynamics, including the timing of snowmelt and peak-flow in summer as well as the future spread between high and low flow years. However the models disagreed regarding the evolution of glacier melt rates thus leading to a considerable difference in predicted annual runoff figures. The findings suggest that snow-glacier feedbacks require particular attention when predicting future runoff from glacio-nival watersheds.     

Storage-discharge characteristics of an active rock glacier catchment in the Innere Ölgrube,Austrian Alps

The active rock glacier “Innere Ölgrube” and its catchment area (Ötztal Alps, Austria) are assessed using various hydro(geo)logical tools to provide a thorough catchment characterization and to quantify temporal variations in recharge and discharge components. During the period from June 2014 to July 2018, an average contribution derived from snowmelt, ice melt and rainfall of 35.8%, 27.6% and 36.6%, respectively, is modelled for the catchment using a rainfall-runoff model. Discharge components of the rock glacier springs are distinguished using isotopic data as well as other natural and artificial tracer data, when considering the potential sources rainfall, snowmelt, ice melt and longer stored groundwater. Seasonal as well as diurnal variations in runoff are quantified and the importance of shallow groundwater within this rock glacier-influenced catchment is emphasized. Water derived from ice melt is suggested to be provided mainly by melting of two small cirque glaciers within the catchment and subordinately by melting of permafrost ice of the rock glacier. The active rock glacier is characterized by a layered internal structure with an unfrozen base layer responsible for groundwater storage and retarded runoff, a main permafrost body contributing little to the discharge (at the moment) by permafrost thaw and an active layer responsible for fast lateral flow on top of the permafrost body. Snowmelt contributes at least 1/3rd of the annual recharge. During droughts, meltwater derived from two cirque glaciers provides runoff with diurnal runoff variations; however, this discharge pattern will change as these cirque glaciers will ultimately disappear in the future. The storage-discharge characteristics of the investigated active rock glacier catchment are an example of a shallow groundwater aquifer in alpine catchments that ought to be considered when analysing (future) river runoff characteristics in alpine catchments as these provide retarded runoff during periods with little or no recharge.     

An enhanced temperature index model for debris-covered glaciers accounting for thickness effect

Debris-covered glaciers are increasingly studied because it is assumed that debris cover extent and thickness could increase in a warming climate, with more regular rockfalls from the surrounding slopes and more englacial melt-out material. Debris energy-balance models have been developed to account for the melt rate enhancement/reduction due to a thin/thick debris layer, respectively. However, such models require a large amount of input data that are not often available, especially in remote mountain areas such as the Himalaya, and can be difficult to extrapolate. Due to their lower data requirements, empirical models have been used extensively in clean glacier melt modelling. For debris-covered glaciers, however, they generally simplify the debris effect by using a single melt-reduction factor which does not account for the influence of varying debris thickness on melt and prescribe a constant reduction for the entire melt across a glacier.In this paper, we present a new temperature-index model that accounts for debris thickness in the computation of melt rates at the debris-ice interface. The model empirical parameters are optimized at the point scale for varying debris thicknesses against melt rates simulated by a physically-based debris energy balance model. The latter is validated against ablation stake readings and surface temperature measurements. Each parameter is then related to a plausible set of debris thickness values to provide a general and transferable parameterization. We develop the model on Miage Glacier, Italy, and then test its transferability on Haut Glacier d’Arolla, Switzerland.The performance of the new debris temperature-index (DETI) model in simulating the glacier melt rate at the point scale is comparable to the one of the physically based approach, and the definition of model parameters as a function of debris thickness allows the simulation of the nonlinear relationship of melt rate to debris thickness, summarised by the Østrem curve. Its large number of parameters might be a limitation, but we show that the model is transferable in time and space to a second glacier with little loss of performance. We thus suggest that the new DETI model can be included in continuous mass balance models of debris-covered glaciers, because of its limited data requirements. As such, we expect its application to lead to an improvement in simulations of the debris-covered glacier response to climate in comparison with models that simply recalibrate empirical parameters to prescribe a constant across glacier reduction in melt.     

Analysis of seasonal snowmelt contribution using a distributed energy balance model for a river basin in the Altai Mountains of northwestern China

Snowmelt water is a vital freshwater resource in the Altai Mountains of northwestern China. Yet its seasonal hydrological cycle characteristics could change under a warming climate and more rapid spring snowmelt. Here, we simulated snowmelt runoff dynamics in the Kayiertesi River catchment, from 2000 to 2016, by using an improved hydrological distribution model that relied on high-resolution meteorological data acquired from the National Centers for Environmental Prediction (Fnl-NCEP) that were downscaled using the Weather Research Forecasting model. Its predictions were compared to observed runoff data, which confirmed the simulations' reliability. Our results show the model performed well, in general, given its daily validation Nash–Sutcliffe efficiency (NSE) of 0.62 (from 2013 to 2015) and a monthly NSE score of 0.68 (from 2000 to 2010) for the studied river basin of the Altai Mountains. In this river basin catchment, snowfall accounted for 64.1% of its precipitation and snow evaporation for 49.8% of its total evaporation, while snowmelt runoff constituted 29.3% of the annual runoff volume. Snowmelt's contribution to runoff in the Altai Mountains can extend into non-snow days because of the snowmelt water retained in soils. From 2000 to 2016, the snow-to-rain ratio decreased rapidly, however, the snowmelt contribution remained relatively stable in the study region. Our findings provide a sound basis for making snowmelt runoff predictions, which could be used prevent snowmelt-induced flooding, as well as a generalizable approach applicable to other remote, high-elevation locations where high-density, long-term observational data are currently lacking. How snowmelt contributes to water dynamics and resources in cold regions is garnering greater attention. Our proposed model is thus timely perhaps, enabling more comprehensive assessments of snowmelt contributions to hydrological processes in those alpine regions characterized by seasonal snow cover.     

Using an inverse modelling approach with equifinality control to investigate the dominant controls on snowmelt nutrient export

There is great interest in modelling the export of nitrogen (N) and phosphorus (P) from agricultural fields because of ongoing challenges of eutrophication. However, the use of existing hydrochemistry models can be problematic in cold regions because models frequently employ incomplete or conceptually incorrect representations of the dominant cold regions hydrological processes and are overparameterized, often with insufficient data for validation. Here, a process‐based N model, WINTRA, which is coupled to a physically based cold regions hydrological model, was expanded to simulate P and account for overwinter soil nutrient biochemical cycling. An inverse modelling approach, using this model with consideration of parameter equifinality, was applied to an intensively monitored agricultural basin in Manitoba, Canada, to help identify the main climate, soil, and anthropogenic controls on nutrient export. Consistent with observations, the model results suggest that snow water equivalent, melt rate, snow cover depletion rate, and contributing area for run‐off generation determine the opportunity time and surface area for run‐off–soil interaction. These physical controls have not been addressed in existing models. Results also show that the time lag between the start of snowmelt and the arrival of peak nutrient concentration in run‐off increased with decreasing antecedent soil moisture content, highlighting potential implications of frozen soils on run‐off processes and hydrochemistry. The simulations showed TDP concentration peaks generally arriving earlier than NO₃ but also decreasing faster afterwards, which suggests a significant contribution of plant residue Total dissolved Phosphorus (TDP) to early snowmelt run‐off. Antecedent fall tillage and fertilizer application increased TDP concentrations in spring snowmelt run‐off but did not consistently affect NO₃ run‐off. In this case, the antecedent soil moisture content seemed to have had a dominant effect on overwinter soil N biogeochemical processes such as mineralization, which are often ignored in models. This work demonstrates both the need for better representation of cold regions processes in hydrochemical models and the model improvements that are possible if these are included.     

Modelling runoff using optical satellite remote sensing data in a high mountainous alpine catchment of Italy

Development of hydrological models for seasonal and real-time runoff forecast in rivers of high alpine catchments is useful for management of water resources. The conceptual models for this purpose are based on a temperature index and/or energy budget and can be either lumped or distributed over the catchment area. Remote sensing satellite data are most useful to acquire near real-time geophysical parameters in order to input to the distributed forecasting models. In the present study, integration of optical satellite remote sensing-derived information was made with ground meteorological and hydrological data, and predetermined catchment morphological parameters, to study the feasibility of application of a distributed temperature index snowmelt runoff model to one of the high mountainous catchments in the Italian Alps, known as Cordevole River Basin. Five sets of Landsat Multispectral Scanning System (MSS) and Thematic Mapper (TM) computer-compatible tapes (CCTs) were processed using digital image processing techniques in order to evaluate the snow cover variation quantitatively. Digital elevation model, slope and aspect parameters were developed and used during satellite data processing. The satellite scenes were classified as snow, snow under transition and snow free areas. A second-order polynomial fit has been attempted to approximate the snow depletion and to estimate daily snow cover areal extent for three elevation zones of the catchment separately. Model performance evaluation based on correlation coefficient, Nash–Sutcliffe coefficient and percentage volume deviation indicated very good simulation between measured and computed discharges for the entire snowmelt period. The use of average temperature values computed from the maximum and minimum temperatures into the model was studied and a suitable algorithm was proposed. © 1997 John Wiley & Sons, Ltd.     

Snow cover and runoff modelling in a high mountain catchment with scarce data: effects of temperature and precipitation parameters

Snowmelt is an important source of runoff in high mountain catchments. Snowmelt modelling for alpine regions remains challenging with scarce gauges. This study simulates the snowmelt in the Karuxung River catchment in the south Tibetan Plateau using an altitude zone based temperature‐index model, calibrates the snow cover area and runoff simulation during 2003–2005 and validates the model performance via snow cover area and runoff simulation in 2006. In the snowmelt and runoff modelling, temperature and precipitation are the two most important inputs. Relevant parameters, such as critical snow fall temperature, temperature lapse rate and precipitation gradient, determine the form and amount of precipitation and distribution of temperature and precipitation in hydrological modelling of the sparsely gauged catchment. Sensitivity analyses show that accurate estimation of these parameters would greatly help in improving the snowmelt simulation accuracy, better describing the snow‐hydrological behaviours and dealing with the data scarcity at higher elevations. Specifically, correlation between the critical snow fall temperature and relative humidity and seasonal patterns of both the temperature lapse rate and the precipitation gradient should be considered in the modelling studies when precipitation form is not logged and meteorological observations are only available at low elevation. More accurate simulation of runoff involving snowmelt, glacier melt and rainfall runoff will improve our understanding of hydrological processes and help assess runoff impacts from a changing climate in high mountain catchments. Copyright © 2013 John Wiley & Sons, Ltd.     

The importance of aspect for modelling the hydrological response in a glacier catchment in Central Asia

Understanding how explicit consideration of topographic information influences hydrological model performance and upscaling in glacier dominated catchments remains underexplored. In this study, the Urumqi glacier no. 1 catchment in northwest China, with 52% of the area covered by glaciers, was selected as study site. A conceptual glacier‐hydrological model was developed and tested to systematically, simultaneously, and robustly reproduce the hydrograph, separate the discharge into contributions from glacier and nonglacier parts of the catchment, and establish estimates of the annual glacier mass balance, the annual equilibrium line altitude, and the daily catchment snow water equivalent. This was done by extending and adapting a recently proposed landscape‐based semidistributed conceptual hydrological model (FLEX‐Topo) to represent glacier and snowmelt processes. The adapted model, FLEX^G, allows to explicitly account for the influence of topography, that is, elevation and aspect, on the distribution of temperature and precipitation and thus on melt dynamics. It is shown that the model can not only reproduce long‐term runoff observations but also variations in glacier and snow cover. Furthermore, FLEX^G was successfully transferred and up‐scaled to a larger catchment exclusively by adjusting the areal proportions of elevation and aspect without the need for further calibration. This underlines the value of topographic information to meaningfully represent the dominant hydrological processes in the region and is further exacerbated by comparing the model to a model formulation that does not account for differences in aspect (FLEX^G,nA) and which, in spite of satisfactorily reproducing the observed hydrograph, does not capture the influence of spatial variability of snow and ice, which as a consequence reduces model transferability. This highlights the importance of accounting for topography and landscape heterogeneity in conceptual hydrological models in mountainous and snow‐, and glacier‐dominated regions.     

General Secretary's Notes

Abstract

The following paper presents two different snowmelt submodels requiring a minimum input and suitable for estimating the monthly snowmelt-runoff in seasonally snowcovered watersheds. The first submodel discussed is based on the degree-day method, relating the variation of the degree-day factor in the melt season to temperature range as an index for the available radiation energy. The second submodel estimates the melt rate according to each of the components of the total energy contributing to the melt process of the snowpack. The submodels were used as a subroutine of a monthly water yield model in conjunction with a nonlinear optimization algorithm for identification of the parameters. The inputs used are the monthly precipitation, the average of daily air temperatures over a given month, and the extra-terrestrial radiation.     

Understanding the hydrological system dynamics of a glaciated alpine catchment in the Himalayan region using the J2000 hydrological model

This paper provides the results of hydrological modelling in a mesoscale glaciated alpine catchment of the Himalayan region. In the context of global climate change, the hydrological regime of an alpine mountain is likely to be affected, which might produce serious implications for downstream water availability. The main objective of this study was to understand the hydrological system dynamics of a glaciated catchment, the Dudh Kosi River basin, in Nepal, using the J2000 hydrological model and thereby understand how the rise in air temperature will affect the hydrological processes. The model is able to reproduce the overall hydrological dynamics quite well with an efficiency result of Nash–Sutcliffe (0.85), logarithm Nash–Sutcliffe (0.93) and coefficient of determination (0.85) for the study period. The average contribution from glacier areas to total streamflow is estimated to be 17%, and snowmelt (other than from glacier areas) accounts for another 17%. This indicates the significance of the snow and glacier runoff in the Himalayan region. The hypothetical rise in temperature scenarios at a rate of +2 and +4 °C indicated that the snowmelt process might be largely affected. An increase in snowmelt volume is noted during the premonsoon period, whereas the contribution during the monsoon season is significantly decreased. This occurs mainly because the rise in temperature will shift the snowline up to areas of higher altitude and thereby reduce the snow storage capacity of the basin. This indicates that the region is particularly vulnerable to global climate change and the associated risk of decreasing water availability to downstream areas. Under the assumed warming scenarios, it is likely that in the future, the river might shift from a ‘melt‐dominated river’ to a ‘rain‐dominated river’. The J2000 model should be considered a promising tool to better understand the hydrological dynamics in alpine mountain catchments of the Himalayan region. This understanding will be quite useful for further analysis of ‘what‐if scenarios’ in the context of global climate and land‐use changes and ultimately for sustainable Integrated Water Resources Management in the Himalayan region. Copyright © 2012 John Wiley & Sons, Ltd.     

Amplification of hydrological model uncertainties in projected climate simulations of the Upper Indus Basin: Does it matter where the water is coming from?

There is a growing appreciation of the uncertainties in the estimation of snow-melt and glacier-melt as a result of climate change in high elevation catchments. Through a detailed examination of three hydrological models in two catchments, and interpretation of results from previous studies, we observed that many variations in estimated streamflow could be explained by the selection of a best parameter set from the possible good model parameters. The importance of understanding changing glacial dynamics is critically important for our study areas in the Upper Indus Basin where Pakistan's policymakers are planning infrastructure to meet the future energy and water needs of hundreds of millions of people downstream. Yet, the effect of climate on glacial runoff and climate on snowmelt runoff is poorly understood. With the HBV model, for example, we estimated glacial melt as between 56% and 89% for the Hunza catchment. When rainfall was a scaled parameter, the models estimated glacial melt as between 20% and 100% of streamflow. These parameter sets produced wildly different projections of future climate for RCP8.5 scenarios in 2046–2075 compared to 1976–2005. Assuming no glacial shrinkage, for one climate projection, we found that the choice among good parameter sets resulted in projected values of future streamflow across a range from +54% to +125%. Parameter selection was the most significant source of uncertainty in the glaciated catchment and amplified climate model uncertainty, whereas climate model choice was more important in the rainfall dominated catchment. Although the study focuses on Pakistan, the overall conclusions are instructive for other similar regions in the world. We suggest that modellers of glaciated catchments should present results from at least the book-ends: models with low sensitivity to ice-melt and models with high sensitivity to ice-melt. This would reduce confusion among decision makers when they are faced with similar contrasting results.     

Assessing the hydrological and geomorphic behaviour of a landscape evolution model within a limits-of-acceptability uncertainty analysis framework

Landscape evolution models (LEMs) have the capability to characterize key aspects of geomorphological and hydrological processes. However, their usefulness is hindered by model equifinality and paucity of available calibration data. Estimating uncertainty in the parameter space and resultant model predictions is rarely achieved as this is computationally intensive and the uncertainties inherent in the observed data are large. Therefore, a limits-of-acceptability (LoA) uncertainty analysis approach was adopted in this study to assess the value of uncertain hydrological and geomorphic data. These were used to constrain simulations of catchment responses and to explore the parameter uncertainty in model predictions. We applied this approach to the River Derwent and Cocker catchments in the UK using a LEM CAESAR-Lisflood. Results show that the model was generally able to produce behavioural simulations within the uncertainty limits of the streamflow. Reliability metrics ranged from 24.4% to 41.2% and captured the high-magnitude low-frequency sediment events. Since different sets of behavioural simulations were found across different parts of the catchment, evaluating LEM performance, in quantifying and assessing both at-a-point behaviour and spatial catchment response, remains a challenge. Our results show that evaluating LEMs within uncertainty analyses framework while taking into account the varying quality of different observations constrains behavioural simulations and parameter distributions and is a step towards a full-ensemble uncertainty evaluation of such models. We believe that this approach will have benefits for reflecting uncertainties in flooding events where channel morphological changes are occurring and various diverse (and yet often sparse) data have been collected over such events.